Macroeconomics for Developing Countries

Macroeconomics for Developing Countries, 2nd Edition This comprehensively revised and updated edition develops the the...

Author: Raghbendra Jha

311 downloads 2736 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Macroeconomics for Developing Countries, 2nd Edition

This comprehensively revised and updated edition develops the themes contained in the first edition. Students and teachers who are familiar with the book will notice that half of the chapters are entirely new, with the other half having changed significantly to take into account the changes that have occurred in the global economy since the turn of the millennium. With questions for discussion and excellent use of case studies, the book covers such themes as: • • • • •

standard closed and open macroeconomic models a full evaluation of the post-Washington consensus model IMF stabilisation programmes and their effects on developing economies the pressing problems of indebtedness financial sector reforms in developing countries.

This informative, accessible and lucid textbook is the ideal accompaniment for students of development economics; not only that—it will prove popular with lecturers and academics alike. Raghbendra Jha is Professor at the National University of Australia. Another of his books, Modern Public Economics, is also published by Routledge.

Macroeconomics for Developing Countries, 2nd Edition

Raghbendra Jha

LONDON AND NEW YORK

First published 1994 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Second edition first published 2003 Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 1994, 2003 Raghbendra Jha All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been requested ISBN 0-203-42282-1 Master e-book ISBN

ISBN 0-203-42464-6 (Adobe eReader Format) ISBN 0-415-26213-5 (hbk) ISBN 0-415-26214-3 (pbk)

Dedicated to Mataji

Contents

Preface to this edition Preface to the first edition PART I

ix xiii

THE BASIC MACROECONOMIC FRAMEWORK

1

Macroeconomic problems of developing countries

3

2

National accounts and the macroeconomy

9

3

The basic IS-LM-AS model in the closed economy

27

The process of money creation and the demand for money

75

5

Macroeconomic policy in an open economy

93

6

Current account and asset demand approaches to balance of payments

139

MACROECONOMIC MODELS OF DEVELOPING COUNTRIES

175

IMF-type macro models for developing countries

177

1

4

PART II

7

viii

Contents

8 9 10

A structuralist macro model for developing countries

193

Dualistic models of output and inflation in developing countries

211

Growth theory and developing country macroeconomics

233

PART III POLICY DILEMMAS FACED BY DEVELOPING COUNTRIES 11

263

An evaluation of the IMF programmes in developing countries

265

Macroeconomic dimensions of fiscal policy in developing countries

285

13

The inflation rate and seignorage

315

14

The problem of indebtedness of HIPC countries

343

Exchange rate issues in developing countries

363

12

15

PART IV ISSUES IN DEVELOPING COUNTRIES’ MACROECONOMICS 16 17 18

385

Financial market liberalisation and economic growth in developing countries

387

The international financial architecture and the developing countries

419

A final word

457

Bibliography

463

Name index

487

Subject index

491

Preface

The first edition of this book was published in 1994. The enthusiastic response it received and the momentous developments in the global economy since then prompted me to write this second edition. An abiding concern of the previous edition was that there was a need to examine a variety of models in order to understand the macroeconomic performance of developing countries. Developing countries are at different stages of development and it would be simplistic to assume that one model suits all of them. Without much risk of oversimplification one can state that there is far more homogeneity in the macroeconomic structures of developed countries. Yet there is disagreement on the macroeconomic model to be used even in their case. It would seem natural then to conceive of alternative models for developing countries. This new edition takes this theme further. Apart from the versions of the structuralist models discussed in the first edition, the present edition also carries a fuller evaluation of the Washington consensus model (the IMF version) since the magnitude of the point of departure of the alternative models cannot be fully appreciated without a study of the dominant model. This new edition contains an exhaustive evaluation of IMF stabilisation programmes. In order to contrast this group of models with standard neoclassical doctrine as well as to build bridges with it, this new edition has a self-contained examination of both closed and open economy ‘standard’ models. The study of economic growth and monetary policy assignment has undergone important changes since the previous edition. Some of these important changes are reflected in this edition. There are separate chapters on economic growth as well as on inflation and monetary policy. These include discussions on ‘new’ theories of economic growth and the determinants of growth as well as issues of seignorage, credibility of monetary policy and inflation targeting. In view of the continued, even enhanced, emphasis on financial sector reforms, the links between financial liberalisation and economic growth are examined. In view of the recent emphasis on sustainability of internal and external debt of developing countries, there is a chapter on developing these notions of sustainability with applications to

x

Preface

select developing countries. In view of the persistent economic crises in heavily indebted countries, there is a chapter on the problems of heavily indebted poor countries. Students and teachers who are familiar with the first edition of the book will recognise that half of the chapters of the present edition are new. Even the other half have been thoroughly revised. Hence this edition marks a major departure from the first one. The plan of this book is as follows. Part I entitled The Basic Macroeconomic Framework’ consists of six chapters dealing with the standard closed and open macroeconomic models. Part II considers specialised models of developing country macroeconomics and has four chapters dealing with IMF and alternative structural models as well as theories of economic growth. The third section deals with policy dilemmas being faced by developing countries. These include dealing with IMF conditionalities, internal and external debt issues and the emergent problems facing heavily indebted poor countries as well as the choice of the exchange rate regime. The final section of the book is entitled ‘Issues in Developing Countries’ Macroeconomics’ and deals with issues of financial sector liberalisation and the international financial architecture. There is a final concluding chapter. This book is primarily intended to be used in upper level undergraduate courses in macroeconomics for developing countries. It would also be useful in standard intermediate courses in macroeconomics. Since the book covers very recent literature, it would be helpful to graduate students and research scholars as well. In writing this book I have run up a long list of debts far too many to mention; hence the following list is indicative, not complete. Many students in the Quantitative Development Economics course I taught at University of Warwick in 1996 gave excellent comments on the first edition, as did batches of students in the Macroeconomics course at IGIDR, Mumbai. Jerry Epstein of the University of Massachusetts was helpful through these revisions. In addition, I would like to thank C.Rangarajan, former Governor of the Reserve Bank of India, Ashima Goyal, Mridul Saggar, Deba Prasad Rath, Santosh Kumar, Anurag Sharma and Partha Ray from IGIDR for their thoughts. Priti Torvi at IGIDR and Stephanie Hancock at the Australian National University provided excellent secretarial support. Two anonymous Routledge reviewers gave very helpful comments. Comments—both published as well as those communicated to me—on the previous edition were insightful as well as encouraging. This book was written in its final form at the ANU. I would like to thank the University, particularly James Fox, Director of the Research School of Pacific and Asian Studies, and Warwick McKibbin, the convener of the Division of Economics, for providing such a congenial atmosphere for my work. I owe a debt of gratitude to Edmund Phelps who, as my PhD supervisor at Columbia several years ago, encouraged me to appreciate and take seriously the variety within macroeconomic thought

Preface

xi

and eschew intellectual regimentation. Robert Langham and his staff at Routledge have throughout been generous with their advice and support. My wife, Alka, and son, Abhay, have patiently seen me through yet another major book. Without their encouragement, love, and unflinching support through occasionally trying times, this book would not have been possible. I cannot even begin to thank them. Raghbendra Jha Canberra, Australia May 2002

Preface to the First Edition

Macroeconomics for developing countries has recently become important in the curricula of many universities. This subject, however, is usually taught at an advanced level either in graduate school or in a ‘Topics’ course at the senior undergraduate level. The rationale for this philosophy seems to be twofold. On the one hand, it is felt that students should have a thorough grounding in the basic tools of economic analysis before they study specialised topics such as less developed country (LDC) macroeconomics. Second, most of the material on LDC macroeconomics is available in esoteric papers in specialised journals and is therefore inaccessible to undergraduate students. In writing this book I have tried to bridge both these gaps. Students working through this book should be able to pick up the tools of modern macroeconomic analysis at the same time as they learn about the special macro issues confronting developing countries. It then becomes possible to present the advanced material pertaining to LDC macroeconomics as relatively straightforward extensions of the tools of analysis already learnt. In doing this I have tried to make the treatment of both the macroeconomic analysis and the special problems of developing countries exhaustive. We begin with elementary tools of analysis. In Part I of the book we develop the tools of analysis of the closed economy, and in Part II the open economy is studied. A good part of this material is common to ‘standard’ intermediate macro courses. However, issues of relevance to developing countries are brought up in key instances. In Part III we study some alternative views of LDC macroeconomics that have gained considerable prominence and respectability of late. We follow this up in Part IV with a review of some key policy questions confronting developing countries. The material in this part harks back to a long-held belief of LDC macroeconomic theory that long-term growth rather than short-term stabilisation is the key question facing developing countries. This book can be used in a variety of courses on macroeconomics. Parts I and II can be used for a ‘standard’ one-semester course on intermediate macroeconomics. In a one-semester ‘Topics’ course, Parts III and IV and

xiv

Preface to the First Edition

some sections of Part II can be used as material for studying special macroeconomic problems of LDC. Finally, the whole book can be used for a two-semester course on intermediate/advanced macroeconomics with special reference to developing countries. The prerequisites for reading this book are a first-year course in economics and a first-year course in mathematical analysis. I have relied primarily on graphical methods, and advanced mathematical tools, whenever they appear, have been fully developed and explained in the text. There is an exhaustive bibliography at the end to help students plan additional reading. In writing this book I have been the recipient of the help of several people. The economics department at Queen’s University has provided a congenial and stimulating environment for my work. Professors Robin Boadway, Martin Prachowny, Dan Usher and Isao Horiba have been especially helpful to me during my tenure here. (The late) Professor Douglas Purvis provided critical encouragement and number of ideas when this work was in its infancy. The comments of an anonymous referee provided absolutely critical input in restructuring and extending the original manuscript. The consistent and unflinching support of Alan Jarvis, economics editor of Routledge, throughout the three years over which this book was conceived and written is a pleasure to acknowledge. Dr Bagala Biswal was immensely supportive throughout and helped with the diagrams. Parts of this book were tried on students in a ‘Topics’ course at Queen’s University. Their forthright appraisals of the manuscript have been invaluable. I am deeply indebted to all these people for their help, patience and friendship. My greatest debt, however, is to my wife Alka and our son Abhay. I shall not be so presumptuous as to thank them. Raghbendra Jha Kingston, Canada

Part I

The Basic Macroeconomic Framework

1


AN INTRODUCTION This book is about macroeconomic problems of developing countries. Until quite recently, there was scarcely any independent analysis of such problems within the purview of mainstream macroeconomics. There were several reasons for this and two of them may be mentioned here. First, it was generally perceived that a traditional concern of macroeconomic theory— countercyclical stabilisation—was not very relevant to developing countries. It was felt that the most important concern of policy makers in developing countries was medium- and long-term growth and not short-term stabilisation. Second, there was a perception, almost up to the late 1980s, that the developing countries were not different from developed countries except for levels of per capita income and the like. The developed country represented to the developing country a mirror image of its future. To be sure, so far as macro problems of developing countries are concerned, there was a significant and distinguished line of theory called structuralism that had always challenged this point of view. It argued that the problem of developing countries were qualitatively different from developed countries and, therefore, required separate treatment. Structuralism provided alternatives to the received doctrine. More recently, structuralism has been enriched by significant contributions from economists such as Sweder van Wijnbergen and Lance Taylor. The most significant impetus to structuralism, however, was provided not by concerted intellectual debate (although that certainly played a role) but by a historical turn of events in the 1970s. The price of fuel oil quadrupled overnight and the whole trading world was dealt a very significant shock. Developed and developing countries alike faced ever increasing bills for imported fuel oil. Developed countries were more able to deal with this shock for several reasons. First, they produced goods that the Organisation of

4


Petroleum Exporting Countries (OPEC) needed and the prices of these goods could be increased to compensate to the sharp deterioration in the terms of trade. Second, OPEC placed much of its revenues from fuel oil exports in the banks of the developed countries. Third, developed countries had the technological capacity to develop fuel efficient processes and substitutes. None of these conditions was satisfied in the case of developing countries.1 The prices of less developed countries’ (LDCs) exports could not be increased substantially; indeed the prices of primary exports from many developing countries have stagnated, LDCs faced adverse balance of payments positions and their technological abilities were and are limited. The LDCs therefore faced a sharp deterioration in their terms of trade as the prices of fuel oil and goods imported from developed countries went up (see Gilbert, 1987). This led to a sharp drop in output potential, which the LDCs tried to counteract by increasing domestic money supply. The ensuing inflation and large balance of payments deficits sent LDC economies into deep crisis. They borrowed from international banks, since borrowing in home markets could only be very limited, and from the International Monetary Fund (IMF). The IMF imposed severe conditions on loans advanced to LDCs. They were required to devalue currencies, reduce import and other indirect taxes and reduce government expenditure at home. The rationalisation for these policy measures came from the monetary approach to the balance of payments—a theory developed to analyse the balance of payments problems of more developed economies. The logic behind this approach was as follows. Devaluation was necessary to make the price of foreign exchange realistic. In most LDCs it was felt that the ‘official’ exchange rate of the home currency was overvalued. Devaluation was often made a precondition for granting of assistance. Devaluation was to be followed by reduction in credit and reduction of indirect taxes and imported goods. The former, by controlling aggregate demand, would reduce inflationary tendencies. Reduction in indirect taxes would lead to more efficient allocation of resources, which would, in turn, improve aggregate supply. This step and devaluation would increase exports. Hence, inflation and the balance of payments deficit could be brought under control. This is the premise on which IMF conditionalities are based. The record of the performance of countries that had agreed to the IMF conditionalities, however, is somewhat mixed. The immediate impact was to cause hardships to LDC populations as social programmes and employment were sharply curtailed to accommodate lower ceilings on government expenditure. It was hoped that the set of policy measures advocated by the IMF would result in LDC economies becoming more competitive and efficient in the medium run. In some cases this did turn out to be the case but, in other instances, LDC economies experienced sharp stagflation—high unemployment and high inflation—for protracted periods.


5

The crisis due to the oil shock helped to focus attention on the theoretical underpinnings of the IMF policy prescription. The IMF had been following the point of view that it is not necessary to develop a separate theory to understand the macroeconomic problems of developing countries. It had been focusing almost exclusively on the demand side under the implicit assumption that supply side responses would improve under structural reforms. The structuralist school, with a completely different theoretical perspective, had been arguing that the IMF type demand stabilisation would only exacerbate stagflationary tendencies in developing countries since supply conditions would be adversely affected. The qualified success of the IMF stabilisation package gave the structuralist considerable support. The debate about whether a different theoretical approach was needed for studying the macro problems of developing countries gained credence. That debate is, as of now, still unresolved. However, the alternative point of view is now treated with considerably more respect. Several medium-term implications of the IMF stabilisation programme strengthened this. For instance, since the 1980s many of the poorest countries have kept accumulating large stocks of foreign debt. This accumulation shows no sign of abating with recent figures indicating that many of the poorest countries are now attracting a negative inflow—i.e., the debt servicing payments are far in excess of current financial inflows. This vast accumulation of debt far exceeds the capacities of these countries to pay with or without free trade. The situation is emergent and needs to be treated as such. If problems arising from international exposure for the poorest countries were not bad enough, even the richest among the developing countries started having serious problems in response to international exposure beginning in 1997. Many South-east Asian and East Asian economies, which had hitherto been variously described as ‘tigers’ and ‘newly industrialised’, went through a period of intense crisis. The reasons for this are many and varied but excessive short-term external borrowing in order to augment domestic savings for purposes of investment seemed to be the most prominent. The collapse engendered massive reduction in wealth and prosperity across the regions. Recovery from this has been slow and uneven and is yet incomplete. Some noted economists (including the Nobel Laureate Joseph Stiglitz), not necessarily of the structuralist persuasion, have argued that IMF type policies are wrong and that the consequences of such policy errors have been severe. However, as of yet, there is no consensus on what is the ‘right’ macro model for developing countries. PLAN OF THE BOOK In this book we do not adhere to one point of view or the other. We present the received macroeconomic theory and point out the differences, in key areas, brought about by considering the special characteristics of developing

6


countries. This is true of the first six chapters in Part I entitled ‘The Basic Macroeconomic Framework’. In this section of the book, then, the implicit assumption is that refinements to the received macro model are enough to understand the macro issues confronting the developing countries. Both closed and open economy versions of the received model are studied. The section begins with national accounts and elementary Keynesian economics. We then proceed, in Chapter 3, to the classical-Keynesian synthesis with the IS-LM and aggregate supply and aggregate demand models. We then generalise the IS-LM analysis to a variable inflation rate model. Chapter 4 discusses alternative models of money supply and Chapter 5 discusses open economy IS-LM analysis. Finally, Chapter 6 discusses alternative approaches to current account and asset market approaches to the balance of payments. A number of important policy issues such as demand management and the assignment of policy to target are also taken up. When dealing with these policy issues the special circumstances of developing countries are also taken into account. The second part of the book is entitled ‘Macroeconomic Models of Developing Countries’ and attempts to go into the underpinnings of alternative macroeconomic models for developing countries. It begins with an elucidation of the IMF’s approach to macroeconomic adjustment in developing countries. The next chapter discusses a variant of the standard structuralist model and points out the differences that this makes to the analysis of key policy questions. The model, as laid out, is aggregative and could be said to be representative of the experiences of some Latin American countries. This is followed by dualistic versions of the structuralist model. Two versions of the dualistic model are considered—one with a South Asian orientation and the other African. Many LDCs, particularly African ones, have a very significant agricultural sector. Typically, the agricultural/rural sector is qualitatively different in these countries from the nascent industrial sector. It would, then, be inaccurate to study the macroeconomic problems of these countries within the context of an aggregative model. We consider, therefore, a dual economy along these lines. We also consider a model of an LDC which is definitely dual but where the industrial sector is fairly important. These characteristics can be claimed to be true of South Asia. Finally this section reviews the implications of modern economic growth theory for the macroeconomic performance of developing countries. The third section entitled ‘Policy Dilemmas Faced by Developing Countries’ consists of five chapters. Chapter 11 is an evaluation of the IMF programmes designed for developing countries. Chapter 12 discusses fiscal policy and the macroeconomic implications of the twin deficits (fiscal and current account deficits) in developing countries. Chapter 13 considers the problem of inflation in developing countries and the incentive on the part of many LDC governments to earn revenue through an inflation tax. In Chapter 14 we turn to the pressing problems of indebtedness being experienced by many poor countries—the so-called Heavily Indebted Poor Countries (HIPC).


7

Chapter 15 considers alternative exchange rate regimes in developing countries. The final section of the book is entitled ‘Issues in Developing Countries’ Macroeconomics’. It begins (in Chapter 16) with an analysis of the role of financial sector liberalisation in aiding the growth process in developing countries. This is followed in Chapter 17 by an analysis of international financial architecture—an issue that has gained considerable credence since the Asian currency crisis of 1997. Chapter 18 concludes the book with some final remarks on international institutions and developing countries. DISCUSSION QUESTIONS 1

2

Make a list of important areas of differences between developed and developing countries. Both quantitative and qualitative issues can be considered. Empirically quantifiable issues could include differences in (a) per capita incomes; (b) degrees of monetisation; (c) the type of exchange rate regime; (d) structure of the economies; (e) skill composition of the labour force; (e) existence and efficiency of unemployment insurance and other welfare programmes; (f) structure of exports and imports; (g) level of technological attainment; (h) representation in international financial institutions such as the IMF, the World Bank, the Bank of International Settlements (BIS) and the like. Qualitative factors could include differences in (a) efficiency of fiscal and monetary policies; (b) perceived credibility of government and financial institutions; (c) extent of tax evasion; (d) maturity of political institutions and the like. On the basis of these differences argue the case for and against a separate macroeconomic treatment of developed and developing countries.

NOTE 1

Sah and Stiglitz (1989) present a useful analysis of the profound technological differences that exist between developed and developing countries.

2


INTRODUCTION Aggregate income accounting is the centrepiece of Keynesian macroeconomics. A key element of this analysis is the distinction between stocks and flows. A stock is a value of the macroeconomic variable under consideration at a point in time, say the money supply in Kenya on 1 January 2002 or the capital stock of Sri Lanka on 1 December 1900. A flow measures a rate per unit time. National income during the calendar year 2001 or additions to a country’s plant and equipment during 2001 or private consumption expenditure during 2001 are all examples of flows. In this chapter we will learn about the basic framework of national accounting. Just as private economic agents have balance sheets expressing the relations between their incomes and expenditures, a nation’s ‘books’ must also balance. This expression of the relationships between a nation’s output, income and expenditure is called the national accounts. These national accounts also have the additional role of providing a framework for macroeconomic analysis. This is crucial since without it macroeconomic analysis cannot proceed. In modern times, most countries follow the United Nation’s System of National Accounts, and the United Nations has a permanent statistical commission for this purpose. In addition, the International Monetary Fund (henceforth the IMF) has developed General Data Dissemination Standards (GDDS) and Special Data Dissemination Standards (SDDS) for countries to adhere to in reporting their national accounts statistics. See, for example, Bloem, Dippelsman and Maehle (2001). We now turn to a rudimentary account of the national accounting framework. AGGREGATE INCOME, OUTPUT AND EXPENDITURE There are five sets of economic actors in the macroeconomy: domestic firms, domestic households, the domestic government, financial institutions and firms, households and governments in the rest of the world, encapsulated

10


Figure 2.1

here under the heading of ‘Rest of the world’. The financial flows that take place between them are depicted in Figure 2.1. On this basis, economists often make distinctions between three different measures of aggregate economic activity: national income, national product and national expenditure. N.B. Double arrows (↔) indicate Financial flows between the Rest of the world and Domestic economic agents. Thus domestic households may invest in foreign stock markets and foreign institutional investors may invest in domestic stock markets. In modern macroeconomics the economy is thought of as consisting of a number of firms and households interacting with each other and with the government. These institutions, in turn, interact with foreign households, firms and governments. This is what is portrayed in Figure 2.1. During this interaction output is produced, incomes are generated and expenditures are made. These three activities form the core of national accounting. We describe each of these very briefly now. In addition, we comment on the nation’s accounts with the rest of the world—the balance of payments. National product The sum of value added during an interval of time at various stages of production is a measure of a nation’s output. Value added at any stage of production is the value of output produced at that stage of production minus


11

the cost of intermediate inputs. Value added at the bakery is the value of bread minus the cost of flour and other intermediate inputs used in the production of bread. The sum of all value added is sometimes called the gross domestic product (GDP) of a country. The adjective domestic is used here because all value added within the geographic boundaries of the country are considered irrespective of whether these values are produced by foreign or domestic firms. The emphasis in calculating value added is on the production of final goods and services produced and sold through the market during the current time period but not resold during the current time period. The goods and services produced are valued at market prices (this is what is meant by saying sold on the market). This recognises, for instance, that a new house costs ten times as much as a new car. Hence the value of the house is assessed to be ten times the value of the car. GDP refers to final goods and services since intermediate goods (that are used in the production of other goods) are netted out. In measuring gross national product (GNP) we deduct from GDP the value added by foreign-owned firms and add to it the value added by domestic firms working in foreign countries. It should be emphasised that when measuring the national product the values of goods and services as expressed in the market are used. Omitted are important non-market activities. Thus the services a housewife provides are not measured in the national accounts whereas if the same services had been provided by an employee these would have been included. Furthermore, national accounts typically ignore the impact of economic activity on the environment. To the extent that economic activity causes a net loss to environmental resources, current economic activity reduces future economic potential and this fact should be reflected in the value of current output. 1 However, in macroeconomics such issues are generally eschewed not because they are unimportant but because the agenda of macroeconomics is very full otherwise. National income This is the sum of all factor incomes generated during an interval of time. In computing this we add up all wages, salaries, profits, interest, rent, dividend incomes and the like, i.e. all remuneration received for productive economic activity2 by all factors of production. This is called net national income at factor cost. If we add to it indirect taxes and subtract subsidies we get national income at market prices. If depreciation is also added we get gross national income. National expenditure A third way of measuring aggregate economic activity during an interval of time is by summing up final expenditures. Hence we add up consumption,

12


investment and government expenditures. To this we add net exports, i.e. exports minus imports. Consumption is defined as the direct utilisation of goods and services by consumers not including the use of means of production such as machinery and factories. Typically, this is the largest component of national expenditure in all countries. Consumption is divided into three major categories: consumption of durable goods such as cars or furniture, consumption of non-durable goods such as food, and consumption of services such as health and financial services. Investment expenditure is defined to include all expenses incurred in augmenting the productive capacity of the economy—say in plants and equipment. Thus final goods that firms and businesses keep for themselves (and do not consume) are called business or private investment. Expenditure on residential investment such as the construction of homes and apartment buildings is part of investment. Business investment includes investment in plants and equipment as well as changes inventories. By including changes in inventories during a given time period as part of expenditure we ensure that national product equals national expenditure. The third major component of national expenditure is government expenditure. In this category are included expenditures by government—national, state, local—on final goods and services. Also included are transfers made by governments. These include items like social security payments. The final category of national expenditure is net exports—exports less imports. Net exports are a source of net demand (hence expenditures) for the home economy. Net exports are also part of balance of payments accounts, which are discussed later in this chapter. These three measures of aggregate economic activity are usually calculated for a year.3 Hence these are all flows. When a year closes national income must equal national product, which in turn must equal national expenditure. This is because they are realised magnitudes. Desired magnitudes, however, might differ systematically from realised ones. Each of these magnitudes can be measured in real or nominal terms. When the GNP for 2001 is measured with the prices of 2001 used to value output, then we have measured GNP in current prices. This is also called nominal GNP. If, however, we wish to compare GNP figures of 2001 with those of 1980, we realise that there are some difficulties. Comparing nominal GNP for 1980 with nominal GNP for 2001 may not be such a good idea, because during the period 1980–2001 the price level would have risen and the structure of relative prices would have undergone several changes.4 One way of getting around this difficulty is to calculate real GNP. Let be the year 0 (or base year) prices of all n goods produced in the economy. The corresponding quantities transacted in the base year are We may evaluate current year quantities in terms of base year prices. Let the current year quantities traded be Real GNP for year 1 (with year 0 as base) can be written as


13

This measure has obvious disadvantages, such as the fact that the same bundle of goods may not be produced in year 1 as in year 0, but in macroeconomics we must learn to be pragmatic rather than fussy. We can similarly compute an index of how prices may have moved during the period. One such indicator is

Similarly one can define a GNP deflator (a measure of price change) as nominal GNP/real GNP. Let P2001 be the value of the GNP deflator in 2001 and P1980 be its value in 1980. The rate of inflation between 1980 and 2001 was

Price index numbers and GNP in real and nominal terms are some of the most important macroeconomic variables. Another important macroeconomic variable is the aggregate rate of unemployment. This is defined as the aggregate number of people unemployed and actively searching for work during a period of time divided by the total size of the labour force during that period. Unemployment data are also collected annually but are available for shorter durations of time as well. BALANCE OF PAYMENTS ACCOUNTING The balance of payments accounts of a country record all payments into and out of the country within a given time period (typically a year). All flows— financial or consisting of goods and services—between the residents of a country and the rest of the world are so recorded. The balance of payments accounts are made up of two major components each with its own subcomponents. The two major groups are: the current account and the capital account.5

14


The current account is divided into four major categories Transactions in goods Transactions6 in goods are subdivided into four major categories: (a) general merchandise consisting of the bulk of commodities included in the balance of payments. (b) Goods for processing are those that are imported with the intention of being re-exported after being transformed in some way through processing or incorporating into another commodity. (c) Repairs on goods and services includes repairs on movable property (such as cars and aircraft) without such property being physically transformed. (d) Non-monetary gold includes exports and imports of gold not held as reserve assets by the central bank of this country, e.g. the gold in jewellery. Services Transactions in services are divided into several categories. (a) Travel. All goods and services, irrespective of type, are classified under this heading if visitors to the economy compiling the balance of payments statistics purchase them (credit) or if residents of this economy visit other countries (debit). (b) Government services. A good or service is recorded under government services if embassies or military units purchase it in the country of location of the embassy or the military unit. (c) Transportation. This consists of costs of transportation of passengers and goods and auxiliary services such as storage, cargo handling, packing and towing. Thus this category includes all international transportation services and transportation services rendered by a non-resident to a resident user. (d) Computer and information related services. Items classified in this category include software implementation, maintenance and repairs of computers, data processing, newspapers and magazine subscriptions. (e) Insurance data includes figures on premiums earned, premium supplements earned and claims. (f) Personal, cultural and recreational services include items in audio-visual, entertainment, sports, museums, libraries and health. Income covers international transactions associated with income accruing to factors of production such as capital and labour. Income on capital is referred to as income on financial assets. Current transfers record unrequited transactions, i.e., those transactions in which an economic agent provides another economic agent with a good or service or a financial asset without any flow in the opposite direction.


15

The capital account The capital account is divided into capital transfers, acquisitions/disposals of non-produced, non-financial assets and the financial account. Items recorded under (a) capital transfers include the following: investment grants (including cash transfers for purchases of investment goods), debt forgiveness and migrants’ transfers. (b) Non-produced non-financial assets are, in the main, licences, franchises and patents. Also included is the acquisition/ disposal of land by representative offices of foreign governments, such as embassies. (c) The financial account records transactions in financial assets and liabilities. Such transactions can be divided into four broad categories. Direct investment includes investment over which the owner exercises control. In practice a distinguishing criterion for inclusion in this category is that the owner should hold at least 10 per cent of the ordinary shares in the company. Portfolio investment is a residual category for transactions in shares, bonds, bills, notes, money market instruments and financial derivatives. It is residual because these instruments are also included under direct investment and reserve assets. Portfolio investment is typically divided into the categories of equity and debt. Other investment lists all transactions in financial assets and liabilities that are not classified under any of the other categories. The most important of these ‘other investments’ are currency, deposits and loans (including trade credits). Reserve assets include gross assets to be used by the central bank of the country for balance of payments purposes where other assets cannot be used. Any asset can be included in the reserve assets category provided it is controlled by the central bank or accessible to the central bank at short notice and is denominated in a convertible currency.7 Residence is defined in economic and not legal terms. The main criterion to determine residence of an entity is centre of economic interest. Thus a domestic branch of a foreign owned firm will be counted to be in residence whereas the foreign branch of a domestic owned firm will typically not. We have so far examined some key flow macroeconomics variables. Let us now look at some important stock variables. Macroeconomists are concerned about several stock variables, e.g. the capital stock or the stock of money in the economy. MONEY Money is anything that is generally accepted as a medium of exchange. What precise assets constitute the medium of exchange varies from one society to another and has varied over time. Gold, silver and several other precious metals have served as mediums of exchange. In developed countries personal cheques and interbank transfers are accepted as mediums of payment. In several developing countries payment by personal cheques is often unacceptable. The use of cash is much more common.

16


In modern societies money is typically a financial asset, which is the liability (most commonly) of the central bank of the country. There are several definitions of money but two are most common. Narrow money (M1) consists of currency in circulation plus demand deposits in commercial banks. Broad money (M2) consists of M1 plus time deposits in commercial banks. Money does not include ‘plastic money’ such as credit cards. These plastic cards are convenient identification tags that enable the holder to create two debts. One debt is between you and bank issuing the credit card and the other is between the credit card bank and the seller whose goods have been bought. Capital stock In economics, capital is referred to as a factor of production (other than land or labour) from which an income is derived. In modern macroeconomic theory capital is deemed to consist of physical capital of tools, machines, stores of merchandise, houses, means of transportation—any materials used to extract, transport, create or alter goods. Marketable intangibles, such as credits, goodwill, promises, patents and franchises, are also included by some economists. Capital goods are those that form a nation’s productive capacity, as opposed to consumer goods, which are bought for personal or household use. A distinction is also made between capital stocks, or circulating capital (such as raw materials, goods in process, finished goods and sometimes wages), and capital instruments, or fixed capital (such as machines, tools, railways and factories). Capital may be classed as specialised, such as railway equipment, or unspecialised, such as lumber or other raw materials having many uses. Economic theorists believe that capital arose out of the need to use the world’s limited natural materials efficiently. The scarcity of the earth’s resources necessitates the creation of materials (capital) that can act on the resources in such a way as to make more goods available to society than would normally exist. Capital goods can be considered a form of deferred consumption, because they produce goods for future consumption, but are not themselves consumable items. The expansion of capital formation—the flow of savings into the creation of new productive facilities—is often the most important target of economic planning. ASSETS, LIABILITIES AND BALANCE SHEETS An asset is something you own. A liability is something you owe. There are two types of assets: financial and real. A real asset is something tangible: cars, plots of land, mines, airplanes, apartments and steel factories are all examples of real assets. Financial assets are different. They are pieces of paper, which constitute an asset for one person and a liability for another. In other words they define a debt relationship between two economic agents. An example is your savings


17

Table 2.1 An example of a personal balance sheet

account in your neighbourhood bank. The amount of money in your account is your asset—you own the money. It is the bank’s liability—they owe that money to you. A $5 currency with you is your asset—the corresponding liability is that of the Federal Reserve Bank of the USA. The balance sheet of an economic agent is a comprehensive statement of that agent’s assets and liabilities. An individual’s balance sheet may look like Table 2.1. The balance sheet drawn here follows the principle that the sheet must balance. Hence wealth is shown as a liability. One can think of it as what the individual owes to herself. BALANCE SHEET OF SIX AGENTS We may now examine the balance sheets of six major economic actors: firms (F), households (H), commercial banks (B), central bank of the country (C), the government (G), and the rest of the world (R). Their balance sheets are represented in Table 2.2. A plus (+) denotes an asset and a minus (-) a liability. An explanation of the terms used in the table is given below. 1

2

3

Commercial bank deposits with the central bank are commercial banks’ accounts with the central bank of the country. These are the central bank’s liability and the commercial banks’ assets. Currency consists of all the notes and coins held by (and, therefore, assets of) households, firms and commercial banks. The corresponding liability is that of the central bank. Demand deposits are accounts from which funds may be withdrawn on demand. They constitute a liability of the commercial banks and are assets of firms, households and the government.

18


Table 2.2 Structure of financial indebtedness

4

Other (term) deposits are held at commercial banks by firms and households. They may not be withdrawn immediately. 5 Government securities are loans issued by the government and held by many agents; in the typical developing country they are held principally by the central bank of the country. 6 Bank loans are personal and business loans, which form an asset for the commercial banks, and are a liability of the people who have borrowed the funds. 7 Equities are issued by funds to raise capital for their activities. An equity holder in a firm is in fact an owner of the firm, i.e. the householders and foreigners who own equity shares in a firm really own a share of the firm’s capital stock. Actually, of course, the owner of a share can only sell the share. The firm has a liability and the households and firms own the corresponding asset. 8 Corporate bonds are held by bondholders in a corporation. A bondholder, unlike an equity holder, is not an owner of the corporation. Rather such a patron has simply made a loan to the corporation. A corporate bond is a liability to a firm and an asset to those advancing the loan (households and foreigners). 9 Foreign securities can in principle be held by the households, domestic firms, commercial banks and the government. In practice, however, in most developing countries foreign securities are held by the central bank and the government. 10 Foreign exchange reserves in developing countries are typically assets of the government and the central bank and the liabilities of the foreign governments that have issued them.


19

NET FINANCIAL ASSETS If we add up the elements along any column of Table 2.2 we shall get the net financial assets of that sector. Those for the commercial banks and the central banks will approximately add up to zero. This is because although the absolute size of the real assets owned by these institutions is very large, in comparison with the size of their financial liabilities it is minuscule. Households and firms have positive net financial assets. The government, on the other hand, typically has a net financial liability referred to as the national debt. The net financial asset position of the country vis-à-vis the rest of the world may be positive or negative. REAL AND FINANCIAL ASSETS Real assets are owned primarily by households, firms and government, and the total of all these constitutes the non-human wealth of the economy. If the government has liabilities that exceed its real assets then it is the household and firms that will be responsible for meeting these liabilities. The government will have to levy taxes on them, which equal in value the excess of its liabilities over its assets. This might be thought of as an implicit financial asset. Households are the ultimate holders of wealth. Firms can be regarded as owing to households net undistributed profits, which are exactly equal to the difference between the firm’s real assets and net financial liabilities. In the case of firms that have issued equity, undistributed profits are already taken into account provided that the equity has been valued correctly. The value of the future income of individuals constitutes the economy’s human wealth. The net wealth or the real assets plus human wealth is the economy’s wealth. This is the same thing as the household sector’s wealth. The reason why the households own all the wealth is because of the implicit asset/ liability items, which take into account future tax liabilities and undistributed profits. Government has no wealth of its own. It owes any excess of assets over liabilities to households who are liable for its net debts. Similarly, firms have no net wealth because they owe to households any undistributed profits. These ideas are summarised in Table 2.3. Table 2.3 Net asset positions of economic agents

20


THE NATIONAL ACCOUNTING FRAMEWORK Aggregate demand in an economy can be written as the sum of consumption expenditure C, investment expenditure I, government expenditure G and net sales to the rest of the world (exports X less imports Z). We know that ex post aggregate expenditure equals national income y. Thus we can write

Domestic expenditure is often referred to as domestic absorption A and is called the balance of trade, so that we may write (2.1)

or (2.2)

In other words the balance of trade is equal to the difference between national income (which in equilibrium equals domestic supply) and domestic demand (absorption). Realise that if net factor payments from abroad are zero then also equals the current account balance. Now any current accounts surplus (CAS) is equal to the net acquisition of foreign assets (∆NAFA). Thus

upon substitution for A. Now

(savings). Hence we have (2.3)

Hence the trade account can improve only by increasing net savings. Equivalently the trade account can only be increased by increasing national income or reducing absorption. SIMPLE MULTIPLIER ANALYSIS Our analysis so far has been conducted in terms of the accounting framework. In the simple Keynesian framework these relationships have been used to predict casual linkage between aggregate demand and national income. Let us dwell in this somewhat. Consider an economy in which the absolute price level, the nominal wage rate and the interest rate are fixed. The price level and the nominal level wage rate may be assumed fixed because we are considering only the short run—a time period too short for any significant wage and price adjustments. We abstract from interest rate effects by assuming that the monetary authorities keep the interest rate pegged. The rate of interest is allowed to vary in the IS-LM model, which is discussed in Chapter 3.


21

Desired consumption C depends on disposable real national income y . d Keynes hypothesised that with higher disposable income people consumed more. But part of every additional dollar of disposable income is saved. The consumption function relates desired consumption in the aggregate to disposable income. A particularly simple example would be (2.4)

where b is often referred to as the marginal propensity to consume (MPC). The average propensity to consume (APC) is defined as the ratio of consumption to income, C/yd, and varies inversely with yd. Real national income equals disposable income plus direct (restricted to income) taxes Hence if T is a relatively simple magnitude it is easy to convert the consumption function from the (C, yd) plane to the (C, y) plane. We shall assume that this is the case. Desired investment I is treated as exogenous. The rate of interest is fixed. Other factors such as expectations of profitability that might affect desired investment do not depend on y. Government expenditure G is treated as exogenous since the government’s expenditure is affected by factors other than y. Finally, exports X are exogenous and imports Z depend on imports. However, the marginal propensity to import (the additional import following an increase in income of $1) is assumed fixed and less than one. Total desired aggregate expenditure E is the sum of desired consumption, investment and government expenditure: (2.5)

Algebraically, let

The equilibrium condition is that national expenditure equals national income. Thus,

Hence equilibrium national income is (2.6)

22


This equation expresses equilibrium national income as a function of exogenous variables and parameters of the system. Suppose we wish to model taxes explicitly. Let Then we would have the consumption function The other components of aggregate demand are as before. Equilibrium national income would be given as or (2.7)

Where y** is equilibrium national income with taxes explicitly modelled. The multiplier is defined as the change in equilibrium national income consequent upon the change in some exogenous component of aggregate expenditure. From equation (2.7) we get

Consider now the simplified model with lump-sum taxes:

We will get so that

so that and whence dy/T=(1-b)/(1-b)=i.e. if G and T change by the same amount (say $1) output would increase by $1. This is often referred to as the balanced budget multiplier.


23

There is another interesting interpretation of the equilibrium condition:

Savings S are the surplus of disposable income over consumption,

Now

Now if (2.8)

we get our equilibrium condition that aggregate income equals aggregate expenditure. Now is the sum of leakages from the income stream whereas are the injections. Hence equilibrium is attended where injection into the income stream equal leakages. In the analysis in this chapter changes in demand drive the macroeconomy. A higher exogenous demand always leads to a higher income. Other critical macroeconomic variables like the price level, real wage and interest rate do not adjust. More output always means more employment from the production function. More workers are always willing to work at the prevailing wage rate. Output demand constrains labour demand, which in turn, constrains output demand. The labour market equilibrium consistent with this view of the economy is depicted in Figure 2.2. The labour supply curve is horizontal at the going wage rate. Employment is constrained by the demand for labour. The modelling of the labour market is very simplistic here. A more sophisticated analysis is definitely possible. However, it needs to be remembered that in

Figure 2.2

24


most developing countries labour is plentiful and output is constrained not by labour so much as by availability of foreign exchange and capital. The assumption about the labour market made in this chapter, therefore, may not be as outlandish as it may seem at first glance. CONCLUDING REMARKS The analysis in this chapter is the starting point of almost all texts on macroeconomics, even those that have the most developed countries as their focus. Does this mean that the same analysis can be used for developed and developing countries? This question is rather complicated and controversial. Complicated because there are many differences in the economic structures of developed and developing countries. To name just a few of these differences: developed countries, apart from the fact that they enjoy much higher living standards than developing countries, also have much more developed financial and banking institutions, their currencies enjoy much more confidence than the currencies of developing countries in exchange markets and they have much better supply responses than developing countries. Many economists believe that these differences are quantitative rather than qualitative in nature. In other words the same economic mode with different values of the parameters can be used for developed as well as developing countries. The aggregate supply curve, for instance, is upward sloping in developed as well as developing countries. The slopes might be different. Several other economists, however, argue that the differences between developed and developing countries are so deep that they amount to being qualitative rather than merely quantitative. They argue that models used to study the macroeconomic problems of developed countries are inapplicable to developing countries. A whole new approach has to be developed. This debate is controversial because, after all, the economic policy measures that one advance would depend on the model used to study the economy. The approaches of the two groups of economists to economic policy are fundamentally different. Given the severe economic predicament in which many developing countries have found themselves since at least the mid1970s, this controversy assumes significance. The approach taken in this book is agnostic. We do not necessarily subscribe to the point of view of any school. We shall develop both approaches and, whenever possible, resolve differences by appealing to the facts. DISCUSSION QUESTIONS 1

The simple Keynesian model studied in this chapter implies that an increase in autonomous expenditure leads to a rise in income. What


2

3

4

25

assumptions do you think are necessary for this to hold? To what extent do you think these assumptions are satisfied in developing countries? We have studied the balanced budget multiplier in this chapter. Can you point out other instances when increasing one component of autonomous expenditure and reducing another will imply a multiplier of one? When are such multipliers equal to zero? Suppose the demand for money depends upon consumption rather than income. Consumption, in turn, depends upon after-tax income. Analyse now the impact of a tax cut on equilibrium output and employment. Suppose the central bank is considering two alternative policies—holding the money supply constant or adjusting the money supply to hold the interest rate constant. In the fixed price IS-LM model which policy would be better if there were (i) shocks to consumption only, (ii) shocks to money demand only? The classic paper analysing this question is Poole (1970).

NOTES 1 2 3 4 5 6 7

Such considerations have led to the notion of ‘green national accounting’. See, for example, Aronsson, Johansson and Lofgren (1997). Payments for intermediate goods are not included. Recently the Special Data Dissemination Standards of the IMF have called for quarterly figures on GDP as part of its Special Data Dissemination System (SDDS). See the IMF website: www.imf.org. An additional problem would be that the basket of goods in the market would have changed between 1980 and 2001. In some classifications, the capital account is split between a capital account and a financial account. Transactions are divided as acts in which two parties exchange goods, services or assets. The two parties need not be separate legal entities. Thus two branches (in two different countries) of the same firm may enter into a transaction. Part of these are special drawing rights (SDR) issued by the IMF which the central bank of a country holds to settle claims between itself and other central banks.

3


INTRODUCTION In this chapter we shall further develop the Keynesian analysis initiated in the last chapter. We begin with an IS-LM analysis with fixed prices. We then relax the assumption of fixed prices and study output and price level determination in a model of aggregate demand and supply. We then generalise the model to include wealth effects in aggregate demand. Finally we study Phillips curves and the natural rate of unemployment. MACROECONOMIC ANALYSIS WITH FIXED PRICES: IS-LM In the previous chapter we developed the Keynesian model for the case where both the interest rate and the price level are fixed. In this section we relax the assumption of a fixed interest rate but keep the price level fixed. This is called IS-LM analysis. The IS schedule To develop the IS curve, consider the model developed in Chapter 2. Equilibrium national income is given by (3.1)

In the model presented in the previous chapter I, G and X were treated as exogenous. Hicks, and later Hansen, in an attempt to generalise this model to include some concerns of the classical economists, allowed investment to depend negatively on the rate of interest, r. (Since prices are fixed there is no difference between the real and the nominal rates of interest.) Thus (3.2)

A simple rationale for the investment function postulated in (3.2) is that the higher rate of interest the greater is the cost of borrowing for investment,

28


ceteris paribus, and the lower, therefore, will be investment. Equilibrium national income is given by (3.3)

or, equivalently, (3.4)

Equation (3.4), or equivalently (3.3), is the equation of the IS schedule. This schedule gives us the r, y combinations for which leakages form the stream of national expenditure equal injections Totally differentiating equation (3.4) gives

for fixed T, G and X. A subscript denotes a partial derivative: Hence the IS schedule in the r, y plane is downward sloping,

The intuition behind the IS schedule is rather straightforward. The lower the rate of interest, ceteris paribus, the higher is going to be aggregate demand. With higher aggregate demand, equilibrium national income is going to be higher. We depict the IS curve in Figure 3.1 as the locus of combinations of combinations of r and y. To the left of the IS schedule at any level of income the rate of interest is too low to give us commodity market equilibrium. If the rate of interest is ‘too low’ then the rate of investment will be too high. Correspondingly, aggregate demand will be too high. This last step in the argument follows from the ‘Keynesian’ assumption retained by Hicks and Hansen that output and employment respond to aggregate demand changes in the commodity market. All points to the left of the IS schedule then denote r, y combinations that are associated with excess demand (EDG) in the commodity market. Analogously, at any point to the right of IS, at the relevant level of income, the interest rate is ‘too high’ to maintain equilibrium in the commodity market. This implies that investment is too low and so will be aggregate demand. All points to the right of IS then denote r, y combinations that are associated with excess supply (ESG) in the commodity market. This is shown in Figure 3.1. Horizontal arrows in the diagram denote pressure on y. When there is excess demand for goods there is pressure for y to rise. Conversely, when there is excess supply in the goods market there is pressure for y to fall. The IS curve gives a relationship between r and y that must obtain to have equilibrium in the commodity market. It is clear that, by itself, the IS curve cannot determine equilibrium r and y. To do this we have to introduce another


29

Figure 3.1

equilibrium relation between r and y. To get this other relation let us dwell a little on the general equilibrium underpinnings of the Keynesian model. We have four aggregate markets: the labour market, the commodity market, the bond market and the money market. Keynes simplifies the analysis by assuming that the wage rate is fixed in the labour market. Employment, then, is entirely dependent on the demand for labour that is derived from the demand for goods. Thus employment rises when demand for goods rises and, correspondingly, falls when the demand for goods falls. Hence the labour market’s behaviour is entirely explained by the commodity market. We exclude the bond market by Walras’ law (which says that excess demand in any market in an n–1 market system is the sum of the excess supplies in the other market). Hence, we get the other relationship between r and y (the LM schedule) from the money market. The LM schedule Keynes had made a spirited departure from the classical tradition by assuming that monetary influences determine the rate of interest. In the classical tradition, r is determined by the equality of real saving and investment, and monetary influences primarily affected the price level. Keynes’ major departures from classical thought (at this level of aggregate general equilibrium analysis) are 1) that the price level and nominal wages are fixed and national income adjusts to clear the commodity market, and 2) that monetary influences determine the rate of interest. Keynes visualised the demand for money (real cash balances) as rising from the transactions, speculative and precautionary motives. People hold

30


cash balances for transactions purposes because of the immediate liquidity of money and because of the non-synchronisation of payments and receipts. People get salaries at discrete intervals of time, usually a month, whereas expenses have to be incurred more or less continuously. These expenses have to be paid for with cash. Hence people hold money for transactions purposes. We might reasonably expect that the higher a person’s income the higher would be expenses and the greater would be the transactions demand for cash. For the economy as a whole the higher the national income the greater would be the transactions demand for cash balances. People often hold cash for precautionary purposes—for an unforeseen contingency that requires extra expenses—an illness in the family, for example. We would reasonably expect the higher a person’s income the higher would be the demand for cash for precautionary purposes. For the economy as a whole the higher the national income the greater would be the demand for cash for precautionary purposes. People demand money for speculative purposes as well. Suppose that people can hold their financial assets in two forms: a riskless asset (cash) and risky asset (bonds). For simplicity the bond is assumed to be an annuity— the bond pays an interest of r and the principal is never paid. The price of the bond, therefore, is 1/r: the higher the rate of interest the lower will be the price of bonds. If we assume that at high interest rates people will expect the interest rate to fall (bond prices to rise), it would be rational for an investor to buy more bonds and reduce the demand for money. Analogously, when interest rates are low (bond prices are high) investors will expect interest rates to rise (bond prices to fall) and will therefore tend to sell bonds to make capital gains. In this process they will increase their demand for money. Therefore the demand for money for speculative purposes depends negatively on the interest rate. Hence we may write the aggregate demand for money as L(r, y). When the interest rate rises, ceteris paribus, the demand for money will fall. When y rises, ceteris paribus, the demand for money will rise. Equilibrium in the money market is given by (3.5)

where M is the nominal stock of money and P is the aggregate price level. M/ P, therefore, is the real value of cash balances. Equilibrium in the money market is shown in Figure 3.2. If the real value of existing money supply is (M/P)0 and the national income is y0, the equilibrium rate of interest in the money market would be r0. At a lower rate of interest, ceteris paribus, demand for money would be too high and at a higher rate of interest it would be too low in comparison to the existing money supply.


31

Figure 3.2

Equation (3.5) represents the LM schedule—the locus of r, y combinations giving us equilibrium in the money market. Totally differentiating we have (for fixed M/P) so that

since

and

Figure 3.3 displays the LM schedule. Above the LM schedule for a given y the interest rate is too high. This means that the demand for money will be too low relative to money supply. This is labelled as ESM. If, along with Keynes, we assume that monetary factors affect r (primarily) then there will be pressures for r to fall in this region. Vertical arrows in Figure 3.3 denote this. Below the LM schedule at any level of income the rate of interest is too low to maintain equilibrium in the money market. There will be excess demand for money (labelled as EDM) and there will be pressures for r to rise. Equilibrium in the commodity and money markets Figure 3.4 denotes simultaneous equilibrium in commodity and money markets with (at point Z) the equilibrium interest rate being r* and equilibrium national income y*. At this point there is simultaneous equilibrium in the goods and money markets. Students should satisfy themselves that the bond market equilibrium schedule would go through quadrants I and III and pass through

32


Figure 3.3

Figure 3.4


33

Z, by Walras’ law. In any event, the bond market schedule is redundant and we shall work with the IS and LM schedules. Shifting the IS schedule Assume that we have equilibrium at Z in Figure 3.4 with rate of interest r* and national income y*. A rise in any exogenous component of aggregate demand would shift the IS curve to the right. To see this, suppose government expenditure goes up. Now totally differentiate equation (3.4) with the understanding that This gives At any given r we have Similarly for given y we have With higher government expenditure aggregate demand is higher, ceteris paribus, at every rate of interest. Hence every rate of interest is associated with a higher level of national income (correspondingly at every level of income the interest rate required to give us equilibrium in the commodity market is higher). Consequently the IS curve shifts outwards. The analysis here is not confined to an increase in government expenditure: an increase in any component of aggregate demand would shift the IS curve outwards. Conversely, any fall in any exogenous component (an increase in T) will shift the IS curve inwards. Shifting the LM curve A change in money supply will shift the LM curve. Totally differentiating equation (3.5) with the understanding that gives d(M/P). Hence for given r we get Hence with an increase in the money supply every interest rate is associated with a larger value of real national income. In other words, the LM schedule shifts outwards. POLICY ANALYSIS IN THE IS-LM FRAMEWORK In this section we analyse economic policy. However, the kinds of policy issues that can be addressed within the IS-LM framework are rather limited: it is difficult to analyse price level changes. Within this framework we can say that expansionary fiscal policy (any increase in exogenous expenditure which, as we know, shifts outwards the IS schedule) will most likely increase output. So will expansionary monetary policy (an increase in the money supply in real terms). Totally differentiating equations (3.4) and (3.5) and writing in matrix form gives (3.6)

Setting d(M/P) = 0 and solving by Cramer’s rule we get

34


Figure 3.5

(3.6a)

(3.6b)

(3.6c)

These are the fiscal policy multipliers. An increase in exogenous exports or government expenditure or a reduction in taxes shifts the IS curve outwards without disturbing the LM schedule. It can be checked from equation (3.6) that in all these three cases r also rises (Figure 3.5). Clearly the magnitude of these multipliers depends upon the sizes of the coefficients Ly, Ir and the like. In particular we realise that as (the absolute value of) Ly rises these multipliers drop. In the extreme case the LM is nearly vertical, Lr is very large and the multipliers are nearly zero. This is the so-called ‘classical case’. In another extreme case Lr is very small and the LM schedule is nearly horizontal. In this case the multiplier is large (the so-called liquidity trap case). Classroom expositions of IS-LM analysis often depict a schedule as in Figure 3.6. This shows an LM schedule with three distinct ranges. At low rate of interest everyone expects the interest rate to rise so that the LM schedule is horizontal


35

Figure 3.6

(the liquidity trap case). At moderate rates of interest the LM schedule is upward sloping; it becomes vertical at very high rates. Any increase in exogenous expenditure simply raises the rate of interest and ‘crowds out’ private investment with no effect on equilibrium output. Some economists have often used this last situation to argue against expansionary fiscal policy. If the LM curve is neither completely inelastic nor perfectly elastic, then the effect of an increase in the supply of money will depend on the slope of the IS schedule. To analyse the effects of expansionary monetary policy we go back to equation (3.6). Setting and solving by Cramer’s rule gives (3.6d)

It can similarly be checked that We depict the situation in Figure 3.7. An increase in the money supply (leading to a shift in the LM schedule from LM to LM’ leads to a drop in the equilibrium interest rate and a higher equilibrium national income. Clearly, for any given shift of the LM schedule, the flatter the IS schedule the greater will be the increase in equilibrium y. AGGREGATE DEMAND WITH VARIABLE PRICES To study the dynamics of inflation we generalise the above IS-LM model and admit continuing inflation.

36


Figure 3.7

The IS schedule Define V as the natural log (henceforth written as ln) of saving plus taxes and J as the ln of investment expenditure plus exogenous government expenditure. Let Y be the ln of real national income, i the nominal rate of interest and πe the expected rate of inflation, so that denotes the real rate of interest. We can write (3.7) (3.8)

Savings are assumed to depend on real national income alone whereas investment depends on real national income and the real rate of interest. ß1 and ß3 are, respectively, the income elasticities of saving and investment. ß4 is an interest semi-elasticity. (Remember that is in terms of levels not logs.) All parameters are positive.


37

What is the difference between the real and nominal interest rate? Suppose you lend out $100 at an interest rate of 5 per cent for one year. You expect to get back If there is no inflation your real rate of return would be 5 per cent or, in this case, $5. Suppose, however, that the price level was expected to rise at 6 per cent? How much should you get back in order for you to get a real rate of return of 5 per cent? The answer is In other words, if r is the real rate of return and πe is the expected rate of inflation you should get back for every dollar you lend out. Now Now rπe is a very small magnitude and can be ignored. Hence the gross of inflation rate of return or the nominal rate of interest, i, can be written as (3.9)

This is the famous Fisher equation after Fisher (1930). Sometimes a corollary (the so-called Fisher effect) is implied that so that r is impervious to monetary disturbances. This, however, is not always true. See, for example, Jha, Sahu and Meyer (1990). To derive the IS schedule we set leakages from national expenditure equal to injections into national expenditure. In other words, we set savings plus taxes equal to investment plus government expenditure. This gives us a relation between Y and i, which can be written as (3.10)

where and Since the coefficient of i is the slope of the IS equation in the (i, Y) plane is However, the IS will be negatively sloped only if This requires that the response of savings to income be larger than that of investment to income. We will assume throughout that this assumption is justified. The intercept of IS on the i axis is whereas its intercept on the Y axis is The IS curve is shown in Figure 3.8. Disequilibrium zones are also labelled in the diagram. An increase in expected inflation will leave the slope of the IS schedule intact and shift it outwards. α0 is a summary measure of the exogenous components of aggregate demand. A change in α0, therefore, may be taken as a change in fiscal policy. Thus an increase in government expenditure will shift outwards the IS schedule without changing its slope. The LM curve Whereas the IS curve deals with the market for goods and services as flows per period of time, the LM curve deals with asset markets as stocks at a point in time. Let us write the demand for real cash balances as (3.11)

38


Figure 3.8

where m is the natural log of the nominal supply of money, p is the log of the price level and α2 and α3 are positive parameters. Notice that the opportunity cost of holding real cash balances depends on the nominal (not real) rate of interest. Now let µ be the rate of growth of nominal money and π the rate of inflation. Thus real cash balances grow at the rate and we can write the equation for the LM schedule as (3.12)

is real money demand last period so that the right-hand side of equation (3.12) denotes the growth of real cash balances from the demand side, whereas the left-hand side gives the equivalent expression from the supply side. Upon substituting from (3.11) for m–p we have the final form of the LM equation: (3.13)

Let us examine the money market in steady state where real cash balances are constant. The intercept of this schedule on the i axis occurs at and on the Y axis at its slope is This curve is shown in Figure 3.9 where the disequilibrium zones are also labelled. In Figure 3.10 we plot both the IS and LM schedules. If we make the Keynesian assumption that the interest rate responds primarily to pressures from the money market and national income responds to pressure from the goods market we can label i, Y movements in the disequilibrium zones as shown.


Figure 3.9

Figure 3.10

39

40


Figure 3.11

With expansionary fiscal policy and unchanged inflationary expectations, the IS schedule in Figure 3.11 would shift outwards. Both i and Y would rise. With expansionary monetary policy LM would shift outwards leading to a drop in i and a rise in Y. This is shown in Figure 3.12; however, both these policy experiments can be misleading if we keep πe constant in the face of changes in the actual inflation rate. Therefore, Figures 3.11 and 3.12 should be used only to show how macro policies are incorporated into the model and not to show their final effect on the economy. THE AGGREGATE DEMAND SCHEDULE We can now use the IS-LM schedules to drive an aggregate demand (AD) schedule. To drive the AD schedule we rewrite the IS curve as

and then substitute for i from the LM equation to get, after rearrangement, (3.14)

where Yd stands for output demanded as distinguished from Ys (output supplied) which we shall introduce later. We have drawn the AD schedule in


41

Figure 3.12

Figure 3.13

Figure 3.13. Its intercept on the π axis is and the slope of the AD curve is which is negative— hence the AD schedule defines a negative relation between π and Yd. An increase in the rate of inflation reduces the real quantity of money in the

42


economy creating excess demand in the money market. In turn, this causes the nominal interest rate to rise. With πe held constant the real interest rate also rises which, in turn, reduces investment and aggregate demand. Two other facets of the AD schedule should be pointed out. First, a change in fiscal policy will shift the AD schedule without changing its slope. Second, so long as the rate of growth of money and the rate of inflation are not equal to each other, will keep changing and so will the intercept of the AD schedule. An increase in πe will shift the AD schedule by α3 times the change in πe. At any given level of income this would also increase π by α3. In that way expected inflation pursues actual inflation. However, there must be some restraint on this self-fulfilling prophecy otherwise π and πe would chase each other without reference to µ. Students can check that the requirement constraint is THE AGGREGATE SUPPLY SCHEDULE The AD curve, by itself, cannot determine π and Y; to complete the system we introduce an aggregate supply (AS) schedule. The derivation of the AS schedule, which shows combinations of π and Y consistent with production conditions in the labour market, has two key elements: (i) an aggregate production function relates total output produced to labour input, and (ii) labour market supply and demand conditions relating labour supply and demand to the rate of inflation. The production function Without loss of generality we shall take a specific form of the production function: (3.15)

where Ys is the ln of output supply, n is the ln of total labour input and k is the In of total capital input. ß5, ß6 and ß7 are defined as positive parameters. In term of levels, then, we have the Cobb-Douglas production function. In the short run we shall take the amount of capital to be fixed and given exogenously so that we can write the production function as (3.16)

where This relation between Ys and n is shown as the straight line in Figure 3.14. Its intercept is ß8 and slope is ß6. An increase in the capital stock will shift the production function ß7 times the increase in capital. The labour input From our discussion in Chapter 2 we recall that labour demanded by firms depends negatively on real wage rate. Let us write this as


43

Figure 3.14

(3.17)

where w–p is the ln of the real wage rate and ß9 and ß10 are defined as positive parameters. Since wages and prices are rising over time, we need to write the labour demand schedule in terms of rates of changes rather than the absolute levels of wages and prices. Writing equation (3.17) for the previous period and the subtracting from (3.17) we have or (3.18)

where n-1 is labour demand in the previous period and is the rate of change of nominal wages. From this perspective, the number of workers rises only if since, in that case, real wages would be falling over time. Only if is the real wage constant and The general AS schedule By taking first differences of the production function (3.16) we obtain (3.19)

44


Figure 3.15

Into this we substitute the expression for

from (3.18) to get (3.20)

According to this equation, an increase in the inflation rate, with ω given, would start to reduce real wages and increase the demand for labour with the extent of this adjustment given by the parameter ß10. The extra labour will produce more output, as measured by the parameter ß6, allowing Ys to exceed Hence there is a positive relation between Ys and π. The AS curve in the π, Y plane is drawn in Figure 3.15. Its intercept on the Ys axis is and its slope is When the labour market is in equilibrium we must have and so that output is independently of π or ω and the aggregate supply curve would be vertical. Nominal wage determination The last step in the process of determining the AS curve involves the determination of the rate of change of nominal wages (ω). Again a special characteristic of the labour market is involved. A labour market is not a ‘spot’ market where purchases and sales are made for current use. Labour supply and demand usually requires some long-term commitment. In an inflationary setting the rate of change in the nominal wage will be set so as to maintain equilibrium in the labour market since firms want to be on their demand curve for labour, and households want to be on their supply curve. This requires a constant real wage. Given that commitments in the labour market must be made for some time in the future, and given that the actual


45

rate of inflation to prevail is not observed today, it seems reasonable that the expected inflation rate will have to be used instead. This requirement can be written as (3.21)

Substituting from (3.21) we have the final form of the AS curve:

THE COMPLETE SYSTEM The complete macro model can be written as

This complete system is depicted in Figure 3.16. In Figure 3.16(a) we have the IS-LM framework which is similar to Figure 3.8. Figure 3.16(b) combines the AS and AD schedules of Figures 3.15 and 3.13. Because of the connection between IS-LM curves and the AD curve the equilibrium level of the income part (a) cannot be different from that in part (b). The intersection of IS (πe) and LM determine Y and i and from the IS we can read the real interest rate. Then from the intersection of AD and AS we can derive the equilibrium values of π and Y. Long-run equilibrium If left to itself, the economy would settle down to a particular combination of i, Y and π, and would replicate itself year after year. Keynesian and New Classical economists would disagree about the time it would take to reach this equilibrium. Keynesians argue that it would take a long time for the economy to reach its long-run equilibrium so that it would become necessary to intervene in the operation of the economy with suitably designed fiscal and monetary policies. New Classical economists argue that the economy is quite resilient and would settle to its long-run equilibrium quickly. Both sides agree on the notion of equilibrium but the term ‘equilibrium’ means different things to the two camps. Let us, therefore, define two notions of equilibrium: short-run to temporary equilibrium exists when (3.22)

while long-run equilibrium requires, in addition, that (3.23)

46


Figure 3.16


47

We can see that the goods market is in a position of rest under both concepts of equilibrium. If output demanded was different from output supplied there would be unwanted movements in inventories and further changes in output and employment would occur. This process is assumed to be completed every time we observe the goods market in short-run equilibrium. Equation (3.22) also requires that asset markets be in equilibrium at all times since the position of the AD curve would not remain constant if µ or were changing over time. However, it is clear that the labour market may not be in equilibrium in the short run although it must in the long run. Only if equation (3.23) is satisfied is the real wage constant at the level required to set supply equal to demand in the labour market. When expectations of inflation are fulfilled, there is no incentive to make further adjustments and the economy can settle down to a stationary state. If this will not be known until the end of the current period during which no adjustment takes place, but after which a new estimate of πe will be made. In long-run equilibrium and the macro model of this chapter becomes even simpler. The AS schedule becomes vertical and is shown in Figure 3.17. Output is determined independently of demand. Second, given Y, the IS schedule can be written as So far as the LM schedule is concerned, we must have From this we get the real money supply and finally the real rate of interest is determined as The solutions for Y, π and i are determined sequentially, not simultaneously. Output is determined by the AS schedule, the inflation rate is equal to the given rate of growth of money supply and the interest is determined by the location of the IS curve. In Figure 3.17(b) the AD curve plays a limited role; it intersects with the AS curve at a vertical distance of In summary, then, the long-run values of Y, π and i are given by (3.24) (3.25)

(3.26)

It should also be understood that AD plays no significant role in determining Y. In other words, demand management policies, be they fiscal or monetary policies, play no role in determining Y in the long run. Another point that needs to be stressed is that, whereas in the simple ISLM model expansionary monetary policy would lower the nominal rate of interest and increase output; this would not happen in long-run equilibrium. An increase in µ from µ0 to µ1 will shift outwards the LM schedule in Figure 3.18, but we know that in long-run equilibrium the rate of inflation must

48


Figure 3.17


49

Figure 3.18

equal the higher money supply growth rate, µ1. Hence, the IS curve must shift to intersect the retreating LM at Ye. i is higher by the amount of the increase in the rate of growth of the money supply. However, since IS has not changed, the real interest rate will not change and the Fisher equation will apply. The real money supply rose initially when but as the LM started retreating the rate of inflation exceeded the rate of growth of the money supply. Hence, although the central banks started out with the larger growth of money, the economy ended with a lower value of real cash balances. However, this is the only real variable that has been affected by the increase in the growth rate of money supply. The long-run effect of a supply shock Suppose that at every wage rate the supply of labour falls—perhaps because of an increase in the preference for leisure or because of an increase in nonwage income. In Figure 3.19 this shifts the AS schedule to AS′ and real output in the long run would drop to If the AD curve does not shift there would be an increase in the rate of inflation. However, since µ has not changed, π must remain as before and AD will shift to the left to intersect the AS′ schedule at E″. The LM curve in Figure 3.19(a) is also affected by this exogenous event. Since income is lower, transactions demand for money must fall. This means that for some time in order to reduce the value of money. In final equilibrium at E′ both i and will be higher. The

50


Figure 3.19


51

higher real rate of interest would choke off some aggregate demand in the face of lower equilibrium output. THE DESIGN OF ECONOMIC POLICY IN THE GENERALISED MODEL What role does economic policy have in the generalised model? To conduct a meaningful analysis we must introduce uncertainty into the model, since we have to model expectations of inflation. We write the IS-LM model with uncertainty as (3.27) (3.28)

Hence, the error terms are merely added on to the IS-LM equations. The random terms εg and εm are assumed to be white noise processes—uncorrelated random variables with mean zero. On the supply side we introduce productivity shocks into the labour demand equation as (3.29)

where εs is the shock to the real wage. This has the same effect on the demand for labour as an equal proportional decline in the real wage. In the period prior to the shock (3.30)

By taking the difference between the two periods, we get (3.31)

where as

and Let us write the aggregate supply schedule and substitute for n from equation (3.31) to get (3.32)

Workers set wages so that supply schedule is written as:

We write α4 for ß6ß10. Hence the aggregate (3.33)

In Figure 3.20 we depict the effects of changes in the values of these stochastic variables. In Figure 3.20(a) an increase in the value of εg shifts the IS curve outwards whereas a decrease will shift it inwards. In Figure 3.20(b) an increase in εm will shift the LM schedule outwards and a decrease will shift it inwards. Any stochastic disturbance that shifts the IS or LM outwards will also shift the AD schedule outwards (as in Figure 3.20(c)), and any disturbance that

Figure 3.20


53

shifts the IS or LM schedules inwards will shift the AD schedule inwards. In Figure 3.20(d) we depict the effects of supply-side random shocks. With the long-run supply schedule shifts because of a change in the value of εs. In the short run, a change in πs will change the short-run supply schedule. The stochastic version of the AD schedule is derived by substituting the IS curve of equation (3.27) into the LM curve of equation (3.28). This AD schedule is written as

(3.34)

The values εg, εm and εs are not directly observable even after the fact because the data do not distinguish between anticipated and unexpected events. For instance, if the growth rate of money supply is measured at 10 per cent for a current year it is unlikely that the central bank would be able to indicate that 8 per cent was expected and 2 per cent was unintended. However, from the behaviour of the three endogenous variables π, i and Y we can extract information about the behaviour of the random variables. First let us refer to Figure 3.21. An increase in εg will shift the IS curve outwards in part (a) and the AD curve outwards in part (b). Thus i, Y and π will all rise. Let us now refer to Figure 3.22. An increase in εm will shift the LM curve outwards in part (a) and the AD schedule outwards in part (b). Thus Y and π will rise and i will fall. Let us finally take a look at the effects of an exogenous supply shock. In Figure 3.23 an increase in εs will shift the AS curve outwards, leading to a drop in π. The lower π will, in turn, mean a faster growth of real cash balances (given µ). Hence, the LM schedule will shift outwards somewhat and so will AD. Hence Y will rise and π and i will fall. We can summarise this in the following matrix:

Remember, though, that these are all short-run effects. Thus if we observe Y, i and π going up even though policy is steady we can infer that there has been a positive goods market shop (a sudden increase in consumer demand, for example). If we observe Y and π going up and i falling we can infer that there has been a positive money market shock—a sudden influx of cash as people want to buy more bonds and hold less cash, for example—even though policy is steady. If we observe that Y has gone up and π and i have fallen we can infer that there has been a positive supply shock— a sudden increase in labour productivity, for example.

54


Figure 3.21

Forming expectations about inflation In the foregoing analysis an important distinction was made between shortand long-run supply schedules. The difference between the two was ascribed to differences between the actual and expected rates of inflation. In the short run actual inflation may deviate from expected inflation whereas in the long run this would not occur. It therefore follows that the process of expectations


55

Figure 3.22

formation is an extremely important ingredient in the macro model we have been studying in this chapter. How are expectations of inflation formed? Since expected magnitudes are not observed we have to theorize about them without being able to observe them directly. Individuals are expected to use all the information that they have about the economy in the optimal manner. This is what rationality would require. Forming expectations through such a process is called rational expectations.

56


Figure 3.23

The question that then arises is what information do individuals have and what does one mean by optimal use of information? The standard answer runs as follows. First, economic agents know the IS-LM-AS model that we have been working with. They know, for instance, that a decrease in consumption expenditure will cause AD to fall and hence π will fall. As π falls so will πe in the long run so that there will be a drop in the real interest rate and IS and, therefore, AD will shift outwards. Hence, the drop


57

in the inflation will only be temporary. Second, individuals have certain information about the current performance of the economy and the policies in force at the current time, which they would use in forecasting inflation. To formalise this process we first solve the IS-LM-AS model to obtain the determinants of the rate of inflation. This involves finding the solutions for Y, i and π in terms of all the exogenous variables in the system from the AS and AD equations (3.33) and (3.34). The solution for π is

(3.35)

Each individual knows what causes inflation and to what extent. For instance, an increase in the consumption expenditure is caused by an increase in α0; multiplying by the known value of permits us to calculate the effect on π. Similarly an exogenous increase in πe increases π by To ensure that π does not rise more than πe we can impose a limit on the parameter a3; otherwise, we would end up with an explosive cycle of inflation merely through the expectations process without an accompanying increase in µ. Even more importantly we should remember that πe should not be treated as an exogenous variable. If α0 changes and π responds, πe cannot remain at its previous level. If it did we would not be using all the information we have at hand. To get out of the vicious cycle of π determining πe and vice versa we need to eliminate πe from equation (3.35). However, without πe in equation (3.35) the parameters attached to the other variables in the equation would no longer be the same since the underlying model now treats πe as an endogenous rather than as an exogenous variable. To get around this difficulty we ‘invent’ an equation similar to equation (3.35) but without πe in it, which means that π and other exogenous variables are connected by as yet undetermined coefficients (see Blanchard and Fischer (1991)). This equation is written as (3.36)

where the γ s are undetermined—undetermined because in (3.36) we have assumed that π e is endogenous. Next, π e is the mathematical expectation of π from this equation. Therefore, (3.37)

The terms γ2em, γ4εg and γ6εs have been eliminated since the expected value of each of these is zero. We can now substitute this expression of π e into equation (3.35) and thus eliminate π e from the list of exogenous variables. This produces

58


(3.38)

By a careful comparison it can be seen that equation (3.38) has exactly the same variables on the right-hand side as equation (3.36), the one that each individual invents when forming expectations about πe. Since they come from the very same IS-LM-AS model, the coefficients attached to each variable must be the same. With this information we can now solve for the undetermined coefficients, the γs in equation (3.36), by setting each of them equal to the corresponding coefficient in equation (3.38) and solving for the γs. Thus we have

which simplifies to assumed earlier that

which again simplifies to and

This parameter is positive since it was Next we have

Similarly

The undetermined coefficients are now determined and it is possible to get rid of the γs in equation (3.36) which now represents a structural relationship for the inflation rate, including the rate of expected inflation, but without having πe as a variable on the right-hand side. By substituting for the γs in equation (3.36) we arrive at


59

(3.39)

Taking expectations of inflation in equation (3.39) we have

(3.40)

Equation (3.40) differs from (3.39) only to the extent that This completes our examination of the process by which expectations about inflation are formed by rational individuals who use all available information. Policy ineffectiveness With the present version of the model it can be shown that policy is unable either to maintain equilibrium in the economy in the face of an exogenous shock or to reduce unemployment below the natural rate even temporarily. From the AS schedule of equation (3.33) we know that there are two sources of change Y: (i) a deviation of π from πe and (ii) a non-zero εs. To see what policy can and cannot do we need to focus just on these two sources. Substituting from (3.39) and (3.40) into equation (3.33) we get

(3.41)

According to equation (3.41), Y deviates from Ye only when εm, εg or εs take on positive or negative values. Here εs appears twice, once with a minus sign and once with a positive sign. On the one hand, εs>0 improves the marginal product of labour directly and therefore increases Y, but on the other hand it also lowers π below πe and, hence, reduces Y. The net effect, however, is positive: Y increases by Let us now examine the difference between π and πe by examining (3.39) and (3.40). We observe that (i) a change in a predictable or deterministic exogenous variable such as µ, α0 or Ye has exactly the same effect on π and πe; and (ii) random events such as εm, εg or εs affect π but not πe. The difference

60


between π and πe is the error that people make in forming expectations. Our analysis suggests that there is nothing systematic about these differences. These are entirely random. In particular, policy authorities do not have any systematic advantage over rational individuals in predicting π. Short-run effects of policy changes Let us consider the case where the economy is operating at equilibrium and the government attempts to reduce the rate of unemployment below the natural rate (where actual and expected inflation are equal). To do this it increases the rate of growth of money supply µ. We analyse the consequences of this in Figure 3.24. The LM schedule shifts to LM′ as It would thus appear that AD would shift up by the increase in µ to AD′ and, given the positive slope of the AS schedule, it would seem that output would rise which, in turn, would reduce unemployment. However, since everyone in the economy is aware of the policy and its effects and knows how the inflation rate will change because of this move, πe will be revised upward. This will cause AD to shift still further. To measure the combined effects of the change in µ and πe we rewrite the AD equation with πe substituted from (3.40) into (3.34). We then get

(3.42)

Thus the AD schedule shifts up to AD″ by the amount which is larger than the change in µ itself. The AS schedule is also affected by changes in πe. Substituting from equation (3.40) into the AS equation we have

(3.43)

Hence the AS schedule shifts up by the same amount so that the intersection of the AS and AD schedules will occur at the same Ye as before. The other important variables can be deduced from this model. Obviously π and i are higher. From the IS equation of (3.27) we can rewrite the real rate of interest as Since monetary policy cannot alter Y, it cannot alter the real rate of interest. However, the IS curve does shift up by the amount of the increase in πe. Finally, the same Y and higher i means that there will, in the final equilibrium,


Figure 3.24

61

62


be a lower demand for money. To make m–p smaller, π must be greater than µ at least for a while, which in turn, implies that LM must backtrack above its original position. What about predictable fiscal policy? It is clear that higher α0 will imply higher real interest rates. However, from the AS and AD equations an increase in α0 will lead to the same shifts in AS and AD. Hence, Y will remain unchanged. The higher real intersect rate will mean that investment will be lower. Higher government expenditure will ‘crowd out’ private investment. So traditional monetary and fiscal policies are not very useful. We are now ready for the second question. Suppose the workers make mistakes in predicting π. In these circumstances can the government through its policy instruments help to re-establish equilibrium? To answer this question suppose that we are in equilibrium and a shock occurs, say This shifts the LM curve downward. The AD curve will also shift outward by εm. Neither the AS nor the IS schedules are going to be disturbed because of this. The rate of inflation rises and the nominal interest rate falls. The economy is in shortrun equilibrium but not in long-run equilibrium since π and πe are different. Can the government use macroeconomic policy to avert this situation? The answer is ‘no’ since the authorities cannot anticipate εm any better than the people. If the government altered µ to counteract the effect of εm, the result would be no different from the case in which the authorities do nothing. Anti-inflation policy While the New Classical model renders active stabilisation policy ineffective it also enhances the effectiveness of monetary policy as an anti-inflationary measure. We understand from the above analysis that the economy can be at equilibrium at any rate of inflation. It is generally agreed that inflation has several costs attached to it (see Chapter 13). To reduce the rate of inflation permanently, the central bank must reduce the rate of growth of the money supply. Assume that such a policy is announced and implemented. If such a policy is believed it will reduce π and πe by times the change in µ according to equations (3.39) and (3.40). Therefore, in Figure 3.25 the economy will stay on the vertical line at Ye. Reducing the rate of growth of the money supply from µ0 to µ2 lowers the position of the AD schedule to AD′ by As was the case before, the inflation rate will overshoot its final target and ultimately the AD schedule will have to shift to AD″ where the vertical distance between E″ and E is µ0–µ2. Not only is the credibility of monetary contraction important for this result, but in equilibrium all market participants must be aware of overshooting. Consider a case when πe did not adjust at all to the announcement of a lower µ. If workers did not believe this and thought that the central bank announcement was a way to get them to accept lower real wages they may continue to bargain for wages based on the current value of


63

Figure 3.25

π. In that case the AS schedule would not shift whereas the AD would and equilibrium would be established at E″. In general the smaller the change in pe relative to the shift in the AD curve in the wake of a policy change, the greater is the departure of income and employment from its equilibrium level. In other words, in the ultimate analysis, not only does monetary policy influence π, it also influences πe. The greater the credibility of the central bank of a country the greater would be its ability to reduce inflation. THE NEW KEYNESIAN MODEL In the New Classical model we have studied so far the guiding principle is equilibrium. Underlying this belief is the perception that all markets are like auction markets, clearing almost always and almost instantaneously. Nevertheless, we observe markets that are emphatically not auction markets. In particular the labour market is a market characterised by contracting behaviour. There are other features of the New Classical model that have been questioned by those economists who believe that stabilisation policy remains useful and effective. In their view the economy is not as self-regulating as the New Classical economists would have us believe. One area of disagreement between the New Classical and New Keynesian schools is the process of expectations information. Sometimes the New Keynesian school

64


argues that expectations are of the adaptive variety (discussed in the Appendix to this chapter). In terms of the current analysis and are known today and we can use this information to calculate π-1. This variable summarises all the information about the economy that is currently available and is used as a predictor of π in the current period. Thus the following expectations process is applicable:

(3.44)

The parameters attached to the ε-1 variables are determined in the manner of equation (3.35) which shows how es change π. For example, if πe should be lower than π-1, because the inflationary shock is not expected to last. Adaptive expectations leads to systematic errors in πe because πe will always be below π when inflation is rising and it will be above π when inflation is falling; only when π is steady will πe equal π. Although individuals try to eliminate such errors they cannot prove their performance without knowing what has yet to be measured. An economy operating with adaptive expectation as expressed by equation (3.44) can move away, although only temporarily, from its equilibrium level

Figure 3.26


65

of output through deliberate but unknown policy changes. Starting from Ye assume there is an increase in µ. Since πe is geared to π0 the economy must move along the short-run AS curve. The AD curve shifts up by the increase in µ and therefore the new equilibrium is established at E′. But this cannot be the final position for the economy. Since inflation has risen to π1, AS must shift up by to AS′. Further, since the increase in πe shifts the AD schedule up by α3 and, since in the first period, has increased and this puts further upward pressure on the vertical intercept of the AD schedule. Suppose AD″ is the new position so that a new temporary equilibrium has been established at E″. Here Y may be higher or lower than at E′. Subsequently the AS and AD schedules keep shifting until final equilibrium is established at E′″ with income again equal to the natural rate and In the meantime, the policy change has been able to influence income and unemployment rate by lowering the real wage when πe responds to the change with a lag. Government’s ability to exploit such a situation may lead them to conceal information. Productivity and wage growth Labour productivity keeps changing almost continuously. Let us see what this implies for the labour market in our analysis. Our labour demand equation was written as where supply shocks have been incorporated because increases in εs increase the marginal product of labour. Our non-stochastic labour supply schedule was We equate the two and calculate an equilibrium value of ω as: (3.45)

From this equation we realise two important points. First, if the labour market was not in equilibrium last period (in particular if there was an excess demand for labour), nominal wages must, today, rise at a rate faster than the current rate of inflation. Further, since the value of εs is not known today it is not possible to keep the labour market in full equilibrium in the face of supply shocks unless we find out some method of rewarding workers today for productivity gains that will occur only ex post. Productivity of labour keeps changing almost continuously because of accumulation of human and physical capital. Let us suppose that such exogenous changes in the productivity of labour are captured by the term εs where κs represents trend growth in labour productivity and ζ represents unpredictable movements around that trend. The trend is predictable and should be included in the wage increase whereas the random disturbances are not and cannot be incorporated. We may calculate the trend by taking the average value of εs for the past n years and then ζ is the residual.

66


With such growth in labour productivity the equilibrium increase in the nominal wage is A higher value of ζs leads to higher real wages, but not to the same extent because However, one can never be sure whether a particular observation for εs is the start of a new trend or a particularly large random shock. Consider a situation where ζs has fallen to a new lower level: from ζs0 to ζs1 because of smaller growth productivity. If this new trend is not detected, nominal wage growth will be too high for equilibrium to be maintained in the labour market. We can write the difference between the actual and the equilibrium rate of growth of nominal wages as

Since it must be the case that This causes the AS curve to shift up and will lead to a higher π and lower Y. In the next period the fact that ω >π puts pressure on ω to fall, but so long as the new trend in productivity growth is not incorporated in wage equations, the actual rate of wage increase will continue to be greater than the equilibrium rate and the AS schedule shifts again. Problems such as this imply that the labour market continues to be in disequilibrium for a considerable time. Long-term contracts An important difference between New Classical and New Keynesian economists is that the latter believe that nominal wages are not set in an auction market but are set contractually. Suppose a contract lasts for two time periods. In year 1 half the workers and half the firms sign a two-year contract. They examine equation (3.45) for this purpose. Suppose expected inflation, as these firms and workers calculate it, is too high. Real wages will therefore rise by and excess supply appears in the labour market. In year 2 the other half of the work force begins its contract negotiation. However, it is stuck with a disequilibrium from period 1 and, therefore, must accept lower wage increases if the labour market is to clear. To be precise nominal wage increases must be lower by the amount This is a situation of conflicts between the two groups of workers. In this situation it will be possible to reduce or even remove the real wage gains of the first group of workers by increasing µ to increase π. This would increase labour demand and remove the necessity for adjustment on the part of the second group of workers. The argument for intervention by the policy authorities becomes even more convincing if the second group of workers, in the interest of wage equity, asked for still higher real wages. This would further exacerbate the problem of unemployment. In the case of long period wage contracts policy intervention speeds up considerably the adjustment to equilibrium in the labour market.


67

INDEXATION OF NOMINAL WAGES The foregoing analysis points to the possibility the worker would like to protect wage gains in real terms. Workers may want to achieve a real wage ex post rather than ex ante. This is possible through the cost of living adjustment (COLA) of nominal wages. This aspect of the labour market can have different effects depending upon the type of shock that occurs. Our analysis here follows that of Gray (1976). We have been assuming that now nominal wage increases are forward looking in that they are set equal to the expected rate of inflation. With COLA nominal wage setting becomes backward looking. Nominal wage increases get tied to whatever π was last period. Let us suppose that nominal wage increases are partly geared to actual inflation last period and partly to expected inflation, i.e. (3.46)

where is the indexation factor. means that there is full indexation, means that there is no indexation and means that there is partial indexation. The AS schedule is affected by the form that indexation takes. If then the AS schedule remains as before. With indexation of the form given in equation (3.46) we get, upon substitution in the AS equation, (3.47)

The slope of the AS schedule is If the AS schedule is vertical and the distinction between π and πe is unimportant. We analyse the effects of AD shocks with and without indexation in Figure 3.27. As εm rises, for instance, AD floats upwards. With a positive shock and no indexation both output and inflation rise and we move to E′. With full indexation Y remains stuck to Ye and inflation climbs up to the higher level associated with the equilibrium E″. As we have seen, in a world that is subject to shocks, it is not possible to maintain labour market equilibrium. If Ye is a preferred level of output then full indexation ensures that output does not deviate too much from it. To that extent indexation in the labour market is desirable. However, when the economy is subject to supply shocks originating in the labour market the optimum may be only partial wage indexation. To see this let us analyse Figure 3.28. (In this diagram ω is the ln of the nominal wage and p the ln of the price level so that is the ln of the real wage). Initially Now allow εs to take a negative value. This shifts the ns curve down by the extent of εs. To move toward a new equilibrium the real wage should drop to but unless there is just the right degree of wage indexation this will not happen. With full indexation we remain stuck at real wage and there will be unemployment in the excess of the natural rate.

68


Figure 3.27

Figure 3.28


69

With no indexation π will rise by the amount Let us assume, for simplicity, that and then ω–p will fall exactly by the amount of εs which leads us to which is too low for the equilibrium. To reach the new equilibrium, therefore, we require partial indexation. CONCLUDING COMMENTS In this chapter we have analysed two variants of the standard Keynesianneoclassical synthesis model. The basic models studied included the fixed price Hicks-Keynes type IS-LM model as well as a model that admits of inflation. The essence of the new classical equilibrium in which standard stabilisation policy plays no role has been identified. At the same time, the conditions under which the new Keynesian position that stabilisation can affect real output have been identified. APPENDIX THE PHILLIPS CURVE In 1958 Phillips published an empirical study of British data about the relation between unemployment and the rate of change of nominal wages. He discovered that there is a negative relation between the rate of change of nominal wages and rate of unemployment. This is shown on Figure A.3.1. On the abscissa we plot the rate of unemployment, u, defined as the number

Figure A.3.1

70


of people unemployed, U, divided by the total size of the labour force, N. On the ordinate we plot the rate of inflation of nominal wage, i.e. The observation on µ and u for the various years (Phillips covered the period 1861–1957) are clustered fairly close to this Phillips curve. Phillips’ analysis was purely empirical. He did not provide a theory or explanation for this negative relation between u and µ. The link between u and the rate of price inflation is fairly easy to establish. Let η be the marginal product of labour. Remember that in competitive equilibrium firms would equate the marginal product of labour to the real wage rate:

i.e. we obtain the rate of price inflation by abstracting the rate of growth of productivity of labour from the rate of inflation. Once we do this we can define a Phillips curve between the rate of price inflation and the rate of unemployment. If we write the equation of the Phillips curve in Figure A.3.1 as

where u* is the rate of unemployment consistent with wage stability, we can write

We have said nothing so far about the theoretical underpinnings of the Phillips curve, this problem being addressed in an important contribution by Phelps (1969). Suppose that the labour market is characterised by pure, not perfect, competition. This means that there are many employers and many workers. However, the labour market is not characterised by perfect knowledge. Workers do not know the going market wage, nor do firms. Workers search for jobs and firms make wage offers to workers. Suppose workers expect wage inflation to be 5 per cent over the previous period. If a worker were offered a wage that is more than 5 per cent higher than the previous period’s wage it is likely that the job offered would be accepted. By this reasoning, the higher the rate of wage inflation, given the expected rate of wage inflation, the greater would be the number of people accepting jobs and the lower, therefore, would be the rate of unemployment. We show this in Figure A.3.2. Given expected inflation of 5 per cent, if the actual inflation turns out to be 6 per cent, unemployment would be 4 per cent of the labour force. If the rate of inflation was higher, say 10 per cent, unemployment would have been 3 per cent.


71

Figure A.3.2

Suppose that the current inflation rate is indeed 10 per cent. Next time period people realise that their expectations were wrong: they expected inflation at the rate of 5 per cent whereas the actual inflation rate turned out to be 10 per cent. Rational economic agents would revise their expectations. Suppose they revise their expectations of inflation by a fixed fraction of the gap between the actual inflation last period and the expected inflation in that period—the so-called adaptive expectation hypothesis. Suppose the fraction of adjustment is 0.5. Hence expected inflation this period would be When the expected rate of inflation is 7.5 per cent it will not sustain 3 per cent unemployment with 10 per cent inflation. With higher expected inflation fewer workers will accept jobs when the actual inflation rate is 10 per cent. In other words, the Phillips curve will shift upwards. Expected inflation will keep rising and the Phillips curve will keep shifting so long as actual inflation is higher than expected inflation. When unanticipated inflation is zero the Phillips curve will stop shifting. In Figure A.3.3 we begin at point A. As expected inflation rises, the Phillips curve starts shifting upwards. In the medium run we move along a path like the one indicated by the arrows. At rate of employment u* there is zero unanticipated inflation. A couple of points about this should be made clear. First, it is argued that no matter where the process of expectations adjustment starts we shall always end up at the unemployment rate u* in the long run. At point B expected wage inflation is 30 per cent whereas the actual rate of wage inflation is 20 per cent. Hence inflationary expectations have to be adjusted downwards. We trace a path from B to C as shown by the arrows— once again to a level of unemployment u*. It is hypothesised that u* is unique.

72


Figure A.3.3

It does not matter whether the long-run (fully anticipated) rate of inflation is 5 per cent, 10 per cent or 30 per cent. This unemployment rate has been called the natural rate of unemployment. In the medium term, unemployment and inflation can both rise as in the movement from A to D or fall as in the movement from B to C. This is the so-called accelerationist hypothesis. An important conclusion that emerges from the analysis is that in the long run the economy will supply the natural rate of output—one that is consistent with the natural rate of unemployment. In the short run output can deviate from the natural rate by a factor that depends upon magnitude of unanticipated inflation: (A3.1)

Actual output at time t (yt) is equal to the natural rate of output (y*) plus a factor that depends on the magnitude of unanticipated inflation. This equation is often referred to as the Lucas supply function. AN ALTERNATIVE INTERPRETATION OF THE NATURAL RATE OF UNEMPLOYMENT We have interpreted the natural rate of unemployment as a situation in which the actual and the expected rates of inflation are equal to each other. There are other possible interpretations. We outline one now. Consider the labour market depicted in Figure A.3.4. The demand for labour is a decreasing function of the real wage. The supply of labour is drawn as an upward sloping function of the real wage. Equilibrium in the labour market is denoted by J where the demand and supply of labour are the same.


73

Figure A.3.4

The labour demand consists of actual employment and vacancies. In every economy there are some vacancies that are unfilled coexisting with workers who are searching for work. We can then define a labour use schedule ( in Figure A.3.4) which shows actual employment. The higher the real wage it is responsible to expect, the smaller will be the number of unfilled vacancies. The lower the real wage the greater the number of workers looking for work. With labour market equilibrium at J in Figure A.3.4 actual employment is 0N1 whereas labour demand is 0N2. The distance N1N2, therefore represents unfilled vacancies. Since labour demand equals labour supply unfilled vacancies must equal quits to search for jobs. An alternative interpretation of equilibrium in the labour market is, thus, a situation where the number of vacancies equals the number of quits. If this is the case, there will be no pressure for the wage to change. Expectations of inflation will, therefore, also be fulfilled. The distance N1N2, therefore, represents the natural rate of unemployment. DISCUSSION QUESTIONS 1

Consider an economy that has just experienced an increase in its rate of inflation. The central bank of the country has announced that it intends pursue a restrictive monetary policy. Suppose that you were a neoclassical economist, would you consider this policy appropriate? If you were a new Keynesian economist what reservations would you have against this policy?

74


2

Milton Friedman is famous for saying that ‘inflation is always and everywhere’ a monetary phenomenon. Using the quantity theory of money relation: where M is nominal money stock, y is real income and P the price level, indicate what assumptions are needed to validate Friedman’s claim. Suppose you are a new Keynesian economist. Assume that financial innovation makes the value of the interest elasticity of money demand more uncertain. How would this affect your relative preference for the use of monetary and fiscal policy in stabilising aggregate demand? How might it affect your overall willingness to engage in stabilisation policy?

3

4

The process of money creation and the demand for money

INTRODUCTION In this book we have used the demand for money and its supply without inquiring into the foundations for such relationships. It is time to dwell on these issues somewhat. Students have, no doubt, been exposed to these foundations in earlier courses in macroeconomics. The discussion here therefore will be brief. We begin with a discussion of the importance of the role of the financial system in economic development. We then move on to a discussion of elementary models of money supply. The discussion here includes the effects of ‘financial expression’—a characteristic of many developing countries—on the money supply. We then proceed to discuss the underpinning of the demand for money. THE IMPORTANCE OF MONEY AND FINANCE IN ECONOMIC DEVELOPMENT Students often begin their study of macroeconomics by appreciating the advantages of monetary economy over a barter economy. Unless there is a ‘double coincidence of wants’—the buyer wanting to buy exactly what the seller wants to sell at a mutually acceptable ratio of exchange—a barter economy is likely to be extremely inefficient, we are told. No modern economy can be based on the barter system and expect to prosper. Most developing countries stated their post-colonial phase of development as near barter economies. However, the mere introduction of money does not make for a modern monetary economy. A fully developed financial system consists of a wide range of institutions, instruments and markets. Further, a well-developed financial system is not a static system—but one that keeps changing over time. How does the growth of the financial system affect the development of the economy? At one extreme it can be argued that if the financial system is completely underdeveloped there will be no transfers of funds from savers to investors.

76


Savers and investors are usually different people and the financial system plays a crucial role in mediating the transactions between them. Savers may not be willing to invest directly in the enterprises of firms and other investors but these savers may be willing to keep their funds in banks. These banks may, in turn, be willing to lend funds to investors. To perform well their role as mediators of the saving-investment relationship, banks and other intermediaries must be attractive both to savers as well as to investors. In other words, they must provide the incentive to save, increase the volume of investment and improve the efficiency of investment. If banks perform their role well savers will find that financial investment with these banks is preferable to increasing consumption or investing directly in enterprises. Potential investors now have access to savings other than their own. Thus the volume of investment will pick up. Moreover, the development of markets in money and financial instruments will tend to channel funds toward the more productive and profitable ventures. Apart from facilitating the development of a productive relationship between savers and investors, a well-functioning financial system performs other roles as well. Financial instruments and institutions perform the extremely important task of maturity transformation and risk transfer. Maturity transformation permits savers to save short term and investors to acquire long-term funds. Typically savers do not want their funds for long periods of time. They want to have access to fairly liquid funds should they want to change their financial portfolio or consume more. But the typical investor wants long-term funds. Financial institutions perform the essential function of maturity transformation between savers and investors. Risk transfer is another very important role of financial institutions. Many savers do not wish to undertake the risk of physical investments themselves. Moreover, they may also be reluctant to lend to investors who can take these risks. Apart from the purely psychological factor of not wanting to take risks, savers may, quite reasonably, feel that they do not have the technical expertise to evaluate the worth of the projects being contemplated by the investors. Such savers prefer to use financial intermediaries, in which they have confidence and which they perceive to be financially sound. These financial institutions then provide funds to many investors thereby diversifying and reducing average risk to themselves. Other important financial institutions such as stock and bond markets also perform very important roles in the development of the economy. However, stocks and bonds provide relatively quicker liquidity and, in any case, their significance is relatively limited in most developing countries. The relationship between financial development and economic growth has been the focus of considerable empirical research. Three competing empirical hypotheses have vied for attention. One suggests that financial development must precede economic development. The second suggests that financial and economic development occur together. The third, rather extreme,

Money creation and the demand for money

77

view is that financial development follows rather than leads economic development. This last view has not been supported by the facts and very few economists espouse it. For a statement in the case of Nigeria see Woolmer (1977). The large empirical literature with respect to the first two views has been surveyed by Gupta (1984). He also conducts causality tests along the lines of Granger (1980) between various indices of financial development and economic growth. His conclusion was that financial development leads to economic development. This was true of a sample of fourteen fairly representative developing countries that he studied. THE SUPPLY OF MONEY AND MODELS OF MONEY CREATION In this book and in scores of theoretical and empirical analyses it has been assumed that the money stock (M1, i.e. demand deposits plus currency) is determined by the monetary authorities. However, in most countries it is probably more appropriate to assume that the money stock is determined by a process that results from the complex interaction of the behaviour of various economic agents, rather than as one controlled exclusively by the central bank of the country. In this chapter, we shall try to understand the complexities of the money supply process. To do this we shall have to make assumptions about the behavioural responses of various economic agents involved in the money supply process and the institutional setting within which day function. Our approach will be to introduce a number of simplified models to understand the process of money creation. From this it will become clear why it is inherently difficult for the monetary authorities to control the money supply. Model of money creation I In this model we assume that the domestic money supply consists of currency holdings and demand deposits alone and, further, that banks wish to hold no excess reserves. We can define the monetary base (B) of the economy as total commercial bank reserves (R) plus currency in circulation (C): (4.1)

Under our assumption, commercial banks will be in equilibrium (in the sense of desiring no expansion or contraction of loans and deposits) only when all reserves they hold are required against their demand deposit liability. If RRD are reserves legally required against demand deposit liabilities, ? is the average legal reserve requirements on commercial bank demand deposit

78


liabilities and DD are demand deposit liabilities of commercial banks, then it must be the case that (4.2)

and that (4.3)

Substituting equations (4.2) and (4.3) into (4.1) we have (4.4)

Equation (4.4) makes clear an important characteristic of the model. The monetary base is determined by the monetary authority; however, its composition is determined by the actions of the non-bank public. From equation (4.4) we can see that (4.5)

Thus, changes in the stock of deposits are related to the changes in the base by equation (4.6): (4.6)

Now we can examine the impact of the change in the monetary base. Suppose that for every dollar held in a bank account people wish to hold a fraction s as currency so that

From equation (4.4) we can write

or (4.7)

We can similarly write the increase in cash holdings as equation (4.7),

or, from

(4.8)

Total increase in money supply equals increases in currency held and deposits: From equations (4.7) and (4.8) we have (4.9)


79

Equation (4.9) is the final expression we want. It relates total money supply increase to an increase in monetary base. It also determines the split up of this ∆M between ∆DD and ∆C. Model of money creation II In this model we continue to assume that commercial banks want no excess legal reserves. They will expand or contract their outstanding loans any time total reserves differ from required reserves. We now define the monetary base

where (4.10)

where RRD are reserves legally required against demand deposit liabilities and RRT are reserves legally required against time deposit liabilities. Thus and where TD is total time deposits and b is the average legal reserve requirement on commercial bank time deposit liabilities. Thus (4.11)

Hence

In terms of differences, (4.12)

We want again to get the total increase in money supply in response to an increase in the monetary base. Suppose the public maintains fixed proportion between time and demand deposits where the factor of proportionality is γ, i.e. Hence

so that (4.13)

This gives us the increase in demand deposits. We know that and We know that

(4.14)

80


Equation (4.14) then relates total change in money supply to a change in monetary base. Model of money creation III It is reasonable (and historically accurate) to assume that commercial banks may not act like the ‘automatic dispensers of credit’ pictured in models I and II. Suppose that banks wish to maintain a fixed proportional relation between their holdings of excess reserves (RE) and their deposit liabilities, i.e. where w is the factor of proportionality. Now as before,

Substituting appropriately we have whence (4.15)

so that (4.16)

This gives us the required relation between changes in the monetary base and changes in demand deposit. To get the effect of changes in B on total money supply we add ∆C to equation (4.16) to get (4.17)

Models with variable interest rates In reality the public’s asset preferences may be well described only by quite complicated relationships. The interested reader is referred to Cagan (1965) for a thorough analysis of the many complexities involved. In this section, we shall give students only a minor taste of some of the richness that can be built in. Time deposits are markedly different from demand deposits and currency because they pay fairly high rates of interest. Hence, giving up time deposits for demand deposits or cash involves a substantial opportunity cost. The higher the rate of interest on time deposits, ceteris paribus, the more costly it would be in terms of foregone income to hold money in the form of demand deposits may influence the size of time deposits. One may postulate, for example, that (4.18)


81

where iT is the interest payable on time deposits. We can then write

with have

and

Upon substitution we

or (4.19)

To take another example, let us suppose that the asset preferences of the banks and the public are dependent on some market interest rates. For example, the public may consider Treasury Bills an alternative to their holding currency or time deposits. Commercial banks’ behaviour may also be influenced by market rates of interest. If we let im be a summary measure of such interest rates, we may write (4.20a) (4.20b) (4.20c)

Equations (4.20a)–(4.20c) denote the fact that market rate of interest may affect holding of cash and time deposits by the public and excess reserves by banks. Now substituting from (4.20) and solving for DD, we have (4.21)

where and Equation (4.21) is a reduced form equation. In this form ϕim represents the total influence of im on the volume of demand deposit liabilities through its effect on the preferred asset holdings of the public and the banks. Since g, φ and ϕ are negative and b is positive, ß is also negative. From equation (4.21) then or, as the market interest rate increases, the equilibrium quantity demand deposits increases. This is shown in Figure 4.1. Our discussion above shows that there can be many ‘money multipliers’ and the supply of money can be influenced by several different interest rates.1 The form of money multiplier and the choice of interest rate variables that affect the money supply depend directly on the form of the behavioural relations employed to explain the asset preferences of the banks and the public.

82


Figure 4.1

These questions could easily be expanded to include additional variables such as income, wealth, expectations of inflation and so on. The ‘best’ money supply model will be the one that explains the actual data (actual past values of the money supply) and predicts future money supply better than any other model. Some special characteristics of money supply in developing countries Many developing countries do not have free capital markets. These markets are often characterised by ‘financial repression’. Financial repression takes many forms in developing countries—the most common being ceilings on interest rates or government-controlled interest rates—particularly interest rates on deposits or/and loans. During periods of inflation, then, real rates of interest (nominal interest rates minus the rate of inflation) are often negative. At the artificial low rate of interest there exists an excess demand for investment. But investment must equal saving so that both saving and investment are lower. Sometimes repression also takes the form of government restrictions on the free development of financial instruments and institutions. It is often the case that the reserve requirements of commercial banks, set by central banks in developing countries, are very high. In developed countries reserve requirements of 10–15 per cent prevail. In developing countries, however, reserve requirements of 50 per cent or more are quite common. These high requirements have two serious implications. First, a substantial portion of investible funds is directed away from potential investment. Second, the banks’


83

interest rate structure remains distorted because they must retain a wide margin between borrowing and lending rates in order to compensate themselves for the fact that they have to retain substantial funds in the form of reserves. Sometimes governments compound matters even further by regulating investments by banks. In many cases banks have to buy government bonds since there is not much of a private demand for these bonds. Over and above this, governments encourage investment by banks in certain ‘priority’ sectors. Bank investment then gets guided by the government’s preferences that may not be an accurate reflection of true economic productivity or profitability. Clearly, financial repression retards the development of the financial sector and, hence, of the economy. We examine these consequences of financial repression and liberalisation in Chapter 16. In this section we realise that financial repression may have very serious implications for the money supply in developing countries. Total money saving is roughly equal to bank and quasi-bank deposits and currency in circulation. This is the M2 definition of money supply. Since bank deposits and cash are the principal forms of savings in developing counties, these savings and M2 are approximately equal to each other. Hence, in equilibrium, savings and the demand for money are equal to each other. Suppose now that inflation rises. At controlled nominal interest rates a higher rate of inflation means a lower real rate of interest. People will tend to invest in inflation hedges such as gold, jewellery, real estate and the like. Productive investment will therefore decline. This will have the effect of lowering the rate of economic growth. It should also be noted that investment in the inflation hedges is itself inflationary since this tends to push up the prices of these hedges and the general price level has a tendency to rise with these prices. Moreover, expectations of higher real inflation will tend to have the same effect as higher inflation itself. THE DEMAND FOR MONEY We have studied the demand for money earlier in Chapter 3. At that level of simplification, we could distinguish between two different sources of the demand for money: the transactions motive and the speculative motive. The precautionary motive is not considered too significant. The transaction motive Economists believe that money is demanded, like any asset, because of the flow of services it yields. With most assets, the value of this flow of services is measurable in terms of dollars per period of time. With money, however, this flow is not measurable in dollar terms. Cheque accounts in banks pay low rates of interest and the value of these accounts gets depleted in times of inflation. The flow of benefits provided by money is implicit and recognisable

84


in the services it performs. Money is a unit of account, a medium of exchange and a store of value. The unit of account function of money simply means that the money is used as numéraire in which the prices of all other goods and services are measured. In other words, the price of money in terms of itself is one and the prices of all other goods are measured in terms of money— dollars per unit. The store of value function of money is discussed with the speculative demand for money. As a medium of exchange, money is held between the receipt and disbursement of income because of the cost of converting into and out of other, earning, assets which themselves are not generally acceptable as media of exchange. A monetary economy is more efficient than a barter economy. The amount of money demanded to finance expenditure, in principle, thus depends on these conversion costs as well as the size of the expenditures and how far in the future they will be made. Most economists, nevertheless, have emphasized the size of expenditure as the primary constraint affecting the demand for money as a medium of exchange. In fact, the transactions demanded for money, associated with its use as a medium of exchange, is usually depicted as varying in direct proportion with a level of expenditures. Following this tradition, the ith individual’s transactions demand for money function is (4.22)

with being the average quantity of real transaction balances held by the individual, where PT is the price level of all transactions, Mi is the nominal cash balances held by the ith individual, Ti is the level of the ith individual’s total real expenditures over the relevant time period, say a month, and αi the proportionality factor with a value for the relevant period. The order of magnitude of αi relates the size of the average money stock needed to finance a flow of expenditures to the size of that total flow. For instance, suppose an individual is paid a monthly income of $3,000 on the first day of each month. Assume further that she spends this income at a uniform rate of $100 per day for 30 days until she is paid again on the first day of the next month. The average transactions balance will be Therefore,

Since our major concern is the analysis of aggregate economic behaviour we must aggregate over all individuals in the economy as a whole, and show


85

what determines the value of α for the economy as a whole. We can define In an economy with N individuals

Now (4.23)

so that

The value of α for the entire economy is the summation of individual αi each weighted by the individual’s share of aggregate total expenditures for a given time period (in our case here, a month). It will be useful to reformulate the demand for transactions balances as a function of the level of output of newly produced goods and services (real national income) rather than the level of total expenditures. Total expenditures exceed the expenditures on new goods and services since total expenditures include expenditures of intermediate goods, second-hand goods, financial assets and the like. Define α0 as the ratio of real aggregate transaction balances to the level of aggregate output (4.24)

where P is the aggregate price level and y is the level of real national income. Hence

(4.25)

we can now write the aggregate demand for money function as (4.26)

Early theorists like Fisher and Keynes pointed out that income was not the only variable affecting the transactions demand for money. But it took some

86


time for economists to develop vigorously a theory of how conversion costs affect the transaction demand for money. William Baumol was the first economist to formally incorporate conversion cost in the transaction demand for money. Baumol proceeded from the assumption that the stock of cash is the holding inventory of the medium of exchange. The rational individual will try to minimise the cost of holding this inventory. Suppose transaction costs are perfectly foreseen and occur in a steady stream over the month. Real income is $H per month and it is assumed that the individual pays out all her $H per month at a constant rate. Thus during the month the individual will be holding an ever diminishing stock of assets. In what form will these assets be held? If assets are held in the form of money then some interest income is forgone. If assets are held as bonds then the individual will have to pay a conversion cost every time she tries to convert bonds for money for purposes of carrying out transactions. Suppose the individual begins the period holding all her income in bonds. She is assumed to withdraw the money from bonds in units of $C evenly spaced. For example, if the individual’s income is $3,000 per month then C might be $3,000 every 30 days or $1,500 every 15 days and so on. There will be H/C withdrawals made during the period. For each such withdrawal there is assumed to be real transaction cost of $b. The expression b(H/C) is a kind of ‘inventory replenishment cost’ which includes not only the explicit cost (e.g. brokerage fees) of selling assets to get cash but also the implicit cost (the inconvenience of doing so). We assume that the individual’s withdrawal of $C is expended at a constant rate. Hence the individual’s average cash balance is $C/2. Hence the interest opportunity cost of holding bonds is iC/2 where i is the interest payable on bonds. Total cost J of holding this inventory can now be written as (4.27)

the individual will choose C to minimise J. To do that set solve for C to get

and

(4.28)

since the optimal average balance is C/2 we can rewrite equation (4.28) as (4.29)

where


87

This result shows that (i) the transaction for money demand does not generally vary in proportion to the level of total expenditure, i.e. since there are economies of scale in the management of cash balances; (ii) the quantity of transaction balances is negatively related to the interest rate; and (iii) because of the transaction cost of exchanging cash for bonds it generally pays to the hold cash. This square root rule of Baumol is justifiably famous for several reasons. Two of the most important are: (i) it provides a justification for the negative relation between the interest rate and the demand for money independent of the speculative demand for holding cash, and (ii) it predicts that the demand for money varies less than in proportion to variation in income. Hence, a given increase in the money supply will, ceteris paribus, require a larger increase in income to increase money demand sufficiently to equilibrate the money market. This means that the monetary policy is a potent tool for stabilisation. Speculative demand for money balances Suppose we live in a world with two assets: cash M/P carries zero yield but, by definition, the distribution of returns has zero variance. The other asset, R, is a perpetuity with a mean return of µ and standard deviation of returns σ. We think of this standard deviation as a mean of the risk of holding R. The wealth holder has a utility function defined over risk and return with risk ‘normally’ being a bad and return a good. What do we expect the individual’s difference curves between risk and returns to look like? The possibilities are plotted in Figure 4.2. In Figures 4.2(a) and 4.2(b) we plot the indifference map of a risk averter. More risk will be undertaken only if the mean return goes up. In Figure 4.2(a) we plot the indifference map of a diversifier because at low absolute risk levels a 1 per cent increase in risk requires a less than 1 per cent increase in mean return to keep the wealth holder on the same indifference curve; whereas at high risk levels a 1 per cent increase in risk requires a larger than 1 per cent increase in mean return to keep the individual on the same indifference curve. Thus this wealth holder diversifies. For the opposite reasons, the individual whose preferences are shown in Figure 4.2(b) is a plunger. The individual whose preferences are shown in Figure 4.2(c) is a risk lover since she will trade off return for a higher risk to increase the chances of her capital gain. Let us analyse the choice problem in this situation. Suppose we are dealing with a risk averter who is a diversifier. She can hold her speculative balances in cash, earn zero return and incur zero risk, or she can invest some in the risky asset. Figure 4.3 visualises the possibilities for this individual. In Figure 4.2 we have plotted the risk/return combination offered by the individual’s portfolio. With cash she gets zero risk and zero return. As she

88


Figure 4.2


89

Figure 4.3

increases the proportion of the risky assets in this portfolio she gets higher µ and higher σ. With interest payment on the risky asset at i1 0A1 shows the investment opportunities available to this individual. As a portion of assets held in R increases, the mean return increases and so does the risk. Hence 0A1 is upward sloping. With a higher interest rate i2, this opportunity locus twists to 0A2. With i1 the consumer comes to equilibrium at D2 and with i2 the equilibrium is at D2. The proportion in R can be read off from 0B in Figure 4.2. An increase in the rate of interest has a substitution effect that should encourage the individual to take greater risks. The income effect would make the person want more security, which would mean a reduction in risk. In Figure 4.3 it has been assumed that the substitution effect dominates. Under this set of conditions, then, the higher the rate of interest the lower the demand for money for speculative purposes. The individual’s demand for money for speculative purposes is negatively related to the rate of interest. The economy’s speculative demand for money is obtained by adding up the individuals’ demand for money. This is portrayed in Figure 4.4. The theory of the speculative demand for money outlined above was developed by Tobin (1958) and is a vast improvement on the approach taken by Keynes (1936, 1973). This approach is important also because it is a precursor of the modern portfolio approach for the demand for financial assets of which money is one constituent. For operational purposes we now have a demand for money function which says that the demand for real cash balances is positively related to real national income and negatively related to the rate of interest,

90


Figure 4.4

(4.30)

It is not necessary that this function be separable in its two arguments. Equation (4.30) provides the theoretical foundation for the demand for money function used in this book. ALTERNATIVE VIEW OF THE MONEY SUPPLY PROCESS The description of the money supply process discussed above would be categorised as the money multiplier approach. Its basic characteristic is that it links the monetary base to the supply of money in the economy. However, there are at least two other views as well. The Portfolio Choice model conceives of the money supply as an aggregate of different financial liabilities of the banking system including the central bank. In a fractional reserve system the supply of base money sets an upper bound to money supply. The actual supply of money is then determined by the portfolio choice of firms and households in the economy. An alternative view of the money supply process originates in Palley (1987) and is referred to as a Mixed Portfolio-Loan Model. This approach takes the post-Keynesian view that private initiatives in the banking system may help determine the supply of money. Such forces make money supply endogenous, thus making their targeting difficult in practice. There is some evidence that the money supply in some developing countries does follow this type of process. For an application to India see Rath (2001).


91

DISCUSSION QUESTIONS 1

2 3

4

5

Let us see the Baumol-Tobin model in action. Make an estimate of the amount of currency (as opposed to credit cards or cheques) you spend each month. Estimate the amount of time it takes for you to go to the bank or the automatic teller machine (ATM). What interest do you earn in the money you leave in your account? What does the Baumol-Tobin model say about the number of times you should visit the bank? Is your behaviour consistent with it? Why or why not? How would your answer change if there were no ATM and you had to go to the bank during working hours to withdraw cash? Explain how the demand for money would change if suddenly there was an increase in risk aversion of the population. How credible do you find the speculative demand for money in a world of ATM and credit cards? Do you think that the transaction demand for money is also affected by these innovations? Why? Suppose the economy under consideration is opened up to international capital flows. How does this affect the process of money creation discussed in this chapter? Gresham’s law of tastes says ‘bad money drives out good money’. As an illustration of this ‘law’ consider the implications of having counterfeit currency introduced into the money creation models of this chapter.

NOTE 1

For some evidence on the endogeneity of such multipliers in the Indian case see Jha and Rath (2002).

5


INTRODUCTION So far we have assumed that the economy we have been studying is closed, i.e., it does not have any interaction with the outside world. It is time to relax this assumption and realise that since most economies are open, we must understand the impact and design of macroeconomic policy in an open economy. Throughout we shall assume that we are dealing with a small open economy (SOE). Thus the economy being studied is too small to affect world prices of goods and services and world interest rates. This is a safe assumption to make for most developing countries. There are some exceptions, however. If India were to sell wheat in international markets, the world price of wheat would likely go down. However, these exceptions are rather rare. In this chapter we first introduce the foreign balance (FB) constraint into the IS-LM framework and then consider the design of optimal policies under fixed and flexible exchange rates. We shall also consider the so-called ‘assignment problem’. We also consider the IS-LM-FB model in its generalised form with variable inflation rates. THE FOREIGN BALANCE AND IS-LM ANALYSIS With an open economy the notion of macroeconomic equilibrium has to be reinterpreted. In the IS-LM framework, full macroeconomic equilibrium is established when there is simultaneous equilibrium in the goods and money markets and the rest of the world (ROW). We shall assume that we can condense all trading partners of this country into this convenient phrase ROW. The domestic currency (the currency of the developing country) is the rupee (Rs). The currency of the ROW is the dollar ($). We write the equation of the IS schedule as (5.1)

94


where I is investment, S is savings, G is exogenous government expenditure, X are exports, Z are imports, T are tax collections and r is the interest rate. The exchange rate between the domestic currency and the world currency (e) is defined as the price of one dollar in terms of rupees. Let P be the domestic price level and P* the foreign price level (asterisks refer to foreign magnitudes). Then we can define the real exchange rate as Exports are assumed to depend positively on the real exchange rate. When the real exchange rate goes up (the rupee price of a dollar goes up or the foreign price level goes up or the domestic price level falls) the goods of the domestic economy become more attractive abroad as the price of domestic goods to foreigners drops. Thus we write (5.2)

Imports are a function of the real exchange rate and the level of real national income at home, (5.3)

As the real exchange rate rises (the price of a dollar rises, the foreign price level rises or the domestic price level falls) home goods become more attractive to domestic residents as compared with foreign goods as the relative price of home goods falls. As national income rises demand for imports goes up because imports are assumed to be normal goods with positive income elasticities of demand. We also assume that the marginal propensity to import lies between zero and one, We write the equation of the IS schedule as

(5.4)

We shall write the money market equilibrium as (5.5)

where Ω is wealth. This formulation of the money market equilibrium will help us concentrate on the demand for money as an asset. The supply of money is assumed to be controllable by the monetary authorities. In an open economy with fixed exchange rates the money supply will increase whenever there is a surplus in the balance of payments (foreign exchange is part of the monetary base of the central bank), and whenever there is a deficit, foreign exchange will flow out and the money supply will decrease. Thus under fixed exchange rates we may write (5.6)


95

where D are nominal domestic assets and R are foreign exchange reserves. A rise in domestic assets unaccompanied by a decline in reserves increases the money supply. This would happen if the central bank purchased government securities. If the central bank purchased foreign exchange in order to stabilise the exchange rate, reserves would rise and so would the money supply. Finally if domestic assets rise and the foreign exchange reserves fall by an equal amount nothing would happen to the money supply. With flexible exchange rates, e would adjust continuously in response to balance of payments disequilibrium and equation (5.5) would hold continuously. Balance of payments equilibrium itself occurs when the net surplus on the current and capital accounts sum to zero. If this is not zero then there is a compensating change in official reserves. Ignoring interest income from abroad we can write the balance of payments surplus as when K is the net inflow of capital. The balance of payments is in equilibrium (equivalently there will be foreign balance or external balance) if (5.7)

We have already discussed the determinants of X and Z. We shall assume that the net inflow of capital is determined as (5.8)

In equation (5.8) E(∆e/e) is the expected appreciation of the foreign currency. The term inside the square brackets in equation (5.8) is the expected gain to a unit of foreign capital by being invested in the domestic economy. Since the domestic country is an SOE, r* is treated as a parameter. For the time being we shall further suppose that so equation (5.8) can be written as (5.9)

The equation for the FB can be written as (5.10)

In the r, y plane we can draw (5.10) as the FB schedule as in Figure 5.1. Let us first look at Figure 5.1(a). Suppose, now, that real income is higher than y0. Since e, P and P* are fixed, exports remain unchanged. But with a higher income, imports are higher. At the same rate of interest (r0) net capital imports are unchanged. Hence a point like ß is characterised by a balance of payments deficit. If the interest rate was to rise from r0, ceteris paribus, more capital would be attracted into the domestic economy and the balance of payments deficit would be eliminated. Thus the FB schedule which shows the r, y combinations giving balance of payments equilibrium is upward sloping.

96


Figure 5.1

Above the FB schedule at any level of income the rate of interest is too high to maintain equilibrium in the balance of payments. Hence r, y combinations above FB denote balance of FB surplus. For the opposite reasons below the FB schedule we must have balance of payments deficit. The slope of the FB schedule reflects the degree of international capital mobility. Figure 5.1(b) shows a situation characterised by perfect capital mobility. The domestic rate of interest cannot be different from the world rate of interest r0. In Figure 5.1(c) the FB schedule denotes completes capital immobility.


97

Figure 5.2

In Figure 5.2 we have plotted the equilibrium of the economy in the balance of payments and the money and commodity markets. a denotes the point of simultaneous equilibrium. In Figure 5.2 we have drawn the FB schedule flatter than the LM schedule. This denotes the high degree of capital mobility. McKinnon (1984) calls this a financially open economy. If the LM schedule were flatter than the FB schedule then the degree of capital mobility would have been low. McKinnon calls this a financially insular economy. It turns out that the distinction between a financially open and a financially insular economy is a useful one. The effect of monetary and fiscal policies under fixed and flexible exchange rates depend crucially on the degree of international capital mobility. The following discussion of macroeconomic policies is classified under two major headings: policy under fixed exchange rates and policy under flexible exchange rates. Within each category we make distinctions on the basis of international capital mobility. MACRO POLICY WITH FIXED EXCHANGE RATES Without any loss of generality let us write the IS-LM-FB model as (5.11) (5.12) (5.13) (5.14) (5.15)

98


where I0, S0, Z0, L0 and K0 represents constants. Write the IS-LM-FB schedules (in linear form) as (5.16) (5.17) (5.18)

where E1 and E2 are appropriately defined constants. We can now write the system as: (5.19)

so that in equilibrium (5.20)

where These results are independent of the money supply. Substitute ye and re in the LM equation to get the equilibrium money supply: (5.21)

The supply of money in equilibrium is constant. Its value is determined solely by the need to maintain a fixed exchange rate between the home and foreign currency. Nothing much can be done with monetary policy except to create disequilibrium and alter the level of foreign exchange reserves. This result is independent of the degree of international capital mobility. To illustrate this we consider the case of perfect capital mobility in Figure 5.3. The domestic interest rate is fixed at the world level r0. Domestic income is y0. The monetary authorities conduct an open market purchase to increase the money supply and the LM schedule shifts outwards to LM’. Point b now characterises ‘domestic’ equilibrium. However, at b there is a deficit in the balance of payments. Since exchange rates are fixed, the monetary authorities are obliged to support the legal value of the rupee. Hence the monetary authorities sell dollars at the existing exchange rate. This means that the money supply declines. A monetary survey shows the following changes in the balance sheet of the central bank of the country.


99

Figure 5.3

The LM curve must, necessarily, go all the way back and the old equilibrium must be restored. Hence monetary policy is completely ineffective. As remarked earlier, this result does not depend on the degree of international capital mobility and so will hold for all cases. The effects of fiscal policy, however, are very different. We realise that G is exogenous. Totally differentiating equations (5.20) and (5.21) we have the following multipliers: (5.22a)

(5.22b)

(5.22c)

We examine the effects of increasing government expenditure in a world with perfect capital mobility in Figure 5.4. Initial equilibrium is once again at a. Expansionary fiscal policy shifts the IS schedule to IS’. The new ‘domestic’ equilibrium is at b where there is a balance of payments surplus. This is the crucial difference between the effects of expansionary monetary and fiscal policies. The surplus at b means that there will be a capital inflow. The monetary authorities prevent the local currency from appreciating by buying dollars at the going exchange rate. This, then, induces an increase in domestic money supply. A monetary survey shows the asset-liability position of the central bank as:

100


Figure 5.4

Following this induced monetary expansion, the LM schedule must shift outwards all the way to LM’ so that we once again have r0 as the domestic interest rate and the balance of payments surplus is eliminated. The expansionary effect on y, however, is very powerful. The above analysis and equation (5.22) suggest that the degree of capital mobility is very important in determining the magnitude of the effects of and the expansionary fiscal policy. If capital is perfectly mobile multipliers become (5.23a) (5.23b) (5.23c)

Realise that the rise in income equals the simple multiplier as in the extreme Keynesian case (i.e. when ) because the money supply automatically adapts by rising in direct proportion to the change in income with the case balance ratio Ly being the factor of proportionality.


When capital is completely immobile,

101

and the multipliers become (5.24a) (5.24b) (5.24c)

The first two multipliers are equivalent to the classical closed economy result with The equilibrium money supply drops because the higher interest rate induces liquidity preference. Foreign exchange reserves drop by an equal amount. When capital mobility is less than perfect the multipliers written in equation (5.22) are relevant. Both the equilibrium level of income and the equilibrium rate of interest rise in response to the fiscal stimulus. However, as can be seen by inspecting equation (5.22c) the equilibrium money stock may rise or fall, depending on whether is greater or less than zero. If it follows that

which implies that the slope of the LM schedule is greater than the slope of the FB schedule. In this case capital is relatively mobile and the equilibrium money stock rises as foreign exchange is accumulated by the monetary authority. However, if the slope of the LM curves less than the slope of the FB curve. Capital is, therefore, relatively immobile, the equilibrium money stock declines and the monetary authority loses foreign exchange. THE FLEXIBLE EXCHANGE RATE MODEL With flexible exchange rates the IS equation may be written as (5.25a)

P and P* are still fixed. The LM equation is (5.25b)

The FB function is (5.25c)

We have three equations in three endogenous variables: y, r and e. Total differentiation of (5.25) yields

102


(5.26)

Solving, we have

where Hence, using Cramer’s rule we can calculate the multipliers as (5.27a) (5.27b)

These are equivalent to the closed economy multipliers. Moreover, (5.27c)

or, using (5.27a), we have (5.27d)

So the depreciation of the home currency increases as income change increases. When capital mobility is perfect and the multipliers become: (5.28a) (5.28b) (5.28c)

With perfect capital mobility no change in the interest rate is possible. The income multiplier is the same as the classical closed economy multiplier with


103

Figure 5.5

For completeness we analyse the effects of monetary policy with perfect capital mobility in Figure 5.5. We start at point a with full equilibrium. Expansionary monetary policy shifts the LM schedule to LM’. ‘Domestic’ equilibrium occurs at b, where we have an incipient deficit in the balance of payments. This deficit is incipient because in a world of flexible exchange rates, actual deficit or surpluses cannot last long. More relevant, however, is the fact that at b the rate of interest is lower than the world interest rate so that there will be an immediate capital outflow. The depreciation of the home currency because of the incipient deficit will increase exports and reduce imports. This will increase aggregate demand in the goods market and the IS curve will shift outwards so long as the domestic interest rate is below the world interest rate (i.e. to IS’). Final equilibrium occurs at c where y has increased substantially. Monetary policy, which was powerless in a world of fixed exchange rates, suddenly becomes very powerful with flexible exchange rates. Now let us look at the effects of expansionary fiscal policy. Expansionary fiscal policy shifts the IS curve to IS’ in Figure 5.6. Domestic equilibrium moves from a to d. The domestic interest rate rises, so that there is a net inflow of capital into the domestic economy. This causes a capital account surplus and the domestic currency appreciates. Correspondingly, exports fall and imports rise and the IS curve shifts to the left all the way back to its old position. There is no increase in y. Thus fiscal policy, which was so powerful in the case of perfect capital mobility and fixed exchange rates, is rendered powerless.

104


Figure 5.6

An increase in the degree of capital mobility raises the money income multiplier because (5.29a)

The multipliers with respect to a change in government purchases are: (5.29b) (5.29c)

Hence, in general, expansionary fiscal policy raises national income and the rate of interest. However, (5.30)

which may be positive or negative because it follows that

whereas

If

The term on the left-hand side of the inequality is the slope of the FB schedule, whereas the term on the right is the slope of the LM schedule. In the case above, the slope of the LM is greater than the slope of the FB. This


105

is the case of relatively mobile capital in which case the local currency appreciates because expansionary fiscal policy causes an incipient surplus. In the absence of capital mobility and equation (5.29) yields (5.31a) (5.31b)

These equations are identical to the closed economy case. Further we must have (5.32)

so the local currency depreciates since fiscal policy affects only the current account. With perfect capital mobility so that we must have (5.33a) (5.33b) (5.33c)

Hence the local currency cannot depreciate since the capital account effect of the fiscal policy is dominant. Consequently, with flexible exchange rates, monetary policy becomes very effective whereas fiscal policy loses some of its potential. This result depends upon the degree of international capital mobility. As the degree of international capital mobility increases the expansionary effect of monetary policy improves and that of fiscal policy declines. In the extreme case with perfect capital mobility, monetary policy is doubly effective and expansionary fiscal policy has no effect on output. We can collect these results in Table 5.1. The assignment problem The economic policy maker often has several objectives. Sometimes these objectives may be in harmony with each other and at other times they are in conflict with each other. Suppose a country is faced with a balance of payments deficit and unemployment. Suppose the policy maker has only demand management policies at his disposal—fiscal or monetary policy, now labelled aggregate demand policy. If the policy maker attacks the problem of

106


Table 5.1 Effects of monetary and fiscal policy under alternative exchange rate regimes

unemployment, he would pursue expansionary demand policies. The consequent increase in real income would increase imports and, if expansionary monetary policy lowered the interest rate, the lower interest rate would cause a capital outflow. Hence the balance of payments deficit would become worse, not better. Such situations have traditionally posed a dilemma for policy makers. A general approach to the problem of economic policy making was formulated by Jan Tinbergen. It was popularised by Robert Mundell. The Tinbergen-Mundell approach to economic policy making is as follows. First it is argued that there must be at least as many independent policy instruments as policy targets. Second, a policy instrument should be assigned to that policy target on which it has maximal effect. Now, we examine some aspects of this assignment problem. We shall first examine this under a fixed exchange rate regime and then in a regime of flexible exchange rates. THE ASSIGNMENT PROBLEM WITH FIXED EXCHANGE RATES The first part of the Tinbergen-Mundell solution to the assignment problem says that we must have at least as many independent policy instruments as policy targets. Let us try to understand what this means. Suppose an economy wanted to attain full employment with price stability (internal balance) (we shall make the simplifying assumption that the price level starts rising only when the economy reaches full employment; at all levels of employment below the full employment level, the price level is unchanged) and balance of payments equilibrate (external balance). By the Tinbergen-Mundell rule there must be at least two policy instruments that


107

must be independent of each other and affect internal and external balance differently. Suppose we choose the level of government expenditure (fiscal policy) and the rate of interest (monetary policy). We know that the level of government expenditure and the rate of interest are not very independent of each other. Higher government expenditure, ceteris paribus, will lead to a higher interest rate. But we can suppose that they are not so related to each other as to imply that a change in one will necessarily and always imply a change in the other. Hence we shall treat them as (relatively) independent policy instruments. In the same vein, changes in the stock of money and the interest rate cannot be treated as independent policy instruments. Now the government expenditure (G) and the rate of interest (r) affect both internal and external balance. A higher G increases real national income and the increased income will increase imports and cause a deterioration in the balance of payments. Expansionary monetary policy will lead to a lower interest rate that will stimulate investment demand and hence raise national income. However, the lower rate of interest will also lead to a capital outflow and a larger deficit in the capital account of the balance of payments. Moreover, it can be argued that the higher real income will also lead to higher imports and, therefore, cause a further deterioration in the balance of payments. The point, however, is the relative magnitude of the effects of the two policy instruments on the policy targets. Fiscal policy has a greater effect on internal balance than on external balance and interest rate changes have a greater effect on external balance than on internal balance. This is explained in Figure 5.7. Figure 5.7 plots government expenditure (G) on the x axis and rate of interest (r) on the y axis. Suppose that at point a in the diagram the economy enjoys internal as well as external balance. Starting from point a, suppose we move to point b where the level of government expenditure is higher than G (the level consistent with internal balance) and the interest rate is still r (the level consistent with external balance). Because government expenditure is higher at b than the level required to maintain full employment, it must be true that there are inflationary pressures at b. Moreover, since income is higher at b than at a it must be the case that imports are higher at b than at a. Hence there must be a balance of payments deficit at b. Now let us try to restore internal and external balance starting from b. Suppose the interest rate is raised. As the interest rate rises two consequences follow. First, more capital is attracted into the domestic economy so that there is a surplus in the capital account. Hence the balance of payments starts moving towards equilibrium. Second, the rise in interest lowers investment and hence increases aggregate demand and inflationary pressures. Thus if the interest rate rises from b we move toward internal balance and external balance. However, the rise in the interest rate has, by assumption, a greater effect on external than on internal balance, i.e. external balance is reached before internal balance. In other words, the external balance schedule (EB) is flatter than the internal balance schedule (IB).

108


Figure 5.7

To the right of IB the level of government expenditure is too high (given the rate of interest) to maintain full employment. Hence there must be inflationary pressure. To the left of IB the level of government expenditure is too low (given the rate of interest) to maintain full employment. Hence there must be deflationary pressures. To the right of EB at any rate of interest the level of government expenditure and, hence, national income is too high to maintain balance of payments equilibrium. This means that points to the right of EB are characterised by excessive imports and hence a balance of payments deficit. Conversely, to the left of EB there are too few imports and hence there is a balance of payments surplus. In Figure 5.7 we have plotted the IB and EB schedules and labelled the disequilibrium zones. Now we consider the second aspect of the Tinbergen-Mundell approach. If the government knew the IB and EB schedules it would follow policies to have interest rate and level of government expenditure G. The point, however, is that the government does not know the EB and IB schedules. When it is faced with a disequilibrium situation it must know how to react without prior knowledge of the exact locations of the IB and EB schedules. The policy advice here is to assign the policy instrument (monetary policy) that has a greater effect on EB—the exclusive task of maintaining external balance. Analogously, since fiscal policy has a greater effect on internal balance, the task of maintaining internal balance is assigned exclusively to fiscal policy. This is the solution to the assignment problem. To consider the significance of this rule consider point c in Figure 5.7. Here we have a balance of payments deficit and deflation. We assign monetary


109

policy to external balance. The money supply is reduced and interest rates raised so that there is a capital inflow and the balance of payments deficit is reduced. We move toward external balance. At the same time, from c we adopt expansionary fiscal policy. G is increased, aggregate demand and national income rise. Hence, proper assignment of instrument to targets moves the economy toward internal balance and external balance. Suppose, however, that the policy authorities carried out an incorrect assignment of instruments to targets. At c the monetary policy would be assigned the target of achieving internal balance. This would call for a reduction in the rate of interest to expand employment. This reduction in the rate of interest causes a capital outflow and exacerbates the problem of external imbalance. Further, fiscal policy is assigned the task of attaining external balance. To reduce imports national income has to be reduced by reducing government expenditure or increasing taxes. This policy, by reducing aggregate demand, further exacerbates the problem of deflation. Thus by wrong assignment of policy to target we get further and further away from simultaneous internal and external balance. The student should carry out similar exercises of proper assignment in each of the disequilibrium zones. Proper assignment of instruments to targets is extremely important. THE ASSIGNMENT PROBLEM WITH FLEXIBLE EXCHANGE RATES In a world with flexible exchange rates one need not have to adjust the interest rate to get balance of payments equilibrium. As a matter of fact, if world financial markets are highly integrated, a country may have only limited control over its interest rate. In such a situation we must recast the assignment problem somewhat. The policy targets are the same as before; however, the policy instruments are different. They are now the exchange rate (e) and domestic absorption (A), which is equal to The options available to policy makers are shown in Figure 5.8. At point a we have simultaneous internal and external balance. Now suppose the economy finds itself at point b, where, at the same exchange rate e, domestic absorption is higher—higher domestic absorption than that required to maintain internal balance means that there are inflationary pressures. If foreign currency were to become cheaper (e were to fall) foreign goods would become cheaper in the home currency and home goods would become more expensive in terms of the foreign currency. Consequently, exports will fall and imports will rise, thus reducing aggregate demand and inflationary pressures. Hence the internal balance schedule in e, A space is downward sloping. To the right of IB at any exchange rate domestic absorption is too high so there are inflationary pressures. To the left of IB there are deflationary pressures.

110


Figure 5.8

Let us now go back to point b and compare it, once again, with point a. At b, because national income is higher than full employment (at exchange rate e), imports will be too high and a balance of payments deficit will ensue. If foreign currency is made more expensive (e is raised) exports will become cheaper in foreign currency and imports will become more expensive in terms of the home currency. Hence balance of payments equilibrium will be restored. Thus the EB schedule—the locus of e, A combinations giving external balance—will be upward sloping. To the right of EB at any exchange rate domestic absorption and, hence, imports, are too high so that there must be a balance of payments deficit. To the left of EB there must be a balance of payments surplus. The disequilibrium zones are labelled in Figure 5.8. To compare the problem in a flexible exchange rate regime with that in a fixed exchange rate regime, let us go to point c in Figure 5.8 where there is a balance of payments deficit and deflation. To cure the deficit the value to e can be allowed to rise. This also helps internal balance—with a higher value of e, ceteris paribus, exports should rise and imports fall. To attain internal balance A should go up. At d one should reduce e and increase A. A proper assignment of policies would, once again, be to alter e to attain external balance and to alter A to attain internal balance. DISCRETE ADJUSTMENT Policy adjustment, as postulated in the solution to the assignment problem, is smooth and continuous. Monetary policy is capable of being continuously adjusted, but a smooth road to equilibrium is not always possible because the effects of monetary policy are not entirely predictable and subject to variable lags. Fiscal policies can be changed only discontinuously.


111

Figure 5.9

Figure 5.10

We may illustrate problems raised by discrete or stepwise adjustment. In Figure 5.9 we consider the case of unstable discrete adjustment and in Figure 5.10 that of discrete stable adjustment. Let us start from a1 in Figure 5.9. We have external balance and deflation. To cure the deflation absorption is increased. We reached internal balance at a2, but we have a deficit in the balance of payments. To cure the deficit e is raised and so on. As illustrated in Figure 5.9 these oscillations are unstable. We follow the same equation of steps in Figure 5.10 but the adjustment is now stable.

112


From Figures 5.9 and 5.10 it is clear that whether discontinuous adjustment is unstable depends on the magnitude of the (absolute value of the) slopes of the IB and EB schedules. To see this formally let us write the internal balance as (5.34)

and let yf be the target level of income. Now because national income rises when absorption rises, but the sign of ye is uncertain. It could be positive if a rise in the exchange rate raises national income (by raising exports and imports). But the opposite can be true if the demand for inputs is inelastic with respect to price. External balance may be written as (5.35)

with Bf as the target current account surplus. A rise in absorption reduces the current account surplus, whereas a rise in the price of foreign exchange increases it. Totally differentiating equations (5.34) and (5.35) gives the slopes of the IB and EB schedules; and so that (5.36) (5.37)

If then IB slopes downwards in the e, A plane. In Figures 5.8 and 5.9 we have assumed We shall continue to assume this. We have been assuming that adjustments in absorption and exchange rate are discontinuous. We may assume that these changes take place in the following manner: (5.38) (5.39)

where A and e are the levels of absorption and exchange rate consistent with simultaneous internal and external balance. It is assumed that correct assignment requires absorption to target internal balance and exchange rate policy to target external balance. Let us write (5.40) (5.41)

Now,


113

(5.42) (5.43)

Now (5.44)

and (5.45)

Substitute equations (5.44), (5.45), (5.41) and (5.40) into equations (5.38) and (5.39) to get (5.46) (5.47)

which is a set of homogeneous, linear, non-homogeneous difference equations. Assume policy attempts to reach its target in one time period so that Substitute (5.47) into (5.46) to eliminate et. Similarly substitute equation (5.46) into (5.47) to write (5.48) (5.49)

These equations are linear second-order non-homogeneous difference equations. Notice that the coefficient of At-1 in equation (5.48) is bd and that the coefficient of et-1 in equation (5.49) is also bd. To eliminate the constant terms and make the equations homogeneous we define (5.50) (5.51)

substituting these into equation (5.48) and (5.49) we have (5.52) (5.53)

In equilibrium Ut and Vt are zero so that equilibrium values are (5.54) (5.55)

114


Substituting equations (5.54) and (5.55) into equations (5.52) and (5.53) we have (5.56) (5.57)

We try a solution like and in equations (5.56) and (5.57) and so that the characteristic equation to get is giving us roots and which are equal but of opposite sign. We may write the complete solution as (5.58)

and (5.59)

If bd would be positive and (bd)1/2 would be real. The sign of [(bd)1 /2]t would be positive and the sign of [-(bd)1/2]t would alternate between positive and negative values. Since neither root is dominant the system would oscillate. Stability of equilibrium requires the roots to be between -1 and +1. Therefore whence so that stability requires Now |d| is the absolute value of the slope of the external balance schedule and |1/ b| is the absolute value of the slope of IB. Hence stability would require the IB schedule to be steeper than the EB schedule. If not, the assignment of policies to targets is incorrect and destabilising and must be reversed. When the IB schedule is negatively sloped. Hence and (bd)1/2 is a complex number. We can write the complex conjugates as (5.60) (5.61)

where α is the real part of the complex root and ßi is the imaginary part. The solutions are now (5.62) (5.63)

Define the modulus of the complex roots as We may factor Mt from the term in parentheses and use DeMoivre’s theorem that states that to get The term in parentheses is an oscillating wave with constant amplitude. The growth or diminution of the cycles depends


115

on Mt. But so that stability depends on whether M is greater or less than one. The difference equation would be stable if the modulus of the complex roots are within the unit circle of the complex plane, i.e. or Hence stability requires that the absolute value of the slope of IB be greater than the slope of EB as in Figure 5.10. Hence we realise that when policy adjustments are discrete, the attainment of stability depends on the relative slopes of the IB and EB schedules. THE ASSIGNMENT PROBLEM IN DEVELOPING COUNTRIES We now attempt to analyse the assignment problem within the context of a ‘typical’ developing country. Our assumptions will reflect some stylised facts about developing countries. We assume that international capital mobility to and from the developing country is low. We shall also assume that the developing country in question is a small open economy. Our analysis here will follow the work of Salter (1959), Hansen (1973) and Prachowny (1984). A mathematical derivation appears in the appendix. Since the terms of trade cannot be affected it is possible to collapse exports and imports into one trade good called, simply, ‘tradable’. There is another good which is called ‘non-tradable’. The supply of tradable and non-tradable goods depends on their relative prices and, in some instances, upon the supply of real cash balance. The domestic economy is highly distorted so that actual production possibilities are less than potential. In Figure 5.11 PP is the potential production possibility frontier between tradables and non-tradables and AA is the actual production possibility frontier. Domestic demand for each of these two goods depends on relative prices, real income and money balance. The price of non-tradables adjusts quickly

Figure 5.11

116


Figure 5.12

to equal the domestic demand and supply of non-tradables. The domestic price of tradables is given by their world price multiplied by the nominal exchange rate. We also assume that net exports always adjust immediately to the excess of domestic production over consumption of the traded good. We can depict internal and external balance for this economy as follows. In Figure 5.12 we measure the price of tradables (Pt) and the nominal exchange rate (e) on the vertical axis. On the horizontal axis we measure the price of non-tradables (Pn). Internal balance requires quality of supply and demand for the non-tradable good since the assumed infinite elasticity of foreign demand means that imbalance in the market for tradables can easily be resolved. We depict the internal balance as II in Figure 5.12. As (Pn) rises, demand for non-tradables falls and supply rises. This excess supply manifests itself, by Walras’ law, as excess demand for tradables. If Pt rises this disequilibrium will be eliminated. Hence II must be upward sloping. The external balance line EE along which the traded goods sector is in equilibrium is also upward sloping. A rise in the price of tradables leads to an excess supply of tradables or, equivalently, an excess demand for nontradables. This can be eliminated if Pn rises. It can be shown that the slope of II is greater than one whereas the slope of EE is less than one. A proportional increase in both prices around a 45° line from the origin would leave relative prices and, hence, supplies unchanged. However, the increase in price would lower the real value of money balances and, therefore, demand. Equilibrium in the market for tradables requires that the price of non-tradables should rise less than proportionately in order to maintain equilibrium in the market for non-tradable goods. Similarly, the price of tradables should rise more than proportionately to maintain


117

equilibrium in the market for tradable goods. Hence the II schedule is steeper than the EE schedule. Figure 5.12 assumes that exchange rates are flexible. Exchange rates can be pegged if the government allows reserve movements (an outflow in the case of a balance of payments deficit and an inflow in the case of a balance of payments surplus) or borrowing. In such cases the EE schedule no longer determines the external balance. The external balance will be given by a horizontal line whose vertical intercept is equal to the value at which the nominal exchange rate is fixed. Domestic prices of tradables are simply equal to the nominal exchange rate multiplied by the world price of these goods. Internal and external balances: response to shocks First, suppose there is an adverse supply shock at home—say a crop failure. We analyse this situation in Figure 5.13. A crop failure will raise the price of non-tradables for every value of Pt. The II schedule, therefore, shifts to the right. The external balance schedule remains undisturbed because of the SOE assumption. With a flexible exchange rate the domestic currency would depreciate from A to B. The price of the non-traded goods rises more than the price of the tradables and the consumption of foreign goods rises and more resources are transferred to the production of non-tradables. If the supply shock is expected to be temporary a more appropriate policy would be to hold the exchange rate pegged at its value at A. This would hold down the domestic price of tradables and reduce hardship in the economy. The economy would move from A to C where there is a balance of payments deficit (since C is below the flexible exchange rate EE schedule). This deficit could be covered by temporary borrowings.

Figure 5.13

118


Now, let us examine the case where there is a difference between domestic and world rates of inflation. This means that the price of tradables rises for every Pn. This would mean that the EE schedule would shift up and the real exchange rate will drop. If the exchange rate were kept pegged the rise in domestic prices of traded goods would generate a surplus in the balance of payments, leading to reserve gains and monetary expansion. We now consider the case of changes in the terms of trade. A change in the terms of trade changes the definition of the traded good. Consider, first, a rise in import prices. The supply of foreign exchange is reduced as supply shifts from exportables to importables. The demand for foreign exchange rises or falls according to whether the demand for imports is inelastic or elastic since the supply of imports is, by assumption, infinite elastic. If demand is inelastic the exchange rate depreciates, further raising the home price of imports and that of exportables. This will lead to a shift of supply toward production of tradables and to a rise in the price of nontradables. The extent of this rise depends upon the strength of the complementarity/substitutability between the two goods. With pegged exchange rates and an inelastic demand for imports, all prices would rise less, since the depreciation of the currency is avoided (or postponed). If the import demand is sufficiently elastic to overcome the reduction in export supply, the exchange rate appreciates, reducing the magnitude of the increase in the import prices and lowering the domestic price of exports. The price of non-traded goods rises by less than in the inelastic case (if at all). Pegged exchange rates would cause import prices to rise more and prevent export prices from falling. Now let us consider a fall in export prices. The reduction in foreign exchange earnings will result in a depreciation of the exchange rate, partially offsetting the reduced domestic price of exports and raising the price of importables. The price of non-traded goods may rise or fall, depending upon the strength of the substitution towards exportables and away from imports. Pegged exchange rates would eliminate the depreciation, so that import prices would not rise. In this case, demand would shift toward exportables and supply toward non-tradables, reducing the price of non-tradables. Exchange rate policies We have examined the impact of various shocks on the internal and external balance. The economy is subject to various shocks from time to time. The optimal exchange rate policy, from a stabilisation point of view, depends on the point of shock to which the economy is most frequently exposed. For example, if most shocks are generated by changes in the level of world prices relative to domestic prices, the goal of external and internal balance will be served best by a flexible exchange rate policy, as it will tend to insulate domestic prices from the effects of fluctuation in world prices. For example,


119

Figure 5.14

Figure 5.15

in Figure 5.14 a country with no net capital imports is faced with random fluctuations in world prices. A flexible exchange rate could allow the EE curve to remain unaffected by world price fluctuations. But a pegged exchange rate policy would cause domestic prices of tradables to fluctuate from, say, GG to G’G’ so that prices, including those of non-tradables, would fluctuate from D to J. If, however, most shocks come from temporary fluctuations in harvests or other domestic production, a pegged exchange-rate policy achieved by use of borrowing from abroad is optimal. In Figure 5.15 we consider

120


fluctuations in domestic production of non-traded goods. Price fluctuations will be substantially reduced by pegging the exchange rate at the level indicated by the line GG since the range is reduced from LL to MM. If most shocks come from fluctuations in the terms of trade the situation is more complex. A GENERALISED VERSION OF THE OPEN ECONOMY MODEL We now want to consider a generalisation to the case where the rate of inflation is variable. In essence, this implies a reinterpretation of the IS-LM-AS apparatus in the case of a small open economy (SOE). We shall retain the same assumption as earlier. To repeat, these assumptions are as follows, (i) the home economy produces one good but consumes two goods—the home good and imports. These two goods are imperfect substitutes for each other, (ii) The home economy has a bond which is perfectly substitutable for the foreign bond when expected yields are equalised, (iii) Domestic currency is not accepted abroad, (iv) In international transactions the SOE takes world prices and interest takes as given. The open economy IS schedule Let S represent the natural log of savings, taxes and imports and I the natural log of investment; government expenditures are exports. We write S as a function of the natural log of output Y and τ, the log of the reciprocal of the real exchange rate where p is the log of the domestic price level, p* is the log of the world price level and ê is the log of the nominal exchange rate. We shall also refer to t as the terms of trade. (5.64)

where denotes the price elasticity of imports. The relative price effect also operates on exports. Hence we write (5.65)

πe is the expected rate of inflation and i the nominal rate of interest. An increase in τ rising from a higher p causes consumers to move away from domestic goods to foreign goods and, therefore, reduces exports. This explains the sign on ß12. Equating S and I we obtain the IS schedule as (5.66)

where and In Chapter 3 we had argued that a condition that is not changed by the inclusion of imports and exports. Hence the denominators of all the as are positive. We further assume that and thus ensure that


121

Figure 5.16

We know that an increase in τ will reduce exports and increase imports thus reducing the trade balance. At a particular level of τ, say τe, the trade balance is zero. In other words To allow for continuing inflation in the home country and the rest of the world and its effects on the terms of trade we can write (5.67)

where Thus τ increases over time if (5.67) into (5.66) we have

Substituting (5.68)

The new IS schedule is drawn in Figure 5.16. We should realise that unless τ is constant the IS curve (even in the non-stochastic case) does not stay stationary. For τ to stay constant π should equal If the demand for home goods will keep falling and, therefore, the IS curve will shift to the left. Conversely if demand for home goods will rise both at home and abroad and the IS schedule will shift to the right. Moreover, there is only one τ(τe) such that the current account is balanced. Such an IS schedule is shown in Figure 5.16 as IS (τe). We have, for reference, also drawn an IS schedule with a deficit at the current τ. The open economy LM schedule The demand for money is still determined by real national income and the nominal rate of interest. However, the domestic interest rate is tied to the world interest rate and the expected depreciation of the home currency through the relation

122


(5.69)

However, the money supply is not exogenously determined by the central bank. We know from earlier in this chapter that if we were operating in a fixed exchange rate system a deficit in the balance of payments would involve a fall in the money supply and a surplus would involve an increase in the growth rate of money supply. The LM curve in an open economy is written as (5.70)

and, in terms of growth rates, as (5.71)

In summary, then, the LM curve looks very much like its closed economy counterpart except for the important fact that the growth rate of the money supply is endogenously determined when the country operates under a fixed exchange rate system. The open economy aggregate supply curve In the open economy aggregate supply responds to two price indices: p, which is of interest to producers, and pc which is of interest to consumers/ workers. We retain the assumption that employment is constrained by demand for labour. Hence remains the relevant wage for determining actual employment. In terms of growth rates employment and output rise if the rate of producer price inflation exceeds the rate of growth of nominal wages, i.e. However, the supply of labour will be determined by and we may expect that workers will negotiate nominal wages to maintain their command over consumption goods, i.e. they will set where is the rate of inflation of consumer prices (part of which is the price of importables). It follows that the terms of trade will play an important role in determining labour and, hence, output supply. Since these terms of trade are subject to changes the supply of labour and, hence, that of output cannot be taken to be fixed as was the case in the closed economy model. We present labour market equilibrium in Figure 5.17. The demand for labour depends only upon which is measured on the vertical axis, but the supply of labour is related to We can write the supply of labour as

(5.72)

Thus the horizontal intercept in Figure 5.17 depends on ß13, which incorporates key factors in the work-leisure choice, as well as which is determined by the terms of trade. In this setting it is possible for a nominal


123

Figure 5.17

change to affect employment and output. To see this define the consumer price index as (5.73)

where is the fraction of consumption coming from home goods. Now suppose p* rises (price of imports goes up), i.e. there is a deterioration in the terms of trade. The labour demand schedule stays in place whereas labour supply shifts to the left. A new equilibrium is established at A’ where, from the firms’ point of view, real wages are higher and employment lower. Naturally equilibrium output Ye also drops. Hence we can write the AS schedule as (5.74) e

Since Y is no longer unique, it has to be dated and equation (5.74) incorporates the assumption that the economy starts with an equilibrium position but may not return there. Substituting for τ we have (5.75)

An increase in import price causes τ to fall and forces the AS curve to shift to the left. Unless there is a subsequent reversal in the terms of trade, the AS curve will stay in its new location and become the new Ye. We must be clear, however, that although output is determined by the terms of trade, equilibrium in the trade market is preserved at any terms of trade without policy intervention.

124


DIAGRAMMATIC REPRESENTATION—THE CASE OF FIXED EXCHANGE RATES The first thing we realise is that, since domestic interest rate is pegged to the world interest rate, the IS-LM apparatus does not determine the interest rate. Moreover, in addition to the inflation rate and real output the model must determine either the exchange rate or the balance of payments. From the interest parity condition (5.69), the IS equation (5.68), the LM equation (5.71) and the AS schedule (5.75) we can solve for Y, π, i and µ. The LM curve is no longer relevant for determining the aggregate demand since the domestic interest rate is pegged to world levels. The IS schedule determines a relation between π and Y with real interest rates held constant. The AD and AS schedules are drawn in Figure 5.18. In both schedules expectational variables are treated as exogenous for the time being. Further i* and π* are assumed to be determined in the rest of the world and impervious to home country actions. Finally, past values of τ and Ye are predetermined and α0 is controlled by fiscal policy. Thus π and Y remain the only endogenous variables and they are determined by the intersection of the AD and AS schedules in Figure 5.18. Once we know all this we can use the equation for the LM curve to solve for µ, given If µ is larger than previously there is a balance of payments surplus and the money supply expands. πc can be calculated from π* and π given θ.

Figure 5.18


125

FLEXIBLE EXCHANGE RATES With flexible exchange rates we substitute the interest parity condition into the LM equation which, then, becomes the AD schedule. The IS schedule becomes superfluous. The negative relation between π and Y comes from the money market alone. Suppose Y goes up. This will create an excess demand for money but with i* and ρe given, i cannot change. The only adjusting variable will be π. We have drawn the AD and AS schedules in Figure 5.19. Together they determine Y and π. We can then write the IS schedule as (5.76)

which has, on the right-hand side, all the information needed to calculate ρ. All the other variables can be calculated. The casual link, therefore, is exogenous i* and ρ determine i; then the LM and AS schedules solve for Y and π; finally the IS schedule determines ρ. POLICY ANALYSIS IN THE PRESENCE OF FOREIGN AND DOMESTIC SHOCKS The model we have considered so far has not dealt with stochastic disturbances. It is now time to change this. We can write down the stochastic version of the IS-LM-AS model as

Figure 5.19

126


(5.77) (5.78) (5.79) (5.80) (5.81)

All previously introduced variables have the same meaning. In addition to the shocks originating in the goods, labour and money markets at home there are foreign shocks as well. επ* is a shock due to random variations in foreign inflation and εi* is a shock due to random variations in foreign interest rate. A positive value of επ* is tantamount to an unanticipated increase in the foreign inflation rate. It causes the IS curve to shift outwards as, ceteris paribus, home goods become more attractive both to domestic residents as well as foreigners. εi* represents unanticipated changes in the foreign interest rate and affects the interest parity condition. Initially foreign shocks are represented by non-zero values for επ* and εi*. If the foreign event is a nontime shock to a particular market the trend values of π* and i* will be unaffected and the foreign shocks should then be treated as temporary; but if the event is permanent, it should be translated into changes in π* and i*. Instruments of monetary policy In a closed economy the central bank can control the money supply or the nominal interest rate. It should be understood that there can be only one control variable—the central bank cannot control both the interest rate and the money supply. If the shock originates in the money market, interest rate adjustment is the optimal policy response. If the shock originates in the goods market then money supply change is the optimal response. In the open economy the choice is not quite so simple since the foreign balance condition (FB) must be satisfied at all times if capital mobility is perfect. In equation (5.80), given i*, assuming that and since ρe cannot change in the wake of a disturbance to the system, there can be only one value of i that is consistent with FB. It is not possible, therefore, for the central bank to choose between control over the money supply and interest rates adjustments but it can choose between the money supply and exchange rate adjustments. In other words, in an open economy with perfect capital mobility, the monetary authorities can allow a shock to be absorbed by exchange rate adjustments (if exchange rates are flexible) with no changes in the money supply or through changes in the money supply (if exchanges are fixed) while the exchange rate is maintained at a pre-designated level. We depict these choices in Figure 5.20 which illustrates the FB condition.


127

Figure 5.20

The FB line is horizontal in i, Y space. There is only one domestic interest rate that is consistent with equality of expected yields between domestic and foreign bonds regardless of the value of Y. If the domestic interest rate was higher than that denoted by the FB line there will be a flood of capital into the country which would, in turn, put pressures on the home currency to appreciate (if exchange rates are flexible) or the money supply to rise (if exchange rates are fixed). Conversely, if the domestic rate is below the FB line there will be an outflow of capital from the home country leading to a drop in the value of the home currency (if exchange rates are flexible) or a reduction in the monetary base and, hence, money supply (if exchange rates are fixed). It thus follows that the interest rate cannot be different from that given by the FB condition. Once the interest rate is determined by the FB condition, the equilibrium level of output is derived from the intersection of FB with either the IS or LM curves. If the monetary authorities control the rate of monetary growth then the LM curve becomes relevant. If, however, the exchange rate is controlled, then the IS schedule becomes relevant and the LM schedule loses its relevance. Thus the FB and LM equation constitutes the AD schedules in the flexible exchange rate case and the IS and FB schedules in the fixed exchange rate case. In either case, the AD equations help us solve for Y and π. Suppose we are interested in finding out which exchange rate regime is better. How do we articulate this question? A fair way would be to find out the exchange rate regime that keeps Y closest to Ye in the presence of random shocks. We realise, of course, that in an open economy there are many values of Ye because of terms of trade changes. However, any movement in Y because of random shocks can be distinguished from changes that occur because of

128


Figure 5.21

terms of trade effects. Any movement from any Ye causes a disequilibrium in the labour market. We measure the efficacy of any exchange rate mechanism by looking at its capability to insulate the economy against various shocks. The initial IS, LM and FB schedules are drawn in Figure 5.21. In all three schedules we assume that es are zero. The three schedules intersect at G: there is only one combination of i and Y that satisfies the long-run equilibrium. Here there is no distinction between controlling the money supply or the exchange rate. In the long run there is only one combination of i and Y that satisfies the long-run equilibrium condition. We must have and We now analyse the effect of (a) a domestic monetary shock, (b) a domestic goods market shock, (c) a foreign monetary shock and (d) a supply shock. These stochastic events can be imposed on this long-run equilibrium and the effect on output or employment compared for the two exchange-rate systems. The system that leads to the smallest deviation of output from its equilibrium level is the best under these circumstances. Our modus operandi is as follows. We hold the rate of growth of money supply constant along with the change in the exchange rate. In other words, if the superscript 1 refers to the situation after the shock, and the superscript 0 to the situation before the shock, we ensure first that and then that We want to find out how Y1 deviates from Y0 when there is a random shock of a particular kind. Let us first deal with the case when the rate of growth of money is constant and the LM schedule becomes the AD schedule. With money market equilibrium we must have Subtract from both sides of this equation to get (5.78')


because and set

and to get

129

Now first difference both sides of (5.78')

(5.82)

The last two terms provide us with the source of a horizontal shift of the LM schedule. We know that from the FB condition. Substituting this into equation (5.82) we have (5.83)

The vertical shift in the AS schedule is given by (5.84)

We substitute this for

in equation (5.83) and solve for

to get (5.85)

This relates changes in Y to the εs via the parameters of the system. Turning now to the regime of exchange control, we have after first differencing the IS schedule (5.86)

which gives We know that

and εs as the source of the horizontal shift of that curve. (5.87)

Substituting into (5.86) we have the horizontal shift of the AD schedule as (5.88)

Now substitute for from the supply side, equation (5.84), and write the final change in output as (5.89)

We are now in a position to analyse the effects of various shocks. This requires an examination of equation (5.85) for the flexible exchange rate case and equation (5.89) for the fixed exchange rate case.

130


Domestic monetary shock Consider first the case where i.e. there is a monetary shock. We shall assume that all expectational variables are kept at their previous levels. We realise immediately that in the fixed exchange rate case a domestic monetary shock shifts the LM curve outwards and it immediately shifts back inwards since it is the IS, FB and AS equations that determine output. The LM schedule is irrelevant to the determination of equilibrium output with fixed exchange rates. To examine the impact under flexible exchange rates let us set all other shocks except εm equal to zero in equation (5.85). The LM shift outwards in

Figure 5.22


131

response to in Figure 5.22(a). So does the AD schedule in Figure 5.22(b). If the central bank does not alter µ, ρ will rise above its previous value to absorb the incipient deficit in the balance of payments, π and Y rise. The rise in π implies a shift back of the LM schedule to LM″. Its location is dictated by the need for the new Y(Y1) to be the same in the IS-LM-FB panel as in the AS-AD panel. The extra output is sold to foreigners because the depreciation of the home currency is not completely offset by the rise in the home rate of inflation. Further, τ also falls so that home goods become more attractive in the world market. From the point of view of workers this outcome is not desirable since all expectational variables remain at their previous levels whereas domestic inflation rose and the home currency depreciated. This leads to a reduction in their real incomes. It would therefore appear that if the home economy was very prone to money market shocks, it would be better to have fixed rather than flexible exchange rates. This conclusion holds irrespective of whether the shocks are transitory or persistent so long as they are monetary in nature. Domestic goods-market shock Let us suppose that there is a goods market shock and all other shocks are zero. An examination of equation (5.85) indicates that with flexible exchange rates there is no effect on output (εg does not appear in equation (5.85)). The IS curve shifts outwards because of the shock but this does not affect output since the IS curve plays no role in output determination. The home currency appreciates and this completely compensates for the good market shock. The IS curve shifts back inwards. We examine the situation with fixed exchange rates in Figure 5.23. A goods market shock shifts the IS schedule outward in part (a) and the AD schedule outwards in part (b). Both π and Y rise. As π rises, the terms of trade rise and demand shifts to foreign goods. This shifts the IS curve back inwards somewhat. Its exact position is dictated by the change in Y in the AS-AD part of the diagram. Thus with a goods market shock there is a rationale for allowing exchange rates to fluctuate. An examination of these two cases points to the desirability of not committing policy to one or the other type of exchange rate regime and preferring, instead, to have a managed exchange rate system. When monetary shocks are more important the economy should stick to a value of π and allow the money supply to fluctuate; and if goods market shocks are more important a value of µ should be adhered to. Foreign monetary shock Now assume that the rest of the world suffers a temporary positive monetary shock. In the domestic economy this will show up as and εi* since an

132


Figure 5.23

anticipated increase in the growth of the world money supply will reduce world interest rates and raise world inflation. The fact that implies that there is a shifting down of the FB schedule by the amount of this fall. There will be a capital inflow as domestic bonds become more attractive—temporarily. With fixed rate of monetary growth the LM schedule does not initially move as in Figure 5.24. But because of the shift of the FB line the AD schedule shifts to the left. Both π and Y fall. As π falls the LM schedule shifts to the right. The new equilibrium occurs at (π1, Y1). Let us move now to a situation of exchange rate control. With the IS curve shifts to the right as home goods become more attractive. The AD curve also shifts to the right by Both π and Y rise and the rise in


133

Figure 5.24

the home inflation reduces demand for home goods and shifts the IS schedule to the left beyond its original position. However, Y must rise. This situation is analysed in Figure 5.25. Initial equilibrium in parts (a) and (b) occurs at E. A foreign monetary shock shifts FB downwards and IS outwards. AD also shifts outwards. IS then shifts back. Domestic supply shock Let us analyse, now, the consequences of an adverse supply shock. Let us suppose that and analyse the consequences with monetary control in Figure 5.26 and with exchange rate control in Figure 5.27.

134


Figure 5.25

With control over the rate of growth of money supply an adverse supply shock has the following effects. The AS schedule shifts to the left, raising π and lowering Y. The higher π shifts the LM schedule to the left. In final equilibrium π is higher and Y is lower. When the central bank controls the exchange rate an adverse supply shock shifts the AS schedule to the left and π rises. This leads to a change in the terms of trade and home goods become less attractive. This shifts the IS schedule to the left. Hence, once again, π is higher and Y lower. We conclude, therefore, neither fixed nor flexible exchange rates provide a defence against a change in the world interest rate. Further, neither provide any protection against domestic supply shocks.


135

Figure 5.26

Random shocks and the exchange rates Let us finally try to understand the exchange rate effects of various random shocks. Let us take first difference in the IS equation of (5.77) and use the terms of trade definition of equation (5.81) to write (5.90)

Substituting the AS curve for into (5.90) yields

and then substituting equation (5.85)

136


Figure 5.27

(5.91)

This equation links changes in the exchange rate directly to the shocks. The first thing to notice about this expression is that all three domestic shocks as well as both foreign shocks affect the exchange rate. Second, some


137

shocks, e.g. the domestic monetary shock εm, have a determinate effect on the exchange rate whereas others do not. For example, only if the sign of εi* is opposite to that of εp* can the effects of foreign shocks be determinate. Further, the term attached to εs cannot be unambiguously signed. This is because of two conflicting effects. On the one hand, the AS curve shifts downwards and reduces domestic inflation which lowers the terms of trade and makes home goods more attractive and the resulting trade surplus puts pressure on ρ to fall. On the other hand, the higher income raises domestic demand for goods, leaving less available to foreigners and putting upward pressure on ρ. CONCLUDING COMMENTS In this chapter we have discussed a generalisation of the IS-LM-FB model. It has often been argued that this representation is inadequate for a developing economy. We consider this argument and some alternative models in Part III. APPENDIX: MODEL OF INTERNAL AND EXTERNAL BALANCE Domestic demand for traded and non-traded goods depends on relative prices, real income and real money balance. Net exports of tradables (X) are equal to the excess of domestic production over consumption. We describe these in conditions in equations (A5.1) and (A5.2) respectively. (A5.1)

(A5.2)

where Di, Si represents, respectively, the demand and supply of commodity i, Y is real income, p is the aggregate price level and M is the nominal stock of money. These functions are assumed to be differentiable and homogeneous of degree zero in the nominal variables (pt, pn, pY, M). The signs of the partial derivatives are given below the arguments. Money income (pY) is equal to the factor cost of production plus the excess of government spending (G)over taxes (T). Thus (A5.3)

the price level is weighted average of the price of tradables and non-tradables, (A5.4)

with and We can write the balance of payments as the domestic value of exchanges in the foreign reserves (∆F), being equal to the sum of net exports (ptX)

138


and exogenous net capital imports (K)valued at the current exchange rate (e), i.e. (A5.5)

We assume that all foreign exchange is held by the central bank and that change in the domestic money supply (∆M) is equal to foreign exchange accumulation plus the government deficit plus credit granted to domestic business (A5.6)

Finally the domestic price of tradable goods is equal to the exogenous world price of tradables pft times the value of the world currency in terms of the home currency. Thus foreign demand for tradables is definitely elastic: (A5.7)

Under pegged exchange rates the value of e is fixed and these seven equations suffice to determine the seven variables pt, pn, p, Y, M, X and F. The solution of this model in the case of flexible rates is also quite similar. Since from equation (A5.6), the money stock is an exogenous variable controlled by the central bank. Net exports must equal the value of capital outflow Then equation (A5.3) can be substituted into (A5.1) and (A5.2) for money income, leaving two equations in two unknowns, the price of tradables (or, equivalently, the exchange rate) and non-tradables. DISCUSSION QUESTIONS 1 2

3

What would cause the world interest rate to change? How would you expect a rise in the productivity of capital abroad to affect the world interest rate? What effect would this have on domestic income and employment? Suppose money demand depends on consumption rather than income. Consumption itself is a function of after-tax income. Analyse the effect on the equilibrium in the Mundell-Fleming model of a cut in the tax rate.

6


INTRODUCTION We have examined the effects of expansionary monetary and fiscal policies (changes in aggregate expenditure) on macroeconomic equilibrium. In addition to such policies there is another set of policies that is often available to the policy makers. This set is called the set of expenditure switching policies. Such policies include exchange rate adjustment (devaluation or revaluation), import controls, tariffs and the like. We saw in Chapter 5 that, in a flexible exchange rate regime, expansionary monetary and fiscal policies had implications for exchange rate adjustment. In a fixed exchange rate regime, exchange rate adjustment often becomes necessary to meet policy objectives. To see this, let us suppose we live in a world of fixed exchange rates and the policy authorities wish to maintain full employment (internal balance) and balance of trade equilibrium (external balance). (We specifically abstract from capital account considerations.) Suppose further that the economy is facing unemployment and a balance of trade deficit. Expansionary aggregate demand policies would increase employment and real income. As real income expands, imports rise. However, with a fixed exchange rate and fixed domestic and foreign price levels, exports do not change. Hence the balance of trade worsens. In this kind of situation the goals of internal and external balance are in conflict with each other. Aggregate demand or expenditure changing policies are going to be inadequate. If this economy could devalue its currency and pursue expansionary aggregate demand policies it could simultaneously attain internal and external balance. The devaluation would encourage exports and discourage imports and the expansionary aggregate demand policy would increase real income. This would be an expenditure switching policy. Chapter 5 also showed instances when it was not possible to pursue the objectives of external and internal balance with just expenditure changing as the policy instrument. Expenditure switching is also often necessary. For developing countries expenditure switching often takes the form of devaluation.

140


In this chapter we examine the extant theories of devaluation as they apply to developed countries. In Chapters 7–9 we shall study some integrated models of balance of payments adjustment in developing countries. Even if (occasional) exchange rate adjustment is allowed for, does it necessarily follow that the devaluation will necessarily reduce the balance of payments deficit? A clear answer to this question is terribly important for a country contemplating exchange rate adjustments in a world of fixed exchange rates. At present, the currencies of several developing countries are tied to a major international currency or a basket of such currencies. Hence the above question remains significant for developing countries even in a flexible exchange rate regime. We will first study two early approaches to devaluation. One of them (the elasticity approach) focuses on the terms of trade effect of a devaluation. The second (the absorption approach) concentrates on the income-expenditure effects of a devaluation. Later we shall examine broader approaches to devaluation—those that combine current and capital account concerns. THE ELASTICITY APPROACH TO DEVALUATION The elasticity approach is the oldest modern approach to analysing the effects of a devaluation. We are here interested in finding out whether a devaluation increases earnings of foreign exchange. If it does, then a devaluation would be considered beneficial, otherwise it would not. Write the net exports in foreign exchange ($ terms) as (6.1)

where NX* is net exports in dollar terms, X* and Z* are, respectively, exports and imports in dollar terms. Total differentiating gives (6.1a)

Write where pf is the dollar price of exports and x is the physical quantity of exports, and define pd as the home price of exports so that where e is the exchange rate. Exports are bought by foreign nationals, so that elasticity of demand for exports (ηx) must be stated in terms of dollar prices, i.e. (6.2)

Since the home country supplies exports, the elasticity of supply of exports must be defined in terms of home (rupee) prices. Therefore, the elasticity of supply of exports with respect to price (ηx) is given by (6.3)


From

141

we get (6.4)

Substitute this into the expression for Sx to get (6.5)

Equation (6.5) expresses supply elasticities in terms of foreign prices and the exchange rate. From equation (6.2) we can write (6.6)

Substitute this into equation (6.5) to get (6.7)

if i.e. if there is a devaluation of the home currency. Substitute this into equation (6.2) to calculate the proportionate change in the quantity of exports as (6.8)

if

i.e. if there is devaluation. Equation (6.7) shows that dpf/pf is negative. The foreign price of exports therefore falls. Adding the proportionate change in foreign price to the proportionate change in quantity given the proportionate change in the value of exports we get (6.9)

This will be positive if the proportionate rise in quantity exceeds the proportionate decline in price. Substituting from equations (6.7) and (6.8) into equation (6.9) we get (6.10)

as the proportionate change in the value of exports. If we divide both sides of equation (6.10) by de/e we shall get the elasticity of supply of foreign exchange with respect to the exchange rate as This elasticity is positive if the demand for exports is elastic. In Figure 6.1 we plot the demand and supply of dollars. Equation (6.10) tells us that the

142


Figure 6.1

supply schedule is upward sloping. If the supply curve will be downward sloping. The supply of foreign exchange would be constant if the elasticity of supply is zero, i.e. In this case, the domestic price of exports rises in direct proportion to devaluation, i.e. there is no decline in the dollar price of exports and, therefore, no increase in volume of exports. Further, if ηx is high it would be advantageous to have a high Sx. But if ηx is low, it would be better to have a low Sx so as to minimise the reduction in foreign exchange earnings. In the ultra-Keynesian case, if the elasticity of supply of foreign exchange rate is Devaluation increases foreign exchange earnings if but decreases foreign exchange earnings if Now let us look at imports. The value of imports in dollar terms is (6.11)

where z is the physical volume of imports and pF is the foreign price of imports. The domestic price of imports is pD. We define

From equation (6.11) it follows that


143

(6.12)

From

we have that (6.13)

Using equation (6.13) and the definition of the elasticities

we have that (6.14a)

(6.14b)

From equation (6.14) we have that devaluation reduces both the physical volume of imports and the foreign price of imports so that the value of imports must decline. Substituting the results in equation (6.14) into (6.12) yields (6.15)

Hence, devaluation reduces the value of imports and the demand for foreign exchange. The demand curve for foreign exchange, therefore, is negatively sloped throughout as shown in Figure 6.1. Substituting from equations (6.15) and (6.10) into equation (6.1a) we get the overall balance of trade measured in foreign currency (6.16)

This is the final expression for the effect of devaluation on net foreign exchange earnings. A much discussed special case arises when supply elasticities are infinite and trade is initially balanced Then equation (6.16) reduces to (6.17)

For it is necessary that i.e. the sum of the elasticity of demand for exports and the elasticity of demand for imports should exceed unity. This is the so-called Marshall-Lerner condition.

144


To summarise then, devaluation is seen by foreign purchasers as an increase in supply. The foreign price drops and the quantity of the exports increases. But if the elasticity of supply of exports is infinite there is no increase in the domestic price of exports despite rise in demand. The foreign price of exports must, therefore, drop in proportion to the devaluation. The elasticity of supply of foreign exchange with respect to the exchange rate, therefore, is exactly the same as the elasticity of demand for exports. Importers see the devaluation as a reduction in supply. The domestic price of imports rises and the volume of imports falls. But if the elasticity of supply of imports is infinite, the additional demand causes no change in the foreign price of imports. Hence the domestic price of imports rises in the same proportion as the proportionate rise in the exchange rate. The elasticity of demand for foreign exchange then equals the elasticity of demand for imports. Another relevant small country assumption is infinite elasticity of supply of imports. The domestic economy’s demand for imports may be too small to affect world demand and price. On the export side it sells in a competitive world market, so the amount it sells has no effect on the world price. So far as the domestic economy is concerned the elasticity of demand for exports, ηx, is infinite. Then or, if trade is initially balanced, (6.18)

In this case devaluation is bound to increase net foreign exchange earnings. The devaluation expands export sales without reducing the foreign price of exports and reduces imports at a fixed foreign price. As one can imagine, economists have expended a great deal of time and effort to measure the supply and demand elasticities for exports and imports. An important difficulty in assessing the effects of devaluation on net exports is that it takes time for trade flows to adjust to relative price changes. In other words, elasticities are low in the short run and increase as time elapses after the relative price change. A very significant illustration of this fact is the oil crisis of 1972–73. The price of fuel oil quadrupled overnight. Elasticity of demand was very low initially. Many countries spent large amounts on the import of oil and this reduced demand for other goods. The world economy slipped into a recession. Over time, however, the elasticity of demand for oil increased. Industries and households learnt to economise on the use of oil. OPEC certainly is not as powerful today as it was in 1972–73. The analysis suggests that the time pattern of foreign exchange earnings in response to a devaluation may therefore follow a J curve, first declining and then rising. This is shown in Figure 6.2. At time t1 the devaluation occurs. Foreign exchange earnings first fall and then rise.


145

Figure 6.2

The J curve describes the time pattern of the balance of trade following a devaluation. The pattern of the effect of the devaluation on domestic absorption can also be similar. To continue with the example of the quadrupling of oil prices in 1973, initially consumers of imported oil found themselves spending a large proportion of their incomes on imported oil. Thus the demand for domestic goods (i.e. absorption) falls. As substitutes are developed for oil, the proportion of income spent on imported oil falls and, hence, domestic absorption rises. An obvious weakness of the elasticities approaches is its failure to go beyond the foreign exchange market to the economy as whole. This weakness leads directly to a discussion of the absorption approach to devaluation. ABSORPTION APPROACH TO DEVALUATION The absorption approach to devaluation also concentrates exclusively on the current account. This approach was launched by Sidney Alexander and further elaborated by Harry Johnson. Alexander criticised the elasticities approach for concentrating purely on the relative price effects of devaluation and neglecting the income effect. He defined domestic absorption (A) as consumption expenditure plus government expenditure:

so that

146


(6.19)

Net exports are then equal to national income minus domestic absorption. A devaluation can increase net exports only if it increased real income (Y) or reduced A. However, policy is complicated by the fact that what increases A often increases Y and a reduction in A often reduces Y. To see this point more clearly, suppose I and G are fixed. Devaluation increases exports and reduces imports and hence increases aggregate demand. and where c is the marginal propensity to So consume. Hence (6.20)

where s is the marginal propensity to save. Equation (6.20) makes it clear that devaluation will increase net exports only in the event that additional income is saved. It does not help to raise production if the resulting increase in income simply raises consumption by an equal amount. Resources will be released to expand net exports only if the rise in income carries with it a rise in savings. This necessitates a positive marginal propensity to save. If income taxation exists, then it also helps, provided the government does not consume the tax receipts. These conditions are likely to be met in industrial countries but are less likely to be met in less developed nations. The rise in real income that permitted net exports to expand was termed the idle resource effect by Alexander. However, one must subtract from this the loss in real income due to the deterioration in the terms of trade caused by devaluation. To understand this let us go back to equations (6.7) and (6.14) to get (6.21)

as the difference between the change in the foreign price of exports and the foreign price of imports. We know that equation (6.21) is negative. Hence

or

hence the terms of trade will deteriorate if Hence the product of supply elasticities must exceed the product of the demand elasticities. The lower is ηx the greater is the reduction in the foreign price of exports. Similarly, a reduction in the elasticity of demand for imports, ηz, means a smaller fall


147

in the price of imports. A country that is too small to influence the world price for its imports is bound to suffer a deterioration in the terms of trade when it devaluates its currency. The above suggests that there is a need to understand the terms of trade effect of devaluation within the absorption approach. We develop the following analysis. Write the macro equilibrium of the economy as (6.22)

We have put without any loss of generality. Totally differentiating equation (6.22) we have (6.23)

where and Further, without loss of generality, assume We can solve (6.21) to get

(6.24)

where

Now and The signs of de/dG and de/dMs depend on the sign of H. For de/dMs to be positive H must be positive. Now and whence Hence for H to be positive the Marshall-Lerner condition must be satisfied. If the elasticities analysis predicts a rise in net exports following a devaluation. This prediction will still be valid even when the adverse effect of income change on imports is taken into account. Thus an approach that incorporates both relative price and absorption effects is desirable. Recognising this, Harry Johnson suggested a general theory of the balance of payments, which says that, for deficit countries, an expenditure switching policy such as devaluation would have to be accompanied by expenditure reducing policies lowering monetary growth or reducing the budgetary deficit. The above analysis refers to a deficit country that is experiencing inflationary pressures. However, other kinds of disequilibria are also possible. These possibilities have been examined in Chapter 5.

148


Having said this, however, we must realise that although devaluation could improve the trade balance if the economy is at less than full employment, the prospects for successful devaluation are less favourable when the economy is already at full employment since total production cannot then be raised. A rise in net exports can then only be achieved by a reduction in absorption. Consequently, we must enquire whether devaluation can reduce absorption when the economy is already at full employment. Devaluation tends to raise the home currency prices of both exports and imports. These price increases feed through the economy and are reflected in a rise in the general price level. This rise in the price level may have expenditure changing effects and thus may change absorption. Let us look at some of the expenditure changing effects of the price level increase following a devaluation. The Keynes effect When the domestic price level rises, the real value of the economy’s money supply falls and the LM schedule shifts inwards. In other words, we travel leftward along the aggregate demand schedule in Figure 6.3. This effect, however, keeps the IS schedule fixed. The real balance effect If consumption expenditure depends upon the real value of cash balances then an increase in the price level will shift the IS curve (Figure 6.3). This will also result in a shift of the aggregate demand schedule as in Figure 6.3(b). The money illusion effect In countries with progressive income taxes based on personal money income, a rise in money income raises the average rate of income tax even if the rise in money income is caused entirely by inflation and not by a rise in real income. This will then reduce real after-tax disposable income and therefore depress consumption and absorption. This is due to the money illusion in the structure of income taxes, i.e. by the failure of the tax system to distinguish between a rise in real income and a rise in nominal income caused by inflation. The significance of this effect is likely to be less important in some developed countries where some indexation of income taxes to the price level is practised. In developing countries such indexation is rare. Inflationary expectations effect The above three effects tend to reduce aggregate expenditure. However, there are some effects that tend to increase expenditure. If people expect


149

Figure 6.3

higher prices in the future they will engage in anticipatory buying, especially of durable goods, in order to beat the expected price increase. If this happens, absorption and import demand will rise. Devaluation is normally seen as a supply restraint on importers. The domestic price of imports rises and quantity of imports is diminished. In the face of anticipatory buying, however, the demand curve for imports shifts to the right, adding to an increase in the price of imports as well as their prices. Foreign suppliers will see no reduction in demand and the foreign price of imports will not fall. Hence there will be no reduction in the amount of foreign exchange spent on imports.

150


On the export side, normally the devaluation will raise the domestic price of exports, therefore providing an incentive for exports and their suppliers to increase production. This may neutralise the price incentive supplied by devaluation. The excise tax effect Consider a developing economy that devalues its currency but discovers that the price elasticity of demand for imports is low. Combine this with the fact that this country is a small open economy so that a reduction in its demand for imports does not alter the world price of imports. Consequently the domestic price of imports rises by the same percentage as the rate of devaluation. Hence the value of imports in home currency rises. Sometimes this increase can be very sharp. Suppose now that the elasticity of demand for imports is low. The proportionate rise in the domestic price of imports vastly exceeds the proportionate reduction in the quantity imported. Hence more real income is spent on imports so that less income is available to be spent on other goods. This leads to a fall in output and employment. This effect has been highlighted by the structuralist school and will be considered in Chapter 8. The terms of trade effect This works against the excise tax effect. Devaluation usually involves a worsening of the country’s terms of trade and reduces real income. This will increase the fraction of income consumed if the consumption function is of the form with Hence real consumption in terms of domestic prices rises, and domestic absorption rises due to the change in the terms of trade. The sum total of the expenditure changing effects of a devaluation in an economy that is closed to full employment (or, more generally, an economy with low supply elasticities) depends on the relative strengths of all these effects. MONETARY AND ASSET MARKET APPROACHES TO BALANCE OF PAYMENTS The approaches to the balance of payments that we have discussed so far have concentrated exclusively on the current account. To arrive at a more comprehensive understanding we must include the capital account into the analysis. There is also the possibility that some of the conclusions we derived within the context of the current account may not hold when we take the capital account into consideration. Furthermore, it can be argued that in these days of highly integrated capital markets capital account effects can,


151

perhaps, occur faster than any effect that works its way through changes in the real national income. The monetary and asset market approaches, which we shall study now, have consistently denied any artificial segmentation of the balance of payments accounts. They have consistently argued in favour of treating the balance of payments as one analytical entity although distinctions between the current and capital accounts may be useful for some purposes. The principal difference between these two approaches lies in their treatment of the substitution possibilities in investors’ portfolios between foreign and domestic assets. Whereas the monetary approach believes that foreign and domestic assets are perfect substitutes, the asset market approach treats domestic money, foreign assets and domestic bonds as imperfect substitutes. Our analysis is expository in nature. In the next section we explore the monetary approach’s analysis of balance of payments under fixed exchange rates. Next, we apply the monetary approach to a flexible exchange rate regime and also attempt to understand the ‘overshooting phenomenon’. We then study the asset market balance approach of financial equilibrium under flexible exchange rates and, then, its explanation of the ‘overshooting’ phenomenon. The chapter ends with some concluding comments. The monetary approach with fixed exchange rates The monetary approach to the balance of payments rejects separate analysis of the current and capital accounts. It is a theory of the overall balance of payments. Let us write the overall balance of payments surplus as (6.25)

where is net imports, i*If is investment income from abroad (i* is the foreign interest rate, If is foreign investment by domestic residents) and K is the capital account surplus. The expression on the left-hand side of equation (6.25) is therefore the change in foreign reserves (∆R). Additions to foreign exchange reserves and addition to domestic assets (∆D) form the base for monetary expansion by the banking system (∆M). Thus foreign and domestic assets are assumed to be perfect substitutes. The monetary approach argues that, under fixed exchange rates, a balance of payments surplus implies an accumulation of foreign reserves, which can serve as the basis for monetary expansion. Conversely, a balance of payments deficit implies a loss in foreign exchange reserves which, therefore, leads to monetary contraction. To the proponents of the monetary approach the most important implication of a balance of payments disequilibrium is its impact on the supply and demand for money in the economy. In line with its counterpart (monetarism) for the closed economy, the monetary approach to the balance of payments views the supply and demand for money as the most important behavioural relation in the economy. Other

152


behavioural relations, like the consumption function, are relatively unimportant. We can write from equation (6.25) (6.26)

This is the fundamental equation of the monetary approach to the balance of payments. Money is created by expansion of domestic credit, ∆D), and by a balance of payments surplus, ∆R, since a surplus requires the monetary authority to purchase foreign exchange, thereby increasing money supply. If there must be a balance of payments surplus, because otherwise the supply of money cannot be brought into balance with the demand for money. Equation (6.26) also makes clear the key assumption of the monetary approach that domestic and foreign assets are perfect substitutes in investors’ portfolios. Consider now the nominal money demand function (6.27)

with (where i is the domestic rate of interest and Y is the level of real national income). From equation (6.27) we can write (6.28)

Equation (6.28) splits up changes in the demand for money into its various components. Hence we have (6.29)

Hence a rise in income causes an increase in the demand for money and a surplus. A rise in the price level increases the demand for money and, hence, increases the surplus. A rise in the rate of interest reduces the demand for money and may therefore cause a deficit. A rise in domestic assets, ceteris paribus, reduces the need for foreign exchange and causes a deficit. Realise that these predictions are exactly the opposite of those made by standard Keynesian theory. Keynesian theory would predict that (i) a rise in Y will increase imports and, therefore, cause a deficit, (ii) a rise in P would increase imports and cause a deficit, and (iii) a rise in i would increase capital inflows and hence cause a surplus. The monetary approach looks upon the balance of payments as a purely monetary phenomenon. Its break with Keynesian theory is fundamental and complete. It must be said, however, that the monetary approach has an advantage over the Keynesian approach. In flow equilibrium, a current account surplus may be offset by a capital account deficit so that the balance of payments is not in equilibrium. However, this equilibrium cannot be maintained because a current account surplus implies that domestic residents are accumulating foreign IOUs. Hence their portfolios are undergoing continuous change and


153

this will feed back into changes in the balance of payments. The monetary approach recognises this stock adjustment and has, therefore, an advantage over the Keynesian approach. We now undertake an analysis of the monetary approach under fixed exchange rates. The monetary approach’s analysis of the fixed exchange rate regime makes three key assumptions. First, it is assumed that there is perfect capital mobility. This is actually a consequence of the assumption that domestic and foreign assets are perfect substitutes. Domestic open market operations, which involve the purchase and sale of domestic bonds for money, have effects, which are identical to purchase, and sales of foreign assets. If the central bank buys bonds, D increases and so does M. If it buys foreign exchange, R increases and so does M. Since it is only the change in M that matters, open market operations and exchange rate stabilisation transactions are, in effect, identical. Hence Second, it is assumed that there is perfect arbitrage in commodity markets so that price levels world-wide are the same so long as we measure price levels in the same currency. In other words, we must have where e is the nominal exchange rate and P* is the foreign price level measured in foreign currency. This is also called ‘the law of one price’: the real exchange rate cannot be permanently altered. Third, it is assumed that wages and prices are flexible and will adjust to ensure that the economy equilibrates automatically (and always) at full employment. Competitive markets combined with the law of one price ensure market clearing because any excess supply of goods in the domestic market is absorbed in foreign markets. Thus aggregate expenditure expansion cannot raise the equilibrium level of real national income. Such increases can come about only through the expansion of aggregate supply by factors such as economic growth or productivity increases. Hence actual income where Yf is the full employment level of output. We can now analyse the implication of the monetary approach in the ISLM-FB framework of Figure 6.4. In this diagram full employment income is labelled Yf. To the right of Yf we have inflationary pressures and to the left of Yf we have recessionary tendencies. The FB schedule is horizontal at given the assumption of perfect capital mobility. Above FB we must have a balance of payments surplus, below FB there must be a deficit. In full equilibrium IS and LM intersect at Yf. Suppose that the foreign price level rises. Commodity arbitrage causes an immediate export surplus as domestic goods are diverted to foreign markets. This shifts the IS schedule to IS’ in Figure 6.4. ‘Domestic’ equilibrium occurs at k. At k the domestic price level starts rising. This would shift the LM schedule to the left. However, at k there is also a balance of payments surplus, which implies that reserves are being accumulated. This would shift the LM curve to the right. On balance, suppose that the LM curve stays unchanged. Hence the domestic price level keeps rising and reserves keep getting accumulated. The rising price level starts reducing the exports surplus and

154


Figure 6.4

the IS schedule starts shifting inwards from IS’. Once the price level has risen sufficiently the exports surplus is eliminated. Thus suppose P* has risen by x per cent. In final equilibrium P would also have risen by x per cent and the law of one price will once again hold. Now suppose that the world interest rate rises. This situation is analysed in Figure 6.5. Initially the world interest rate is and full equilibrium occurs at a1. The world interest rate now climbs up to a1 is now a position of balance of payments deficit. Hence the economy loses reserves, the money supply shrinks and the LM schedule shifts to LM’. Domestic equilibrium is now at a2. The deflationary pressure at a2 causes prices and wages to fall. This increases Y. This also increases net exports by arbitrage so that the IS curve shifts outwards to IS’. This move eliminates the initial deficit. Final full equilibrium is established at a3. Let us now analyse a situation in which economic growth increases the full employment income of the domestic economy. This situation is analysed in Figure 6.6. Initially the full employment level of income is Yf0 and full employment occurs at b1. Economic growth causes the full employment income to rise to Yf1. At b1 there is unemployment. Falling prices and wages shift the IS schedule to IS’. ‘Domestic’ equilibrium occurs at b2 and now involves a surplus in the balance of payments. The surplus means that reserves will accumulate and there will, hence, be an expansion of domestic money supply. An additional reason for the LM to shift rightward is the fall in the domestic price level. LM shifts outwards to LM’; Final equilibrium occurs at b3.


Figure 6.5

Figure 6.6

155

156


Figure 6.7

Let us examine the consequences of devaluation in the context of the monetary approach. This analysis is done in Figure 6.7. Initial equilibrium occurs at c1. A devaluation causes an export surplus as exports expand and imports shrink. This shifts the IS schedule to IS’. At c2 there is inflationary pressure in the economy and a balance of payments surplus. The surplus leads to an accumulation of foreign exchange, which will increase the money supply and shift the LM rightward. However, the rise in the price level will shift the LM leftward. On balance, let us suppose the LM stays in place. The rise in the price level will also reduce net exports. This will shift the IS’ back to IS. In final equilibrium (at c1) the domestic price level will have risen in proportion to the devaluation so that the law of one price is again established at the new exchange rate. Devaluation produces nothing more than a temporary balance of payments surplus. How realistic are the pronouncements of the monetary approach? Some predictions are almost uncannily accurate. Consider the following example. Between 1960 and 1980 industrial production grew at an average annual rate of 4.1 per cent in the USA. Over the same twenty-year period industrial production in Japan grew at a rate of 8.6 per cent—more than double the US rate. During the same period the rate of inflation was 7.5 per cent per annum in Japan and 5.2 per cent per annum in the USA. Standard Keynesian theory would argue that the higher Japanese growth as well as the higher Japanese inflation rate would tend to raise imports in Japan. Conversely US imports would be lower and the US dollar would be stronger compared with the


157

Japanese yen. However, the facts fly in the face of Keynesian theory. It was the Japanese yen that was stronger. The monetary approach has no difficulty in explaining these facts. It would argue that the higher Japanese growth rate would increase the demand for money in Japan, as would the higher Japanese inflation rate compared to the USA. This would, in turn, increase the Japanese surplus and strengthen the Japanese yen. It is true, however, that the monetary approach’s assumptions are hard to defend. For instance, Williamson (1983), writing about the law of one price, says that, ‘The hypothesis…has probably been rejected more decisively by empirical evidence than any other hypothesis in the history of economics.’ Similarly for most countries the FB function is probably steeper than the LM thus pointing to limited international capital mobility. The assumption of full employment is also suspect; moreover it can be argued that under some special assumptions the Keynes theory would give the same results as those offered by the monetary approach. An important question then arises. Are there special contributions of the monetary approach or is it to be regarded as a special case of the Keynesian approach? Kouri (1976) took up this question in an important article. Let us consider his analysis. From the relation subtract the current account surplus (CS). The left-hand side is the capital account surplus. So (6.30)

From equation (6.30) we can write (6.31)

Equation (6.31) represents an integration of the flow Keynesian approach and the stock adjustment monetary approach. It tells is that the change in the monetary supply depends on the change in domestic credit and the levels of P and Y. Hence we have the following conclusions: with fixed exchange rates and fixed P, P* and Y the current account surplus remains fixed while the interest rate is the only variable to equate the demand for money with its supply. If there is excess demand for money there will be a capital account surplus until the interest rate rises by enough to eliminate the excess demand. It follows that in the short run the capital inflow and the interest rate are closely linked. In particular we can write (6.32)

If we lived in a world of flexible exchange rates we would write

158


(6.33)

where E is the expectations operator. Hence there seems to be little conflict between the two approaches in terms of short-run analysis. Nevertheless there are two important differences. First from equation (6.32) we see that a given interest rate differential provokes a uniform flow of capital through time. The flow of capital depends on the level of the interest rate. According to the monetary approach this capital inflow, by raising the money supply, reduces excess demand for money, thereby rendering the capital inflow temporary. This implies that the capital inflow depends on the rate of interest as seen in equation (6.31). Second, if we use equation (6.32) to predict the change in the capital flow But if capital caused by a change in the rate of interest, we get mobility is perfect, Ki is infinite so that this expression would be useless for predicting the magnitude of the change in the capital inflow. Hence an equation like (6.31) is more useful for forecasting capital inflows. Suppose now that capital is completely immobile so that the inflow of capital is zero. From equation (6.31) we have (6.34)

The terms on the left-hand side of this expression add up to changes in the demand for money. Here the change in reserves is the current account surplus. Hence if capital is immobile and the exchange in domestic credit is fixed any change in money supply will require adjustment in the current account of the balance of payments. If P, Y and i are exogenously determined it must be the case that (6.35)

Thus the creation of domestic credit implies a current account deficit. This is a particularly important prediction for LDCs. We shall see in a later chapter the International Monetary Fund (IMF) uses this argument to rationalise its recommendation of a reduction of domestic credit by LDCs, which come to the IMF for assistance to tide over balance of payments difficulties. When the economy is in long-run equilibrium the demand for money equals the supply of money. P, Y and i are at their equilibrium values. If domestic credit does not change, the demand for money will equal the supply of money only if the current account is balanced. Thus


159

Hence when income is at full employment level, balance of payments adjustments with fixed exchange rates require price level adjustment. If a country wishes to avoid long-term balance of payments problems, it must hold domestic inflation close to the world average. Monetary approach with flexible exchange rates Under fixed exchange rates the balance of payments equation of the monetary approach is used to predict movements in foreign exchange reserves. When the exchange rate is flexible the balance of payments equation is used to predict movements in the exchange rate. The money supply is then free to be determined by conscious monetary policy. When analysing a regime of flexible exchange rates the monetary approach makes the following critical assumptions. (a) Portfolios adjust instantaneously to disequilibrium. Domestic and foreign financial assets are perfect substitutes so that there is no distinction between domestic and foreign assets (such as bonds). This assumption is sometimes called uncovered interest parity, i.e. (6.36)

which states that the expected rate of depreciation of the local currency must equal the difference between the domestic and foreign rates of interest, (b) The law of one price holds. Consumers see domestic and foreign goods as perfect substitutes and prices are flexible and determined in competitive markets. Thus To see the implications of flexible exchange rates consider the following example. Let the demand for money function be (6.37)

where M (log of money demand); P (log of price level); α is the income elasticity of demand for money ((dM/M)(dY/Y)); ß is the interest semi-elasticity of demand for money with respect to the rate of interest ((dM/ M)/di)); and Y (log of income). The foreign economy has an identical demand for money function (6.38)

The monetary approach assumes that α and ß are the same world-wide. Subtracting equation (6.38) from equation (6.37) we have (6.39)

Rearrange equation (6.39) with

as the dependent variable to get (6.40)

160


Now from the law of one price we have (6.40) to get

We use this in equation (6.41)

Equation (6.41) says that the price of foreign exchange will rise if the domestic money supply rises relative to foreign money supply, the price of foreign exchange will fall if domestic output rises relative to foreign output, the price of foreign exchange will rise if the domestic-foreign interest rate differential widens. Now substitute the condition for uncovered interest parity in equation (6.41) to get (6.42)

Thus a rise in the expected price of foreign exchange will raise the current price of foreign exchange. Equation (6.42) points to a potential source of instability: the higher the expected rate of appreciation of the foreign currency the greater the actual appreciation, and the greater the actual rate of appreciation the greater the expected rate of appreciation. If the expected rise in the price of foreign exchange equals the difference between the expected domestic and foreign inflation rates, we may write Substituting equation (6.42) we have (6.43)

This equation states that a relative increase in the domestic money supply raises the price of foreign exchange, and a rise in the relative expected rate of inflation rate also raises the price of foreign exchange. It is further assumed that full employment is continuously maintained both at home and abroad. We may now write equation (6.43) as (6.44)

which implies that the country with the higher real growth rate will experience appreciation. If differential interest rates equal expected differences in monetary growth rates, we have If we further suppose that the best prediction of future monetary growth rate is today’s monetary growth rate, we can write equation (6.44) as (6.45)

For a given full employment income differential the exchange rate depends entirely on differential money stock and their rates of growth. The countries with least rapid monetary growth rates will experience appreciation whereas countries with rapid monetary growth rates will experience depreciation.


161

The experience with flexible exchange rates, however, has been that exchange rates have been very volatile—much more volatile than any application of equation (6.45) would suggest. To explain this excessive volatility within the framework of the monetary approach is a challenge. This issue was addressed in a pathbreaking paper by Dornbusch (1976). Dornbusch argues that in the short run prices and wages are sticky. They become fully flexible only in the long run. Dornbusch assumes that equation (6.45) helps determine the long-run exchange rate eL. Thus we write (6.46)

The short-run exchange rate is denoted by e. When the short-run exchange rate is different from its long-run equilibrium value eL, the short-run exchange rate is assumed to change at a rate proportional to the difference between the two exchange rates. The part of this exchange rate due to this discrepancy is written as where γ is the factor of proportionality. Since the long-run, rationally expected, change in the exchange rate is the difference between the monetary growth rates, i.e. the overall expected change is (6.47)

Uncovered interest parity holds so that equation (6.46) and rearranging terms we have

Using this in

(6.48)

Now, since the expected inflation rates equal the monetary growth rates, it must be the case that and where r and r* are, respectively, the home and foreign real rates of interest. Hence the difference between the log of the short-term exchange rate and the log of the long-term exchange rate depends on the difference between the real interest rates: (6.49)

Substituting for eL from equation (6.49) we have (6.50)

From equation (6.49) we realise that as the real interest rate gap narrows, the gap between the long-run and short-run exchange rates narrows. Now, suppose there is a monetary shock in the domestic economy and the money supply grows unexpectedly. In the long run, when prices will have risen sufficiently, the old real interest rate will prevail. In the short run, however, prices respond sluggishly. Hence the monetary shock causes the domestic real interest rate to decline temporarily. Expected inflation is still the same as

162


the growth of the money supply. However, the actual inflation is below its long-run equilibrium value causes of sluggish prices. Thus the real interest rate falls temporarily. This causes a capital outflow, which raises the spot (short-term) exchange rate e. In general, whenever foreign and domestic monetary shocks are different from each other, one might expect differences in real interest rates internationally. This will cause short-run movements in the spot exchange rate unrelated to the forces discussed in equation (6.45). This makes for excessive volatility of the exchange rate as this has to retrace its steps as prices become more flexible in the long run. Dornbusch calls this the overshooting phenomenon. ASSET MARKET ANALYSIS OF FLEXIBLE EXCHANGE RATES The asset market balance (AMB) approach to the balance of payments tries to rectify (in the specific context of flexible exchange rates) some of the principal weaknesses of the Keynesian and monetary approaches to the balance of payments. The AMB approach recognises the fact that monetary approach represents a significant advance over earlier approaches in so far as it looks at the balance of payments as a whole and emphasises the role of stocks in balance of payments adjustments. But the AMB approach does agree with the view that the monetary approach makes some extreme assumptions. In particular, the AMB approach drops the assumptions that foreign and domestic assets are perfect substitutes. Even if domestic and foreign assets are similar (which is certainly not the case in most LDCs) investors may perceive differences in the risk caused by differences in liquidity, tax treatment, exchange risk, political risk and default risk. Another important reason why investors view domestic and foreign bonds as imperfect at any point in time is that international business cycles and national economic policies are not perfectly synchronised with respect to time. Short-run asset market equilibrium Among the important variables that affect the balance of payments are real income levels, price levels, interest rates and exchange rates. The AMB approach believes that total financial wealth may be an important factor affecting the balance of payments. To simplify the analysis we ignore the foreign developments and study the asset market in a single country. There are three assets: money M, domestic bonds B and foreign bonds F. Nominal wealth W is then (6.51)


163

where F is in foreign currency and eF in home currency. Since M, B and F are imperfect substitutes it follows that the rational asset holder’s portfolio will contain all three assets. We further assume that the level of national income is fixed so that the demand for assets depends on the nominal interest rate i and total nominal wealth W. We write the demand for money as (6.52)

A rise in the interest rate lowers the demands for money and a rise in wealth increases it. The demand for bonds is (6.53)

A rise in the interest rate lowers the price of bonds and, therefore, raises the demand for bonds. A rise in wealth will increase the demand for bonds. Finally we write the demand for foreign assets as (6.54)

Realise from equations (6.52), (6.53) and (6.54) that a rise in the interest rate raises the demand for B and lowers that of M and eF. By Walras’ law, then, the fall in the demand for money and foreign bonds must exactly equal the rise in the demand for domestic bonds, i.e. (6.55)

A change in the exchange rate affects asset demands by altering W. We may understand that a change in the exchange rate does not provide a relative price effect that leads to portfolio substitution. There is only the effect on wealth. The model consists of four equations in three unknowns: e, W and i. Hence one equation can be eliminated. For example, once the demand for money and foreign assets are known, the demand for domestic bonds would just be the residual The exogenous variables are nominal money supply M, the nominal value of the stock of domestically held foreign assets. For any value of W we can depict the equilibrium in e, i space as in Figure 6.8. Suppose that at point a there is simultaneous equilibrium in all three financial markets. Suppose that at the same exchange rate e the interest rate was higher. This would lower the demand for money so that at point b there must be excess supply of money. Now, starting from b, if the exchange rate were to rise this would raise eF and hence W and hence the demand for money. So the MM schedule, which shows combinations of e and i giving money market equilibrium, must be upward sloping. To the right of MM we must have excess supply of money. To the left of MM there must be excess demand for money.

164


Figure 6.8

Let us go back to point b and realise that at this point the price of domestic bonds is too low to give domestic bond market equilibrium. Hence, there is an excess demand for bonds at b. Now if the exchange rate were to fall the value of W would fall and the excess demand for bonds would be eliminated. Thus the BB schedule, which shows the combinations of i and e giving bond market equilibrium, must be downward sloping. To the right of BB we must have excess demand for bonds and to the left of BB we must have excess supply of bonds. By Walras’ law, then, the FF schedule showing combinations of e and i giving equilibrium in the market for foreign bonds, must be downward sloping and flatter than the BB schedule. The excess supply and excess demand conditions are as noted in Figure 6.8. We can demonstrate all this mathematically as follows. Our behavioural equations are

Differentiating the wealth constraint with respect to W we have


165

since Lw, Bw and Fw are all positive it follows that Now to obtain the slopes of the three schedules we totally differentiate each one of them with respect to e, i and W to get (6.56) (6.57) (6.58)

Now

so that (6.59) (6.60) (6.61)

From equation (6.59) the slope of the MM schedule is

From equation (6.60) the slope of the BB schedule is

From equation (6.61) the slope of the FF schedule is

Walras’ law requires that BB be steeper than FF. In fact it is easy to demonstrate that this condition must be satisfied to ensure stability. We are now in a position to analyse the comparative static properties of the model. Suppose the government runs a deficit and finances it by creating more money. So far as the private sector is concerned The effects are traced in Figure 6.9. Initially there is simultaneous equilibrium in all three markets at a1. The original financial market equilibrium relations are labelled M0M0, B0B0 and F0F0. Now the money supply and, hence, nominal wealth expands. Hence now at a1 there will be an excess demand for foreign and domestic bonds and a corresponding excess supply of money. (The rise in W increases the demand for money as well. Walras’ law, however, requires that there be excess supply of money at a1 to match the excess demand for domestic and foreign bonds.) To restore equilibrium in the money market at any exchange rate the interest rate must be lower. Hence the MM schedule shifts leftwards to M1M1. The excess demand for bonds at M1M1 can be eliminated by raising the price of bonds, i.e. lowering the interest rate. Hence

166


Figure 6.9

the BB schedule shifts down to B1B1. To eliminate the excess demand for foreign bonds the rate of interest must rise. Thus the FF schedule shifts right to F1F1. Final equilibrium occurs at a2 with a lower interest rate and higher exchange rate. We can illustrate the same results mathematically. Totally differentiate the BB, FF and MM schedules to get

(6.62)

The determinant of the matrix on the left-hand side is Now, from equation (6.62) we have

(6.63)

When the government finances its deficits by selling bonds to the central bank, the central bank will expand money supply to the same extent. On the right-hand side of equation (6.63) the column vector terms become


167

Figure 6.10

whence (6.64)

(6.65)

(6.66)

Money-financed budget deficits, therefore, lower the interest rate and raise the exchange rate. Recall that in the case of the monetary approach the domestic interest rate was determined completely by the world interest rate. Now suppose that the government deficit is financed by selling bonds to This increase in wealth the public instead of the central bank. Thus raises the demand for foreign bonds. Thus, higher F prices would be needed to eliminate excess demand. In Figure 6.10 we show this change with a shift of the FF schedule from F0F0 to F1F1. At a1 there is an excess demand for foreign bonds and an excess supply of domestic bonds. To eliminate the latter there must be a fall in the price of bonds (a rise of the interest rate). Hence the BB schedule shifts to B1B1. Further, at a1 there is an excess demand for money, which can be eliminated by shifting the MM schedule from M0M0 to M1M1. Final equilibrium occurs at a2. The rate of interest is definitely higher at a2 than at a1; however, the change in the exchange rate depends on the relative shifts of the BB and FF (and MM) schedules. Mathematically, the terms on the right-hand side of equation (6.62) would become

168


Hence (6.67)

(6.68)

(6.69)

One can see that the home currency will appreciate if i.e. if This condition may be satisfied if the rise in the interest rate leads to a larger substitution of domestic bonds. Now let us look at the effects of monetary policy. Suppose the central bank conducts an open market purchase of domestic bonds. Wealth does not change. Let us now refer to Figure 6.11. Initially equilibrium is at a1. The increased supply of money creates an excess supply of money. To restore equilibrium in the money market the interest rate must fall, i.e. the MM schedule shifts from M0M0 to M1M1. The reduction in the supply of bonds creates an excess demand for bonds. To restore equilibrium bond prices must rise, i.e. interest rates must fall. The BB schedule then shifts from B0B0 to B1B1. The equilibrium interest rate is lower and the exchange rate is higher.

Figure 6.11


169

Mathematically, the terms on the right-hand side of equation (6.62) become

whence we have (6.70)

(6.71)

(6.72)

The monetary authorities can conduct open market operations by buying and selling foreign assets. Suppose, for instance, the monetary authorities buy foreign assets. We analyse this situation in Figure 6.12. Initially the equilibrium is at a1. The government buys foreign assets. This creates an excess supply of money and an excess demand for foreign assets at a1. To remove the excess supply of money the interest rate must fall. Hence the MM schedule shifts from M0M0 to M1M1. The excess demand for foreign assets can be eliminated by raising the interest rate at every exchange rate, i.e. by a shift of the FF schedule from F0F0 to F1F1. Final equilibrium is at a2. Mathematically, the terms on the right-hand side of equation (6.62) become

Figure 6.12

170


from which we will get (6.73)

(6.74)

(6.75)

The AMB approach provides the additional insight that the balance of payments cannot be balanced unless the current account is balanced. It will not do to simply have an overall equilibrium in the balance of payments. This is because a current account surplus or deficit has implications for the accumulation of foreign assets. A current account surplus means that foreign assets are being accumulated whereas a current account deficit means the opposite. Suppose that the domestic economy has a current account surplus. This would mean that foreign assets will be accumulated and wealth would increase. Moreover, the balance between the three assets in the portfolios of investors is disturbed. Investors would like to convert the additional foreign assets into domestic bonds and money. This will tend to appreciate the domestic currency and thus lower the value of the increase in W. Since wealth holders increase their demand for money the rate of interest would tend to rise. However, since the demand for domestic bonds also rises, this would tend to raise the price of bonds, i.e. lower the interest rate. On balance, suppose the interest rate stays unchanged. Then the appreciation of the local currency will wipe out the increase in wealth. In the long run, then, a current account surplus will lead to a proportionate appreciation of the domestic currency whereas a current account deficit will cause an equiproportionate depreciation of the local currency. Asset market balances and overshooting If exchange rates are fully flexible then the overall balance of payments must be in equilibrium with any current account surplus (deficit) being matched by a capital account deficit (surplus). However, as we have realised, asset markets (and hence the exchange rate) will not be in equilibrium. Therefore, in the medium run, the balance of payments cannot be in balance unless the current account is also balanced.


171

Suppose that current income is at full employment level. Then the net inflow of foreign assets consequent upon a current account surplus can be written as

With i*F will keep rising and so, ceteris paribus, will the current account surplus. As domestic residents try to convert foreign assets into domestic bonds and money the value of the local currency rises. Over time, then,

In the short run suppose that a country such as the USA has large investment income from abroad. Then it can maintain a current account surplus even though its exports are less than imports, i.e. The AMB approach predicts, quite correctly, that such a country can continue to export less than it imports and yet experience an appreciation of its currency. This is precisely what seems to have happened in the USA in recent times. It continues to enjoy more imports than exports. However, it had such a large investment income from abroad that it had a current account surplus and the dollar continued to rise. Let us now try to understand the overshooting with output fixed at the full employment level Yf and, to simplify the analysis, with net exports and foreign investment equal to zero. i* and P* are treated parametrically. Further, suppose without loss of generality that initially the exchange rate and the domestic price level are each equal to 1, i.e., Now, suppose there is a monetary shock. The money supply increases suddenly to a higher level. As a consequence the exchange rate and the price level begin to rise. We analyse the situation in Figure 6.13. Initially and the supply of money rises. The economy is in a flexible exchange rate regime with flexible price level. The exchange rate immediately jumps up to e3 and then starts falling. The critical jump in the exchange rate creates an export surplus and investment income rises. Domestic residents starts converting foreign assets into domestic assets and so e starts falling. The jump in the money supply also causes the price level to rise. At time t2 the exchange rate and the price level are again equal to each other and net exports should again equal zero. However, because of the initial rise in e and the consequent accumulation of foreign assets, there would be a current account surplus at t2. Hence, after t2, the exchange rate must continue to fall and the price level continue to rise to balance the

172


Figure 6.13

Figure 6.14

current account. The new equilibrium value of e is e2, that of the price level P2. Unlike the monetary approach, the real exchange rate is permanently altered. Let us now suppose that the price level adjusts slowly. We analyse this situation in Figure 6.14. We have the same initial conditions as in Figure 6.13. At time t1 there is a monetary shock and the exchange rate jumps to e3 and this causes an export surplus. The price level also starts rising but its rise is slow. The exchange rate drops and, at t2 once again, e and P are the same. Net exports are again zero at t2, but because of the accumulation of foreign assets there is a current account surplus at t2. Hence e keeps falling and P keeps rising. Now, since the rise in the price level is slow, the exchange rate


173

may drop to a level below e2 before current account balance is attained. Let us suppose that this occurs at t3. However, the price level is still rising at t3 so that we now develop a current account deficit. As a consequence, the exchange rate now starts rising. We once again reach exchange rate e2 and price level P2; however, the exchange rate reaches its equilibrium value from below rather than from above. The real exchange rate is again permanently altered. CONCLUSIONS In this chapter we have examined various approaches to balance of payments adjustments. Both current account and integrated approaches have been examined. The immediate applicability of these approaches to developing countries is somewhat limited. Nevertheless, they serve as an important backdrop for the analysis conducted in Parts III and IV of this book. DISCUSSION QUESTIONS 1

2

3

Analyse the effects of the following on the current account, the real exchange rate and the nominal exchange rate, (i) A fall in consumer confidence leading to an anticipated reduction in consumer demand and rise in saving; (ii) the introduction of a more fuel efficient variety of Nissan sedans makes people prefer foreign cars to domestic cars; (iii) the introduction of ‘plastic money’ reduces the demand for currency. How will the current account and the real exchange rate respond to an increase in government expenditure? Does it make a difference whether this increase in expenditure is temporary or permanent? Suppose that the many foreign countries start to subsidise investment by instituting investment tax credit. What happens to (a) world interest rate; (b) world investment demand function; (c) home investment; and (d) home real exchange rate?

Part II Macroeconomic Models of Developing Countries

7


INTRODUCTION The International Monetary Fund (IMF) is often called to assist developing countries in financial crisis. To evaluate the needs of the developing countries as well as their progress in overcoming such crises, the IMF needs a conceptual framework. This is what we are calling IMF-type macro models for developing countries. This approach is important for several reasons, not the least of which is the fact that the global financial community’s perception of the performance of the developing country in question is often influenced by the IMF’s evaluation. Hence, the IMF view on their performance has profound effects on the prospects for developing countries. In this chapter we study a simplified version of the IMF’s approach and then a more formal model that has been built around it. Alternate views on the macroeconomic structure of developing countries are provided in the subsequent two chapters. THE BASIC IMF MODEL The intellectual foundation of the IMF approach to LDC stabilisation policies is the Pollack model. As Taylor (1987) has noted, the model of Pollack (1957) has influenced the lives of more people than any economic theory except for Keynes’ General Theory (1936). Pollack’s model has been revived and reinterpreted in a number of important contributions—see for example Khan (1990). The building blocks of the Pollack model are as follows: 1

A budget constraint for the private sector which states that the accumulation of financial assets must be the difference between what the private sector earns and what it spends on consumption and investment. Thus we have: (7.1)

178

2

Macroeconomic Models of Developing Countries

where Yp is private income, Cp is private consumption expenditure and Ip is investment by the private sector. Whatever the private sector does not consume or invest it must accumulate as financial assets. These are as follows: (a) change in money holdings, ∆M, (b) lending to the government, ∆NGBp, (c) private sector borrowing from the rest of the world, ∆NFBp. This must equal the change in borrowing from domestic banks by the private sector, ∆DCp. The budget constraint of the public sector can be written as: (7.2)

where Yg is income accruing to the government, Cg is government consumption expenditure, Ig is government investment, ∆NFBg is the change in net foreign borrowing by the government, ∆NGg is government borrowing from the private sector and ∆DCg is government borrowing from domestic banks. Equation (7.2) states that the difference between income and current consumption is financed by borrowing. 3 The financial constraint facing the external sector of the economy can be written as: (7.3)

Equation (7.3) is written in terms of the home currency. Z is the value of imports, INTge is payment of interest on its external debt by the government, INTpe is payment of interest on its external debt by the private sector. So these terms add up to net external payments by this economy. The next set of terms represents earnings. X represents the value of exports, NTR eg is the net transfers from abroad to the government, NTRep is the net transfers from abroad to the private sector and NFPep is net factor incomes from abroad. The difference between external earnings and external payments has to be made up by borrowings or running down international reserves. Thus ∆NFBp represents additional external borrowing by the private sector, ∆NFBg additional borrowing by the government and -∆R the running down of reserves. Since the economy in question is a developing one, it is assumed to get transfers from abroad. 4 By definition, the change in money supply (∆M) will be made up of changes in the reserves (∆R) and claims on the private and public sectors, i.e., ∆DCp and ∆DCg respectively. Thus we have: (7.4)

5

Finally the national accounts of the economy can be written as: (7.5)

This is the basic Fund model. A few points about this framework should be noted explicitly. First, the equations presented hold ex post and are, as such,


179

identities. Second, considerable importance is placed on the links between domestic credit (to public and private sectors), changes in reserves and the money supply. Third, the supply side does not figure in the model at all. Hence, inherently, it is a completely demand driven model. To emphasise this last point, the Fund in its working with the model assumes that eight key variables are exogenously determined. These are government interest on its external debt, private interest on its external debt, exports, net external transfers to the private and public sectors, net factor payments from abroad, and net borrowings of the private and public sectors. Various justifications are given for this assumption. It is assumed that both the private and public sectors must honour their external commitments so that interest payments on debts owed by them are treated as exogenous. The country in question is assumed to be a small open economy so that its export earnings are determined by world demand and are outside its control. External transfers to the private and public sectors are determined by external conditions, as are factor payments from abroad. External borrowings of the private and public sectors are assumed inflexible in the short run. But the domestic credit offtake of the public and private sectors are assumed flexible. The remaining variables of the model are then determined as follows. The change in reserves (∆R) is treated as a target variable to be determined by government policy. The level of real national income, y, is assumed fixed (underscoring the neglect of the supply side) and nominal national income, Y, is simply the product of real national income and the price level, P. (7.6)

where a bar (-) denotes a parameter. Money demand is related to nominal income through a simple fixed velocity quantity theory of money type equation: (7.7)

It is further assumed that the money market always clears so that money demand equals money supply. Substituting in equation (7.4) we have (7.8)

Equation (7.8) is important because it relates changes in international reserves to changes in the demand for money and in domestic credit. When changes in domestic credit are lower than the changes in money supply, there will be an accumulation of reserves, much like the monetarist approach to the balance of payments studied in Chapter 6. For equation (7.8) to hold, changes in nominal income must be independent of changes in domestic credit. As discussed in this model real national income is taken as fixed. The price level is determined by:

180


(7.9)

where Pd is the price of home goods and Pz the price of imports and e is the value of the exchange rate. θ (a parameter) is the share of imports in the overall price index. The price of home goods and imports are assumed exogenous. Now for small changes in the price level and in real national income, the change in nominal national income can be approximated by: (7.10)

Substituting in equation (7.8) we have: (7.11)

Thus equation (7.11) states that the change in the external reserve position is a function of the change in the price level (which is exogenous), the change in real national income (which is also exogenous and more often than not set equal to zero in the short run) and changes in credit. Thus if the external reserve position is to improve, there must be a drop in domestic credit— public or private or both. Analogously, a drop in domestic credit could lead to a drop in the rate of inflation. Thus in an economy that has a high rate of inflation and a balance of payments deficit a curtailment of domestic credit will lead to an improvement in both. This is the basic message of the FundPollack model. Equation (7.11) can be depicted as a straight line (II) in change in reserves and inflation space as in Figure 7.1. Suppose the government targets the rate of inflation (∆P*) and changes in reserves (∆R*). Equation (7.11) can then be written as: (7.12)

Figure 7.1


181

Equation (7.12) defines a limit for the credit that can be issued to be consistent with the desired inflation and reserves targets. This can be defined as an internal balance condition. For given values of the exogenous variables in (7.12) and given credit changes, a higher rate of reserve accumulation will be associated with higher monetary growth and, therefore, higher inflation. This is then a justification for the credit ceiling imposed by the IMF. One should note, however, that ∆P* and ∆R* cannot be chosen arbitrarily. Changes in the rate of inflation may impact on other variables that have been held fixed. In particular, their chosen values must satisfy the external balance constraint (7.3). In equation (7.3) we make the simplifying assumption that imports are a simple linear function of nominal national income: (7.3a)

All other magnitudes in equation (7.3) are predetermined so that this equation can be written as: (7.3b)

where χ is a constant whose value is given by the predetermined values in equation (7.3). We have to use (7.3b) to relate the rate of inflation to the change of reserves that is sustainable in the sense that it satisfies the external balance condition (7.3). To do this realise that, by definition, ∆Y Using from equation (7.10) we have (7.13)

where ß is a constant. Thus a negative relation between inflation and change in reserves is required to maintain external balance. The higher the rate of inflation, the greater will be imports. Reserves are run down faster to pay for imports. This defines the EE curve in Figure 7.1. Thus there is a unique rate of inflation and a unique rate of reserves accumulation that is consistent with both internal balance as well as external balance for given values of exogenous variables. How does policy work in this model? Suppose we are at point R where we have neither internal nor external balance. We need to re-establish external and internal balance. To do this we must change values of some of the policy variables, which we have kept exogenous so far. How do we attain internal and external balance? Policy changes must ensure that internal and external balance schedules interact at R. If there is ceiling on the expansion of domestic credit, that will shift the II curve up toward R. A devaluation of the home currency will shift both curves up so that they can intersect at R. It is important to realise that both policy initiatives are necessary. A credit ceiling will shift II and not EE. Hence this policy adjustment, by itself, will be inadequate. Similarly, only a devaluation of the currency will shift the EE curve but not the II curve so that it would not be possible to simultaneously attain internal and external balance.

182


This, then, is a rough characterisation of the approach taken by the Fund in its stabilisation programme. There are points of detail such as determining how the credit ceiling will be split up between the private and public sectors. The Fund’s approach to stabilisation then follows a programme. It has been called Financial Programming. The constituents of the programme for any planning horizon (the duration of the stabilisation programme, called the programme period) are as follows: Step 1: Specify the target rates of inflation and reserve accumulation. Step 2: Project the values of exogenous variables for the entire programme period. Step 3: Given what has been achieved in Steps 1 and 2, calculate the adequacy of exchange rate changes required to satisfy the II and EE schedules. Step 4: After arriving at the extent of the required devaluation the implied price level changes are deduced from equation (7.9) and then (7.6) can be used to get nominal income. Step 5: Equation (7.12) can then be used to calculate the implied required change in credit. Step 6: The public sector budget constraint (7.2) is then made consistent with the credit requirements known from Step 5. Several criticisms of the IMF model have appeared in the literature especially by the development community. The most significant of these criticisms has been from the structuralist school. A succinct statement of the structuralist position with respect to the IMF model is given in Taylor (1987). There are many other significant authors contributing to this position. Killick (1995) is a recent restatement of this position. The structuralist school argues that the structure of an economy is important in determining the outcome of market processes. The principal point of contention of these authors is that the IMF model overkills domestic demand by assuming that excessive domestic demand is the reason for a developing country’s balance of payments problems. The structuralist school argues that the root cause of the foreign exchange problems of developing countries lies in their structure. For instance, they need foreign exchange to buy industrial machines whereas they are unable to sell their exports (which consist essentially of primary products) in international markets at favorable terms of exchange. Reducing domestic demand, in this situation, will lead to industrial stagnation without addressing the root cause of the balance of payments difficulties. Structuralists also believe that, unlike what the IMF model assumes, inflation is not just a monetary phenomenon in developing countries. Both monetary and structuralist forces are at work. In particular, the structuralist school questions the appropriateness of the assumptions of constant velocity of circulation of money, law of one price and strong substitution responses. Any of these


183

assumptions may break down in practice leading to control of credit being an inadequate (and probably incorrect) response to high inflation. In a more formal vein, Taylor (1987) suggests that the IMF framework misses one of the crucial linkages found in developing countries:

(7.14)

Taylor (1987) believes that this is the appropriate output equation in the IMF model. This should include the costs of imported inputs like fertilisers, petroleum and energy since production in many developing countries is strongly dependent upon these inputs. Taylor also feels that the IMF model unduly emphasises the financial constraint facing the economy (equation (7.3)), the reserves equation (7.4) and the money income identity (7.7), and ignores (7.1) which brings out the fact that private saving can take the form of physical capital or financial claims on the public sector and the rest of the world. If the Fund considers (7.1) and (7.14), it would realise that a reduction in credit will lead to an increase in private savings (other things being the same). Thus aggregate demand will decline which, under imperfect price adjustment, will lead to a contractionary drop in output and thus lower demand for imported inputs. The trade balance will necessarily improve as a consequence but this improvement would have come at the cost of inducing a domestic recession. The tight monetary policy espoused by the Fund will lead to higher interest rates, which will raise the costs of borrowing for working capital. Under mark-up pricing rules found commonly in many developing countries, this will lead to higher prices. The higher inflation and lower output would amount to stagflation. The last policy measure suggested by the Fund is devaluation. This initiative also has undesirable consequences. Devaluation of the home currency would lead to a rise in the price of imported inputs in domestic currency. The higher the pass-through of these higher prices the greater the inflationary impact on domestic inflation. If money wages rise in response there will be a vicious cycle with higher wages chasing higher prices and higher prices chasing higher wages. If nominal wages do not rise, the consequent drop in real wages will lead to contractionary effects on demand and real output. A FORMAL IMF-TYPE MACRO MODEL FOR DEVELOPING COUNTRIES Some prominent economists have formalized the IMF view of the macroeconomics of developing countries. One of the earliest and still one of the most influential such models was developed by Haque, Lahiri and Montiel (1990). They considered a variant of the Mundell-Fleming model studied in

184


Chapter 5. Theirs is a two-good model. One good is produced domestically. This good is domestically consumed as well as being exported. The other good is imported. The aggregate demand equation for this model is written as: (7.15)

where Yt is real income, Ct is consumption, It is investment, Gt is government expenditure and Xt is exports. These are all measured in terms of the commodity being produced domestically. Zt is imports. is the foreign price of the imported good and et is the exchange rate whereas Pt is the domestic price level. Hence is the real value of imports in terms of the home good. This is then comparable to the values of the other components of aggregate demand. Within this IS-LM framework, the consumption function is specified as: (7.16)

where rt is the real rate of interest,1 and represents disposable real income. Disposable income for domestic residents consists of three factors: real GDP at home and earnings of domestic residents on net foreign assets. From the sum of these two we must deduct interest paid on domestic debt and taxes. Thus we can write the following expression for disposable income: (7.17)

is the foreign nominal rate of interest and it is the domestic nominal rate of interest. Fp,t-1 is the net foreign assets held by domestic residents, whereas DCp,t-1 is the stock of domestic bank credit held by the domestic private sector and Tt is real taxes paid by the domestic private sector. The private sector faces a budget constraint. It says that disposable income and consumer expenditure are linked to the net change in consumer wealth and is written as: (7.18)

Mt is the money supply in time period t. Equation (7.18) states that disposable income equals expenditure on consumption, investment and the accumulation of assets—money (M), foreign assets (FP) and bank credit (DC). The determinants of consumption have already been specified. Investment is deemed to depend upon the change in the interest rate, the change in income and on the lagged value of investment. This is written as:


185

(7.19)

Exports are modelled as a function of the real exchange rate and the level of real economic activity abroad The higher the real exchange rate the greater will be exports as home goods are cheaper. The higher is income abroad, the greater is likely to be the demand for home country exports. To capture partial adjustment a lagged endogenous variable term is included in the export equation, which is written as: (7.20)

Real imports, on the other hand, are negatively related to the real exchange rate: the higher the real exchange rate, the higher the domestic price of imports and the lower their demand. On the other hand, since imports are a normal good, the higher is domestic real national income the higher is import demand. To capture partial adjustment a lagged imports term is included in the import equation. Finally, in many developing countries, the availability of reserves often constrains the purchase of imports. Thus foreign exchange reserves as a proportion of the value of imports last period is inserted into the import equation. This import equation is then written as:

(7.21)

where Rt is the foreign exchange value of international reserves. Aggregate supply in this economy is given by a Cobb-Douglas production function, which relates capital and labour to output: This production function is written as: (7.22)

Estimation of this equation is a problem since it involves getting estimates of the aggregate capital stock, which may be hard to get. To get around this difficulty Haque, Lahiri and Montiel express current capital stock as a function of an initial capital stock and the sum of investments (after depreciation). The final form of this supply function is written in per labour terms as: (7.23)

where is the redefined capital stock term and is the intercept for the redefined equation. Sluggish adjustment is given by the lagged endogenous variable and t represents time, so that the equation incorporates a time trend. Given that equation (7.23) represents a supply function, it is being implicitly assumed that there is perfect wage-price flexibility. Hence, the aggregate demand, aggregate supply function of this model are fairly standard. Nothing

186


except the constraint that imports are constrained by the availability of reserves distinguishes this model from a mainstream macro model for a developed country. Some departures from this standard model are admitted in the money market equation discussed below. Money market The supply of money (M) in the economy consists of reserves and domestic credit: (7.24)

Reserves are determined endogenously by the balance of payments for any given money supply as will be explained below. However, domestic credit both to the private sector (DCp,t) and to the public sector (DCG,t) is determined by policy as follows: (7.25)

The demand for money is written in its usual format as: (7.26)

Having the lagged income term permits the demand for money to adjust more slowly to changes in income than in the interest rate. The lagged value of real cash balances permits partial adjustment type responses in the demand for money. Now, if the economy is fully integrated into the world economy then uncovered interest parity would hold and the domestic interest rate would equal the foreign interest rate plus expected depreciation of the home currency. In a completely closed economy domestic nominal interest rates would not be related to world interest rates. Haque, Lahiri and Montiel permit both these factors to influence domestic interest rates as in: (7.27)

where Et is the expectations operator. When θ=1 we have uncovered interest parity and when θ= 0 we have a closed economy. As θ increases, so does the degree of capital mobility. Now let denote the money supply in this economy had this economy been closed, i.e., had full capital controls. The actual money stock differs from this to the extent of private capital flows. Let denote the interest rate that would be consistent with M′. The value of can be obtained by solving the following equation:


187

(7.28)

With capital controls in this economy, an independent monetary policy is pursued as follows: for a desired level of the domestic interest rate i′ we get the value of the money supply consistent with this by substituting this into (7.26) and solving for Given this, foreign exchange reserves are determined from (7.24). Subtracting last year’s reserves we get the change in reserves, ∆Rt. From the balance of payments equation: (7.29)

CAt is the current account surplus and is written as (7.30)

In this expression, the first two represent the balance of trade. Fp,t-1 represents net foreign assets of the private sector, FG-1 net foreign assets of the government sector and Rt-1 represents net foreign reserves last period. Hence the expression inside the brackets in equation (7.26) represents the stock of net foreign assets of the home economy. Multiplying it by the foreign interest rate (it*) gives the flow of yield (in foreign currency) from this asset and multiplying it by the exchange rate gives the value of this flow in terms of the home currency. The government can choose either the interest rate or the net private capital flows from abroad. Once one is chosen, the other becomes endogenous and equation (7.27) drops out of the model. To see this, suppose the authorities choose it. They can then solve for the ∆Fp,t required to support it. They would then allow that much capital mobility. Hence (7.27) is not necessary. This would be true if the degree of capital mobility could be chosen. Alternatively, we could suppose that the degree of capital mobility is determined by convention. In this situation both it and ∆Fp,t become endogenous. Government sector The government acquires foreign assets (FG,t) and domestic assets (DG,t). It gets revenue from taxes (Tt) and from interest on foreign assets. Its expenditures are Gt and comprise government consumption and interest payments on the domestic debt. Thus the government budget constraint is written as: (7.31)

Haque, Lahiri and Montiel thus consider the following system of equations as comprising their macro model: 1 The consumption function (7.16)

188

2


The investment function (7.19)

3

The export function (7.20)

4

The import function (7.21)

5

The supply function of output (7.23)

6

The money market equilibrium condition (7.26)

7

The interest rate equation (7.27)

They estimate this set of equations for 31 developing countries. The countries studied include Brazil, Chile, Colombia, Costa Rica, Ecuador, Egypt, Ethiopia, Greece, Guatemala, India, Indonesia, Jamaica, Jordan, Kenya, South Korea, Malawi, Malaysia, Malta, Mexico, Morocco, Nigeria, Paraguay, Philippines, South Africa, Sri Lanka, Tanzania, Thailand, Tunisia, Turkey, Venezuela and Zambia. Given the extremely wide variety of macroeconomic experiences across these countries, the authors would appear to consider this model to have wide, almost universal applicability. EVALUATION OF THE HAQUE-LAHIRI-MONTIEL MODEL This model is very much inspired by the IMF view of macroeconomics. Neither aggregate demand nor aggregate supply is radically different from the standard macro model for developed countries. Departures from this are admitted only in the financial markets. From this would follow the prescription that stabilisation policies that intervene primarily in the financial sector will have limited output effects. Since wages and prices are perfectly flexible, employment effects of such stabilisation policies are non-existent or, at best, transitory.


189

Hence the models studied in this chapter would argue that stabilisation policies concentrating on reducing aggregate demand would have limited, if any, impact on employment and output—at least in the long run.2 Similarly, financial sector reforms would have little impact on the real sector. Hence countries in balance of payments difficulties would be able to pursue stabilisation policies without endangering output growth.3 This, then, provides a rationale for the kind of intervention postulated by the IMF. In subsequent chapters we will encounter criticisms of this approach from the structuralist and other schools. In the box below I outline some objections to IMF-World Bank-type policies raised by the Nobel Laureate Joseph Stiglitz.

The IMF-World Bank policies for developing countries: Stiglitz’s critique Professor Joseph Stiglitz of Columbia University won the Nobel Prize for economic science in 2001. He is ex-chief economist of the World Bank and former Chairman of the Council of Economic Advisers to US President Bill Clinton. Since leaving the Bank, Prof. Stiglitz has been highly critical of the policy advice given by the IMF and the World Bank to developing countries. In an interview published in the British newspaper The Guardian, Prof. Stiglitz outlined four principles of his critique of these institutions. 1

2

Stiglitz is opposed to the poorly thought-out privatisation programme put forward by the World Bank and the IMF. Although Stiglitz cited several such cases, the most striking one was the biggest—the 1995 Russian sell-off. It is well known that much of the proceeds from the big sale of states’ assets flowed out of the country. After privatisation, Step Two is capital market liberalisation— particularly capital account convertibility of the home currency, although in recent times the IMF has stopped insisting on this. Capital account convertibility allows investment capital to flow in and out. Unfortunately, as in Indonesia and Brazil, the money often simply flows out. Stiglitz calls this the ‘hot money’ cycle. Cash comes in for speculation in real estate and currency, then flees at the first whiff of trouble. A nation’s reserves can drain in days. And when that happens, to induce speculators into returning a nation’s own capital funds, the IMF demands these nations raise interest rates to the dizzying heights of 30, 50 and even 80 per cent. ‘The result was predictable,’ said Stiglitz. Higher interest rates lead to sharply lower property values as well as savaging industrial production and draining national treasuries.

190

3

4


At this point, Step Three of the IMF-World Bank prescription comes into effect. This involves a lot of measures that restrain fiscal expenditures and deficits. Subsidies are lowered, public expenditure and bank credit are curtailed (as per the models discussed in this chapter) and user cost pricing in respect of utilities and petroleum resources and electricity. Thus as Stiglitz noted when the IMF eliminated food and fuel subsidies for the poor in Indonesia in 1998, the country exploded into riots. There are other examples— the Bolivian riots over water prices and the riots in Ecuador over the rise in cooking gas prices imposed by the World Bank. There is evidence to suggest that the IMF anticipates such social unrest but persists with these price adjustments nevertheless. These riots have often been followed by capital flight. Step Four involves the adoption of free trade. This is free trade by the rules of the World Trade Organisation and the World Bank, which Stiglitz likens to the Opium Wars. ‘That too was about “opening markets”,’ he said. As in the nineteenth century, Europeans and Americans today are kicking down barriers to sales in Asia, Latin America and Africa while barricading their own markets against the Third World’s agriculture. In the Opium Wars, the West used military blockades. Today, the IMF and the World Bank can order a financial blockade, which is just as effective and sometimes just as deadly. Stiglitz has two concerns about the IMF/ World Bank plans. First, he says, because the plans are devised in secrecy and driven by an absolutist ideology, never open for discourse or dissent, they ‘undermine democracy’. Second, they don’t work. Under the guiding hand of IMF structural ‘assistance’ Africa’s income dropped by 23 per cent. Did any African nation avoid this fate? ‘Yes,’ said Stiglitz, ‘Botswana.’ Their trick? ‘They rejected the IMF medicine.’


2

Using the IMF website www.imf.org make a list of countries that opted for IMF stabilisation packages in the ten years up to 1995. Trace the subsequent path of key macroeconomic aggregates of these countries after 1995. Make a list of policy changes. How far and in what cases would you agree with Stiglitz’s critique of IMF policies? Examine the balance sheets of a typical commercial bank in a less developed country—say India—and that in a rich country—the UK. What differences do you find in the asset and liability sides of the balance sheets in the two cases? Comment on the relative importance of public sector bonds in the two cases. Suppose a similar stabilisation


191

policy —consisting of lowering of government expenditure—is being followed in both countries. Do you think there will be differential impacts on the balance sheets of the two banks? Why? NOTES 1 2 3

The real interest rate is the nominal interest rate less the expected rate of inflation. Strictly speaking, this argument obtains even for the short run. To improve long-term responses the IMF-World Bank prescription is to improve economic efficiency through reduction in government expenditure, price and tax reforms and privatisation of public enterprises.

8


INTRODUCTION Macro models specifically for developing countries are not very common. This is surprising given the wide variety of experiences of many developing countries with respect to macroeconomic performances. Perhaps the most important factor behind this is the poor quality of data in many developing countries. This poor data quality leads to difficulties in empirically verifying these models. In this chapter we will look at several variants of these models— some more representative of some groups of countries than others. For instance, supply responses will vary across groups of countries. Developing countries with reasonably well developed financial models such as Thailand will have different supply responses than many African countries where financial markets are rather weak or do not exist at all. STRUCTURALIST MODELS OF DEVELOPING COUNTRY MACROECONOMICS Many eminent macroeconomists would disagree with the Haque-LahiriMontiel view of developing country macroeconomics presented in Chapter 7. In particular, they would disagree with the modelling of aggregate supply and demand. Also they would question the almost universal applicability of the model that Haque-Lahiri-Montiel profess. In addition, since the collapse of the Bretton Woods Agreement in 1971 and the subsequent oil price shocks, and the recent currency crises and contagion experienced by several Latin American and East Asian developing countries, several developing countries have had to take recourse to the International Monetary Fund (IMF) to bail them out of their balance of payments difficulties.1 The IMF imposed some conditionalities on its loans (discussed within the context of the IMF’s monetary approach to the balance of payments in Chapter 7), which took the form of short-term stabilisation measures. The IMF conditions were viewed and continue to be viewed as

194


too harsh, but a cogent intellectual response could be fashioned only if an alternative view of the short-run macro performance of developing countries was available and agreed upon. These IMF conditions had the beneficial effect of giving rise to a whole new school of macroeconomic thinking about the problems of developing countries. Some of this alternative approach drew its inspiration, if not its technique of analysis, from an older and somewhat unorthodox view of the economic problems of developing countries that went under the general rubric of structuralism. In this chapter we shall study a model that outlines the basic contours of the standard model of structuralism. In the next chapter we will focus on some specific models that emphasise the heterogeneity of the developing country macroeconomic experience. The building blocks of this structuralist model are the same as those of the mainstream macro models: aggregate supply and aggregate demand. However, the specifications of aggregate demand and supply are markedly different. We now proceed to a discussion of these. Aggregate supply Aggregate supply in an economy depends, in general, upon product and labour market demand and supply function and the behaviour of entrepreneurs and workers. In each of these cases developing countries are perceived to be different from developed countries. In the aggregate supply schedule studied in Chapter 5 we had assumed that product markets are competitive. The Haque-Lahiri-Montiel model also shares this assumption. The structuralist view disagrees with this. It argues that in LDCs product markets are generally characterised by oligopoly with markup pricing. This is because the size of demand for industrial non-tradable goods is rather small. For firms producing tradable goods the government usually provides tariff protection and other incentives, typically subsidised credit and other forms of protection from foreign competition. In addition, wages in the labour markets are too high and do not reflect the true scarcity of labour. A variety of explanations can be offered for this. First, labour unions often emerge in protected manufacturing sectors at a relatively early date as labour unions follow western patterns. The ruling elite in many developing countries finds it profitable to co-opt a section of the labour force to counterbalance the favour shown to domestic capitalists and for political gains. Correspondingly, the growth of labour unions is tolerated—even encouraged. Furthermore, salaries for civil servants also tend to get inflated. Multinational corporations in developing countries mimic wage structures in their home countries. All this leads to a situation of persistent excess supply of labour. In the short run, anyway, a considerable part of the labour force is unemployed, underemployed or employed at very low wages in the ‘informal’ sector. In the face of all this the money wage for regular employment is fixed.


195

Another important distinguishing characteristic of production in developing countries is that imported intermediate inputs are extremely important in industrial production. They, along with labour, are the major constituents of variable costs. It is often assumed that labour’s marginal product is diminishing. However, it is not uncommon to find unused capacity in the industrial sector. Moreover, cash borrowing largely finance working capital. With capital markets that are not well developed, equity investment is relatively unimportant. Correspondingly, the rate of interest is an important determinant of the cost of working capital and, hence, of aggregate supply. Since imported intermediate inputs are an important element of cost, so is the exchange rate. As the price of foreign exchange goes up, ceteris paribus, we would expect the price of imported inputs to go up and hence supply to fall. We can then write the aggregate supply function as (8.1)

where Y is real output, P is the price level, i is the interest rate, w is the nominal wage rate and e is the exchange rate. In the PY plane the aggregate supply curve (AS) is upward sloping and an increase in i, w or e would shift the aggregate supply schedule left. Furthermore, since excess supply exists in industry we would expect the AS schedule to be relatively flat, not steep. To be more precise, there are several goods in the model. First, there are intermediate inputs, which are not produced domestically and need to be considered. Second, there is the domestic final good, which is consumed as well as invested. Third, there are exports. We shall club the last two goods into one category and study the demand and supply for it. Our analysis follows, in the main, the work of Taylor (1987) and Porter and Ranney (1982). Aggregate demand Aggregate demand in developing countries is also affected by their institutional characteristics. In the first instance, financial intermediation and, sometimes, even the use of money may not be very widespread in many developing countries. As one would expect, economists have developed indicators of financial underdevelopment in developing countries. This underdevelopment is reflected in the low value of the ratio of financial assets to national income. Similarly, the number of banks is low and unduly concentrated in urban areas. Furthermore, bank interest rates are often controlled by the government—a policy which has been called ‘financial repression’ by McKinnon (1973). A large part of investment is self-financed. This has two significant implications for monetary policy. First, large parts of the economy are outside the purview of the central bank. Second, open

196


Table 8.1 Balance sheets for four agents of a typical developing country

market operations are not very important. The central and commercial banks hold government debt at artificially low interest rates (and therefore the prices of these bonds are artificially high). These bonds are only occasionally traded in the bond market—which is rather thin to begin with. In Table 8.1 we detail the prototype of accounts for four major economic agents in a typical developing country. We distinguish between four different types of economic agents: the central bank, the commercial banks, firms and households. We neglect currency holdings since no great insight is obtained by including it. We make the realistic assumption that domestic residents do not hold foreign exchange. This assumption, however, may not always be satisfied. From the table we have (8.2) (8.3) (8.4) (8.5)

Thus equation (8.2) states that the assets of the central bank (reserves, R) equal the sum of current and past government deficits (Σ∆), against which the government has borrowed, and the sum of past balance of payments surpluses, (Σß). The latter is almost always a deficit so that this enters with a negative sign. The liability of the central bank corresponding to these two


197

assets is the sum of the reserves held by commercial banks with the central bank. Equation (8.3) describes the balance sheet of commercial banks. The assets of commercial banks include reserves held with the central bank and the short-term working capital loans to firms. The corresponding liabilities of central banks include the deposits of firms and households. Equation (8.4) denotes the financial balance sheet of firms. Firms’ assets include their deposits with commercial banks and the value of their capital stock. Their corresponding liabilities include borrowings from households and commercial banks as well as their net worth (which ultimately is an asset to households). Finally, we have the household’s balance sheet where the assets are deposits with banks and loans to firms as well as the net worth of firms; the corresponding liability is their net worth. The monetary base is determined completely by the sum of past government budget deficit and the sum of past government budget deficits and the sum of past balance of payments surpluses. In any time period (8.6) (8.7)

where Ng is the size of the labour force employed by the government, Ig is the amount of investment undertaken by the government, T is tax revenue, X are exports and M are imports. We have normalised world price to equal one. Equation (8.6) simply defines the government’s budget deficits as equal to its expenditure on wages and investment less its tax revenue. Equation (8.7) defines the balance of payments surplus as the difference between the home currency values of exports and imports. Any change that increases the government budget deficit or the balance of payments surplus will cause a direct and cumulating increase in the monetary base. The central bank imposes a reserve requirement ratio s on the commercial banks: (8.8)

Using (8.2), (8.3) and (8.8) the supply of bank loans is determined as (8.9)

Any increase in the monetary base or decrease in the reserve requirement ratio increases Lb. The demand for real balances is modelled as a function of real income Y and the interest rate i. For simplicity we ignore the rates of return on capital and firms’ equity as possible determinants of the households’ demand for money. The demand for real balances by households is written as Dh/P where (8.10)

198


Firms demand money in order to finance their variable costs during the period of production. These costs consist primarily of labour and imported raw materials, and hence (8.11)

where w is the nominal wage rate, N is the level of employment and m is the marginal propensity to import. Some empirical studies have noted that the elasticity of money demand by firms with respect to cost is near unity, whereas the elasticity with respect to the interest rate is near zero. Substituting appropriately and eliminating the financial variable we derive a single equation that defines the financial equilibrium of the economy: (8.12)

This is the LM equation for this economy. It says that money demand by firms and households must equal money supply It has three endogenous variables: the price level P, national output Y and the interest rate i. This interest rate is to be distinguished from the fixed bank lending rate ib. The past values of the budget deficit and balance of payments deficits are unalterable; however, the current values of these variables can be influenced. The two obvious policy variables in this equation are the exchange rate e and the reserve requirement ratio σ. Holding all other ingredients constant, the slope of the LM schedule is

Hence the LM curve is upward sloping in the (i, Y) plane. This is the same as in developed countries but in LDCs the LM schedule is definitely steeper, ceteris paribus. Several reasons can be advanced for this. The ratio of money to gross national product is smaller in LDCs. This may be the case because money is largely held for transactions purposes: An increase in the price level will lower the real supply of money and shift the LM curve upward and to the left. An increase in e, σ or w will have the same effect. An increase in ∆ or ß will imply an increase in the monetary base and the LM curve will shift to the right. Goods market equilibrium As before, goods market equilibrium occurs when aggregate goods demand equals aggregate supply. Aggregate goods demand has the usual components. Real consumption expenditure depends upon real disposable income.


199

Moreover, the functional distribution of income is important in LDCs. It is often argued that the marginal propensity to consume out of wage income is lower than that out of capital income. We further introduce an aggregate (proportional) income tax rate into the consumption function and write it as (8.13)

where S1 is the share of labour in total income and τy is the proportional income tax rate. Consumption rises when real income rises and drops when the share of labour in total income or the income tax rate rises. Let us now look at investment demand. Investment is assumed to depend positively on the desired capital stock, which, in turn, depends on output. Investment funds are borrowed from official sources (commercial banks and the like) at the fixed interest rate ib. However, because this interest rate is below the equilibrium interest rate, there will be an excess demand for investment that spills over to the curb market where the interest rate, i, is flexible. In a functional form for investment it will be convenient to express investment as a function of the average of these interest rates. We know that this average interest rate will also be a function of the availability of loans (Lb) at the official rate. We know that Lb increases as total reserves increases. Hence we can write the private investment function as (8.14)

As output goes up, so does the optimal capital stock and investment; as i or ib rise, investment drops for the standard reasons; as total reserves rise, the amount of loans available from official sources rises and hence investment rises. When the reserve requirement rises, the amount of funds available for investment drops. When the real wage or the real exchange rate rise, the cost of financing inputs goes up and investment drops. It is to be noted that changes in the official bank rate, ib, affect the commodity market equilibrium and, hence, the IS schedule but not the LM schedule. Government spending in developing countries is composed largely of wage payments and the wage paid to government employees is kept in line with those in the private sector—or vice versa. Any decrease in the real wage bill would decrease government expenditure. Another major component of government expenditure is public investment. We shall assume that this is fixed in the short run in real terms. We may, therefore, write government expenditure G as (8.15)

The signs of the partial derivatives are as noted above.

200


We may now investigate the determinants of net exports. For most developing countries the assumption of a small open economy is appropriate. Hence the domestic price of tradables depends on world prices, subsidies, tariffs and exchange rate. Because of capacity and other constraints the price elasticity of supply of exports is likely to be small in the short run. In most developing countries imports and domestic goods are very poor substitutes. For LDCs that have embarked on a course of import substitution and have completed most of it, import consists largely of non-competitive imported inputs. We suppose that imports are a fixed proportion of domestic output. It should also be pointed out that many LDC governments derive substantial revenues from ad valorem taxes on imports (τm). We shall also model a tax on exports (τx)—which may actually be a subsidy Further, the trade sector is a very important part of national accounts and many significant macroeconomic problems of developing countries have their genesis in a large deficit in their current accounts. We write net exports (NX) as (8.16)

where X is the value of exports in foreign currency. We can write the goods market equilibrium condition as

(8.17)

The equilibrium relation between i and Y from the above equation will be the IS schedule. This schedule will have a negative slope in the (i, Y) plane for fixed values of other variables. This IS schedule is likely to be quite steep. A major reason for this is the low interest elasticity of investment demand. However, as Leff and Sato (1980) argue, the IS curve may not be too steep on two counts: first, the large value of marginal propensity to consume in developing countries and, second, the high income elasticity of investment demand. An increase in government expenditure will shift the IS curve outwards. An increase in τx or τm will reduce the value of net exports in domestic currency for a given foreign currency balance, thus shifting the IS curve inwards. This is concentrated to some extent if export supply or import demand are price elastic. When there is a devaluation of the real exchange rate (e/P) the terms of trade are adversely affected and investment declines because the cost of


201

intermediate inputs rises. From our study of the Marshall-Lerner condition in Chapter 6 we know that if B is the initial trade balance in domestic currency, ηxs is the export supply elasticity, εMD is the (absolute value) of the import demand elasticity and Z is imports in the foreign currency, the balance of trade will worsen (ignoring taxes) if (8.18)

For developing countries it would seen that both ηXS and εZD are small. Furthermore, in times of severe balance of payments problems, B may be a very large negative number. Hence, under these conditions, it would seem that devaluation will reduce aggregate demand and shift the IS curve inwards (see Krugman and Taylor (1978)). Finally a reduction in the real wage rate (w/P) will shift the IS curve down and to the left through its effect on the wage share S1, but up and to the right through the effect on retained earnings and hence fixed private investment. In most LDCs we would expect the former effect to dominate. The price level is the third endogenous variable in the goods market equilibrium equation. An increase in the price level implies both a revaluation of the real exchange rate (a decrease in e/P) and a fall in the real wage (a decrease in w/P). The net effect would, therefore, be ambiguous. We shall assume that this effect is zero. The aggregate demand schedule The LM schedule and the IS schedule can be combined in the usual way to remove one of the three endogenous variables. We choose to remove the curb market rate i. Since ib is a parameter we are left with a relation between P and Y that can be written as (with Yd denoting output demanded): (8.19)

With all other factors held constant, this AD schedule is downwards sloping in the (P, Y) plane. However, this AD schedule is likely to be much steeper than the AD schedule derived in Chapter 3. There are two reasons for this. First, following a price rise, the IS schedule shifts to the left in the standard model but does not do so in developing countries. Furthermore, the IS curve is relatively steep due to high marginal propensity to import and a low interest elasticity of investment demand. It is also noteworthy that devaluation reduces aggregate demand in the case of developing countries whereas, students will recall, devaluation increases aggregate demand in developing countries. Further, the nominal wage enters aggregate demand with an uncertain sign.

202


Aggregate supply We now move on to consider aggregate supply. In general it is possible to write down the aggregate supply function as (8.20)

Furthermore, removing variables that are inessential for the time being, we can write the aggregate demand function as (8.21)

The sign of the partial derivatives below the Yd equation have already been discussed. So far as aggregate supply is concerned a rise in the price level will evoke the normal supply response, an increase in the interest rate will make production more costly and reduce supply, an increase in the value of foreign exchange will make production more costly and, hence, reduce supply, and so will an increase in the nominal wage. Both will negatively affect output. It ought to be pointed out that, as before, any change in the IS-LM equilibrium and, therefore, the curb market interest rate will be a function of the price level (changes in which induce shifts for the LM schedule) and exogenous variables of the model. Hence, it is possible to write down the curb market interest rate as a function of the price level and the exogenous variable: (8.22)

The signs of the partial derivatives of equation (8.22) are those appropriate to the LDC model discussed above. Since the IS curve does not shift, there is an unambiguous positive sign below P. As P rises, the supply of market effects. Similarly as s rises, the money supply drops and the interest money falls and the interest rate σ rises. This effect is not altered by any goods rate rises. The rationale for the other effects is also straightforward. Using equation (8.22) we eliminate the interest rate from the aggregate supply, aggregate demand framework and write them as functions of two endogenous variables, P and Y. Substituting for i in the aggregate supply equation we have (8.23)

The slope of the aggregate supply function represented by equation (8.23) will be steeper than that in (8.20). This is because the effect of the price level


203

through interest rate changes is captured in equation (8.23) but ignored in equation (8.20). Moreover, changes in exogenous variables may affect the monetary base. An increase in P increases both nominal government spending (to the extent that real government expenditures are fixed) and tax revenues (to the extent that there are direct income-related taxes. We assume that the net effect is negligible. An increase in Y increases imports, reducing the balance of payments surplus and increasing import duty revenues, which in turn decrease the budget deficit. Hence the monetary base decreases. This decrease shifts both the aggregate demand and the aggregate supply to the left so that output falls. We now examine the effects of various policy shocks. Monetary contraction The impact of monetary contraction, of the type recommended by the IMF as part of its stabilisation programme, is shown in Figure 8.1. The initial position of various schedules is denoted by the subscript 0 and the final position by the subscript 1. An increase in the reserve requirement ratio s, for example, will reduce the money supply and will shift the LM curve up to the left and the IS curve to the left and down. Because of the low interest elasticity of money demand the shift in the LM schedule is greater and the interest rate rises to i1 and output falls to Y1. We draw the old and new Yd schedules in Figure 8.1(b). Aggregate supply and aggregate demand will shift inwards. Because of underutilisation of capacity the output supply function is relatively flat and the price level registers a small rise. Had the output supply function not shifted, output would have fallen to Y1. Because the rise of the price level there is a second round fall in Y. In part (a) the LM schedule shifts left again with the IS schedule staying stationary—remember that it has been assumed that the IS schedule does not respond to changes in the price level. Output falls to Y2. In the case of developed economies the price level falls in response to a monetary contraction. In the case of LDCs, however, the price level may increase. It ought to be stressed, however, that the cumulative effects of the increase in reserve ratio requirement will reverse the drop in the money supply and output will tend to revert back to Y0. Restrictive fiscal policy Let us now look at the effects of a restrictive fiscal policy. We analyse this in Figure 8.2. A decrease in government spending shifts down the IS schedule in Figure 8.2(a) from IS0 to IS1. This results in output falling to Y1 and, in part (b), the AD schedule shifting down from to Output would have fallen to Y1 had the aggregate supply schedule been horizontal. However, a fall in government expenditure means that the curb market interest rate also

204


Figure 8.1

falls. This stimulates aggregate supply and the aggregate supply schedule shifts out from to Output starts to rise and the price level starts to fall. We may come to an equilibrium income like Y2 with the LM schedule shifting outwards because of the price fall. (Remember, the price fall does not affect the IS schedule.) The contractionary fiscal policy can turn out to be expansionary. Hence fiscal policy becomes a very important policy tool in LDCs, at least in the short run. The reduction in government spending increases the government surplus and leaves the balance of trade essentially unchanged. The cumulative effect


205

Figure 8.2

is a reduction in the monetary base. Thus, over time, recession appears or worsens. Devaluation The difference between developed and developing countries is most evident when we consider the effects of a devaluation. We have already considered the effects of devaluation in the case of developed countries. As a matter of fact we analysed several different approaches to devaluation. The general result from this was that, except in some extreme cases, a devaluation can

206


Figure 8.3

help improve the balance of payments position of developed countries running a deficit. For the case of developing countries we know that the IMF has almost always insisted on devaluation as a condition for extending loans to deficit developing countries. In Chapter 11 we discuss, in greater detail, the effects of IMF stabilisation packages. In this section we want to analyse the effects of devaluation within the context of the model studied in this chapter. Let us examine these effects in Figure 8.3. For a given price level, devaluation shifts both the IS and LM curves to the left (from IS0 and LM0 to IS1 and LM1, respectively). Aggregate demand falls to But the effect on i


207

is ambiguous (in Figure 8.3(a) we show no change). Even when the interest rate is constant the increase in the price of foreign goods following a devaluation shifts the aggregate supply schedule to Output falls to Y2 ultimately with the price level going up or, perhaps, staying constant. Although the fall in output suggests an expansionary cumulative effect, devaluation has the direct impact of reducing the monetary base. Tax revenues from trade increase in terms of home currency, while balance of payments deficit increases. The net effect is ambiguous. Wage restraint An incomes policy that calls for cuts in workers’ nominal wages affects the model in a variety of ways. First, if we assume that the decrease in government spending and the redistribution of income away from workers have a large impact relative to the increased investment resulting from increased retained earnings, the IS schedule will shift inwards to the left. Second, since there is less demand for financing working capital, the demand for money fall and the LM schedule shifts to the right. Because wage and interest costs fall, supply conditions improve and the aggregate supply schedule shifts to the right. The analysis of these effects is carried out in Figure 8.4. In part (a) the leftward shift of the IS schedule from IS0 to IS1 dominates the rightward shift of the LM schedule (from LM0 to LM1). In part (b) aggregate demand falls from to Had the supply curve been horizontal, output would have fallen to Y1 and the price level would have stayed at P0. However, not only is the supply curve upward sloping but it also shifts during the process of adjustment. Output rises, somewhat, to Y2 and the price falls so that the LM schedule shifts again to intersect IS at Y2 (the intersection of the dashed lines in Figure 8.4(b)). Thus a wage restraint reduces the price level but may cause a drop in output. CONCLUDING COMMENTS From the analysis in this chapter it is clear that the effects of macroeconomic stabilisation policies are model specific. In the immediately preceding chapter we discovered that in a model of the Haque-Lahiri-Montiel variety the effects of stabilisation policies appear standard. A model with some ‘special’ features of developing countries as discussed in this chapter, however, predicts that such stabilisation policies may not be unambiguously beneficial. Most importantly, contractionary policy may fuel rather than dampen inflationary tendencies. It seems to be difficult to get rid of balance of payments difficulties except through severe output contraction to reduce import demand. Devaluation is not able to help very much. Second, all macro policies have effects on the government budget and balance of payments deficits. This severely restricts the government’s ability to conduct stabilisation

208


Figure 8.4

programmes. Third, most restrictive policies are likely to adversely affect the distribution of income. The standard IMF recipe for the balance of payments problems of developing countries is, according to this point of view, counterproductive. According to this school of thought, then, developing countries face a formidable task in stabilising output and the price level. DISCUSSION QUESTIONS 1

How would you empirically verify the model sketched in this chapter? Write down the estimable equations and discuss how you would conduct verification tests.


2

209

Some economists argue that sometimes the objectives of policy makers are at variance with those of the public. In such cases, politicians often opt for a solution that will ensure their re-election. This was put forward as the political business cycle hypothesis by William Nordhaus (1975). Nordhaus discusses this in the context of the inflation-unemployment tradeoff. How would you alter the model in this chapter to admit the possibility of opportunistic politicians?

NOTE 1

These balance of payments difficulties are often compounded by the presence of a banking crisis.

9


INTRODUCTION In Chapter 8 we studied a macro model with some salient features of developing countries. We concentrated on the role of imported intermediate inputs and imperfect capital markets in the determination of aggregate supply. In that chapter as well as the earlier chapters, particular emphasis was placed on the role of the foreign sector in determining output and inflation. In this chapter we wish to introduce some other complications particular to LDCs. In particular we wish to concentrate on the dualistic nature of many LDC economies—a characteristic that makes an aggregative model like that of Chapter 8 somewhat untenable. It is possible to argue that there is a fundamental schism between the traditional agricultural sector and the relatively modern industrial sector. There are several differences between the two sectors. For instance, the agricultural sector is operated by peasantowners, relies mainly on family labour and operates relatively outdated technology. On the other hand, in the industrial sector production is carried out in capitalist profit-seeking enterprises using relatively modern technology and hiring labour and other inputs in impersonal factor markets. There are two important links between the two sectors: food is the most important item of consumption of industrial workers and industrial goods are needed for investment in the agricultural sector. It must be admitted, however, that the differences between the two sectors do tend to narrow down with the development of the economy. In order to focus on the problems associated with dualism we assume that the economy we are studying is closed. The models studied in this chapter are those of closed dual economies rather than the aggregative open economy version studied in Chapter 8. The dualistic view of LDC economies has a long and distinguished history. For a recent survey of this vast literature see Ranis (1989). This literature has, for the most part, concentrated on real (non-monetary) issues and on long-run growth. Recently this model has been adapted to address macroeconomic questions (see, for instance, van Wijnbergen (1983), Edwards and van Wijnbergen (1989) and Taylor (1987, 1988)).

212


In this chapter we will study two simplified versions of these models. First, we emphasise the role of differences in adjustment mechanism between the two sectors. The agricultural sector is modelled as being ‘traditional’. In other words, factors such as the amount of rainfall during the crop cycle are a more important determinant of agricultural output than inputs of capital and labour. Hence, the output of the agricultural sector is taken as fixed. The market for agricultural goods clears through price adjustment. The industrial sector of this economy is nascent. As Taylor (1987) argues, in the context of LDCs, this has two implications. First, in the labour market, labour presses for certain wage demands and, second, the market for industrial goods is characterised by markup pricing. In other words, both labour and output markets are characterised by imperfect competition. Money and other financial assets enter the model through wealth effects. This model is being discussed as a prototype of a low saving-low growth LDC. Hence we focus attention on short-run equilibrium rather than long-run growth. The second version of the dual economy macro model is based primarily on the work of Taylor (1983) and Rattso (1989). We consider questions of adjustment within a broader perspective with Keynesian and neoclassical adjustment mechanism considered seriatim. We also examine the implications of differences in the propensities to save of different income classes. We explicitly model savings and investment in this section although monetary issues are eschewed. It would be only slightly oversimplifying matters if the matter of Chapter 8 is termed ‘structuralist to Latin American countries’; the first model in this chapter can then be said to be representative of the relatively less developed economies of Africa where agriculture has been stagnant, though important, and where the nascent manufacturing sector is characterised by high degrees of imperfection in both factor and product markets. The second model in this chapter is applicable to countries where the industrial sector has taken firm roots although the economy is still essentially dual. Rattso and others have estimated variants of this last model for India—but clearly the model has wider applicability. Both models in this chapter assume that the factor of production constraining output is capital, not labour. A SHORT-RUN DUALISTIC MACRO MODEL WITH DIFFERENCES IN ADJUSTMENT Consider an economy with two sectors: agriculture (A) and manufacturing (M). Agricultural output (sometimes called food) is determined by supply conditions in sector A, which, as we have discussed above, have several nonmarket characteristics. We make the simplifying assumption that agricultural The price of output is exogenously determined and denote this level as food is determined in competitive markets. We assume that this price is flexible.


213

We denote by XM, PA, PM and PR the output of the manufacturing sector, the price of food, the price of manufactured goods and the relative price of sector A to sector M goods, respectively: is total income in terms of sector M goods, α is the fraction of income spent on sector M goods so that is the fraction spent on sector A goods Since the share of income spent on either good is fixed, it follows that we are assuming unitary price elasticity of demand for both goods. Let γ be the marginal propensity to consume. The market for agricultural goods clears when supply equals demand, i.e. when (9.1)

The left-hand side of equation (9.1) is the value of agricultural supply, and the right-hand side is the value of agricultural demand, both in terms of sector M output. From equation (9.1) we can calculate the equilibrium value of PR as (9.2)

Clearly this has a meaningful solution only when In Figure 9.1 we depict equation (9.2) as the line AA. If agricultural price moves sluggishly the economy may not be on the line AA. In Figure 9.1 MM denotes the equilibrium relation between PR and XM in the market for manufactured goods. We shall assume that this market clears through quantity rather than price adjustments. The market clearing level of sector M output is

Figure 9.1

214


(9.3)

where is exogenous government expenditure in terms of sector M output and the first term on the right-hand side represents private consumption. Thus equation (9.2) states that the output of the manufacturing sector must equal private demand plus demand from the government. From equation (9.3) we can write (9.4)

Equation (9.4) is plotted in Figure 9.1 as the MM schedule. If it follows that the MM schedule is steeper than the AA schedule, is a Keynesian multiplier type relation. We shall assume that this multiplier relation is always satisfied so that the economy is always on the line MM. The market for industrial goods is characterised by imperfect competition so that PM is a markup (assumed fixed) over unit labour cost: (9.5)

where ϕ is the markup factor, ß is the amount of labour required to produce one unit of industrial output and W is the nominal wage rate in sector M. Capital is suppressed from the production relation of the M sector. The labour market is also characterised by imperfect competition. Workers are assumed to have a target real wage ω (denoted in units of both goods). Let us define the consumer price index, Pc, as (9.6)

so that the target nominal wage Wτ can be defined as (9.7)

We admit a degree of money illusion in the model by postulating that the actual wage adjusts slowly to the gap between the actual wage W and the target wage Wτ. When there is no inflation of wages or prices. We can solve for non-inflationary relative price by substituting from equation (9.6) into equation (9.5): (9.8)

is positively related to the markup factor, the labour-output ratio and the desired real wage. This is denoted as the horizontal line NIRP (noninflationary relative price) in Figure 9.1. Our assumption about the movement of nominal wages implies that (9.9)


215

where and a dot above a variable denotes a time derivative. Further, since the markup factor is constant it must be the case that the rate of inflation of PM is the same as that of W. In other words, (9.10)

Finally we postulate that the relative price PR adjusts as per excess demand for food: (9.11)

where and the expression inside the square brackets is the excess demand for food. From equation (9.2) it can be seen that, if there is no excess demand for food, i.e. if we are on line AA, then If PR=0. is above AA, PR must be falling. Equations (9.9), (9.10) and (9.11) fully describe the dynamics of inflation in this economy. Stability By assumption the market for industrial goods always clears. We substitute for XM from equation (9.4) into equation (9.11) and rearrange to get a firstorder linear differential equation in PR: (9.12)

Stability requires that (9.13)

in other words that the MM schedule be steeper than the AA line. Equilibrium and comparative statics Equilibrium in this model obtains when both the manufacturing and agricultural markets clear. The equilibrium relative price can be obtained by setting PR in equation (9.12) and solving for equilibrium (9.14)

With PR at its equilibrium value and agricultural output predetermined, the market for XM clears. To find the equilibrium value of manufacturing sector

216


Figure 9.2

output, substitute the value of get

into equation (9.4) and solve for XM to

(9.15)

When relative prices have stopped changing PA, PM, W and the overall price level are all changing at the same rate. Using equations (9.10) and (9.9) we can write the following expression for π, the overall rate of inflation, (9.16)

We now turn to comparative statics. An increase in government expenditure raises aggregate demand and the MM line shifts to the right. This is depicted in Figure 9.2. The increase in government expenditure increases XM via a Keynesian type relation. Given agricultural output, then, PR will rise. An increase in agricultural output implies a downward rotation of the A A schedule and the MM schedule shifts to the right. Since demand for food has unitary price elasticity, the incomes of farmers do not change, i.e. the percentage increases in XA is exactly equal to the percentage drop in PR. This is depicted in Figure 9.3. Asset effects Much of the financial wealth in many developing countries is held in the form of government debt and money. In this section we alter the model we


217

Figure 9.3

have been studying to incorporate asset effects in consumption because of the accumulation of cash balances. No other forms of wealth will be admitted. We shall simplify the analysis still further and assume that the government does not issue debt because the market for such debt is underdeveloped. Hence, nominal wealth is simply money, M. We define real wealth m as nominal wealth divided by industrial prices: (9.17)

Real asset accumulation must be (9.18)

The government earns revenue by issuing new money and by taxing incomes in both sectors (at the common proportional rate v). Hence we can write the government’s budget constraint as (9.19)

Substituting equation (9.19) into equation (9.18) we get an expression for the rate of change of real wealth (9.20)

where δ is the rate of inflation of the industrial price. Using a linear Taylor approximation we write

218


(9.21)

where the subscript ‘0’ denotes initial values. Hence we can write (9.22)

assuming, for simplicity, that initial inflation is zero. From this equation government expenditure must equal the sum of tax revenue, the revenue from the inflation tax and the change in real assets. Consumers consider the change in the real assets at par with income and, therefore, consume a fraction γ of it. Now, the agriculture market clears when (9.23)

Hence, equilibrium relative price PR occurs when (9.24)

For a reasonable solution we must have The market for manufactured goods clears when (9.25)

which can be solved for XM to yield (9.26)

For a meaningful solution The value of the government expenditure multiplier is now altered. It is now It is higher because there is another source of consumption demand, namely real assets. We now discuss short-run equilibrium and comparative static analysis just as in the non-monetary version of the model. The derivation of the condition for local stability is algebraically tedious and unilluminating. It is mentioned below and the formal derivation is left as an exercise for the reader. Short-run equilibrium To simplify the analysis we discuss short-run equilibrium under the assumption that both the food and manufacturing sector markets clear immediately. Substitute (9.24) into (9.26) and simplify to get (9.27)

which is a reduced form equation for the manufacturing sector output. All we need to get now is a reduced form equation for PR. We obtain this by substituting (9.24) into (9.26) and simplifying to get:


219

(9.28)

where is assumed positive. Indeed it must be under plausible assumptions about the values of parameters and if the model is to be locally stable. Comparative static analysis Suppose that government spending G goes up. Suppose, further, that before this increase the government budget was balanced. The increase in government expenditure, then, creates a budgetary deficit. From equation (9.27) the increase in government expenditure causes XM to go up. The budgetary deficit is covered by additional money creation, which, in turn, is considered to be wealth by the consumers. As consumption expenditure goes up, so does PR. Thus, the increase in government expenditure has the effect of stimulating industrial output, increasing output, increasing inflation and increasing the relative price of food. In this model restraining government expenditure is inimical to output growth. When agricultural output rises industrial output remains unaffected as an examination of equation (9.27) reveals. From equation (9.28) though, PR would drop and so would, therefore, overall inflation. An increase in agricultural output would leave manufacturing output unchanged and lead to a drop in the overall rate of inflation and the relative price of food. A MEDIUM-TERM DUALISTIC MODEL WITH AN IMPORTANT MANUFACTURING SECTOR In the model studied in the previous section agricultural output was constant and there was no serious discussion of savings and investment. We now analyse a model that takes these into account while still retaining a dualistic structure for the economy. The model to be studied in this section is complementary to the previous model. Production conditions in both agriculture and industry are considered important and there is much more serious modelling of savings and investment. However, although the relative price of food plays an important role in the analysis, dynamics of inflation are not studied. In other words, we concentrate on the real sector to the exclusion of the monetary issues. Agricultural output is determined by the stock of capital in agriculture whereas industrial output is initially modelled as being demand determined, in a different version of the model, on the capital stock. The analysis in the present section is based on the work of Taylor (1983) and Rattso (1989). Two versions of the model are considered. In both versions there is a well-defined production relation for agricultural output. However, in the first version (called Keynesian), manufacturing sector

220


output is determined purely by demand whereas in the second version (called neoclassical) there exists a well-defined production relation for manufacturing sector output. There is a good reason for considering these different adjustment mechanisms. The Keynesian version corresponds to the case where effective demand constrains industrial output; supply is passive. In the neoclassical version the role of supply is extremely important. This becomes particularly obvious when we consider the comparative static properties of the two models. The Keynesian version Output of food (XA) is equal to its consumption (CA). Food cannot be invested: (9.29)

Output of the manufacturing sector (XM) can be consumed (CM) or invested (I), (9.30)

The economy being considered here has a surplus of labour. (Remember our initial motivation that this is a ‘South Asian’ variant of the ‘alternative’ model.) Hence labour is plentiful, capital is scare. To keep the analysis simple we assume that the output of the agricultural sector depends solely on the amount of capital in agriculture (KA): (9.31)

where λ is a fixed output-capital ratio. Total demand can be written as the sum of demand for food and the demand for manufactured goods, both measured in terms of PM. We distinguish between three income classes according to propensities to save: agriculturists with propensity to save (sa), workers in the manufacturing sector with propensity to save (Sw) and profit recipients with propensity to save (Sz). Now, given the labour-output ratio (ß) in the manufacturing sector and the real wage (θ) in terms of PM, we can write total agricultural and industrial demand in terms of the manufactured goods (D) as (9.32)

The demand relationships for the two sectors are formulated as linear expenditure systems. Let the parameter ε allocate consumption demand between the two sectors and the parameter ρ represents the Engel effect. Thus we can write demand for the outputs of the two sectors as (9.33)


221

(9.34)

with We now move to a discussion of investment in agriculture (IA) and the manufacturing sector (IM). Total investment is the sum of the investment in the two sectors: (9.35)

There is a simple behavioural hypothesis for investment in agriculture. There is a target capital stock which is related in a fashion to the relative price of agricultural goods: (9.36)

with It is hypothesised that the higher the relative price of agricultural goods the higher, ceteris paribus, is the profitability of agricultural operations and the higher, therefore, is the desired capital stock in agriculture. Current investment in agriculture is postulated to be a constant fraction of the gap between the desired and current capital stocks: (9.37)

with Investment in sector M is the residual amount of manufacturing sector output left over after current consumption and investment in agriculture: (9.38)

Equations (9.30)–(9.35), (9.37) and (9.38) provide eight equations in eight unknowns: XA, XM, CA, CM, IA, I, D and PR. The parameters sa, sz, sw, ß, ε, ρ, k, k1 and k2 are fixed. There are two predetermined variables KA and θ. Substituting from equations (9.29), (9.31), (9.33), (9.35), (9.36) and (9.37) into (9.32) we can derive the excess demand equation for sector A goods as (9.39)

We can, similarly, write the excess demand equation for sector M output as (9.40)

The stationary equilibrium of the Keynesian version The stationary equilibrium depends on the values of the predetermined variables KA and θ. As these values change so will the stationary equilibrium. We must, first, discover when KA and θ are constant. From equation (9.37) we have

222


(9.41)

This allows for a negative growth of the capital stock which may be interpreted as depreciation or scrapping. Capital stock in agriculture will be constant when i.e. when Turning to now, we assume a target real wage. The real wage in terms of agricultural goods is θ/PR and the target real wage is θτ. It is assumed that the real wage adjusts according to the relation (9.42)

Hence the real wage will be constant if or We can use this model to study the medium-term dynamics of the economy. The strong rigidities of the production structure reflect the supposition that this economy has surplus labour. The domestic terms of trade, PR, have a central role since it affects both the capital accumulation and the wage formation through equations (9.37) and (9.42). It is natural, therefore, to seek to understand the dependence of PR on θ and KA—the two predetermined variables of the system. When θ goes up, real consumption expenditure goes up if which we assume. Hence, following an increase in θ, there develops an excess demand for food, and the terms of trade move in favour of agriculture. When KA goes up, so does agricultural output and an excess supply of agricultural goods develops. This should lead to a drop in PR. Hence we write (9.43)

where a sign below a variable denotes the sign of the corresponding partial derivative. Let us now examine the impact of changes in KA and θ on the supply of manufactured goods XM. When θ rises there will be an expansionary effect on XM given that the labour-output ratio is fixed. However, a strong Engel effect may disturb the picture through a negative real income effect related to changes in PR. We assume that the Engel effect does not dominate. Using these results in the dynamical equation for KA and θ we have (9.44)

and (9.45)

Let us study these two dynamic equations in the θKA plane.


Totally differentiating

223

gives

or (9.46)

Thus this schedule is upward sloping in the θ, KA plane. Totally differentiating we have

or (9.47)

Hence this schedule is downward sloping in the θKA plane. The two loci and are plotted in Figure 9.4. The equilibrium occurs at point z with the equilibrium real wage rate being θ* and the equilibrium capital stock being It is straightforward to check that z is locally stable. Let us, for instance, look at the path of adjustment from point Q. At Q both the real wage and the agricultural capital stock are below levels consistent with stationary equilibrium. Two alternative paths of adjustment are depicted

Figure 9.4

224


Figure 9.5

in Figure 9.4. Along the first path both the real wage and the agricultural capital stock grow steadily. XM will also increase since growth of θ and KA stimulate XM growth. Along the second path, on the other hand, the real wage grows too fast compared with KA. As KA and XA grow in zone II terms of trade move against agriculture so that θ starts to drop until we reach equilibrium at point Z. Comparative static properties of the Keynesian version Suppose that investment demand goes up. This implies that the KA necessary to have will be higher for every θ. In other words, the schedule in Figure 9.5 will shift to the right. Similarly the increase in investment will imply that the schedule will shift to the right. As is evident from Figure 9.5, KA will certainly rise. We have placed θ as having remained unchanged. Consider now the case when the real wage target goes up. This stimulates demand for food and, therefore, the terms of trade move in favour of agriculture. This, in turn, increases agricultural investment, capital stock and output. In Figure 9.6 an increase in the wage target shifts the line to the right but leaves the line unchanged. This version of the model, then, is entirely demand driven: the greater the demand, the greater will be output and investment. A neoclassical version The preceding version of the model does not consider capital or capacity to be a constraining factor. This is particularly at odds with the facts in most


225

Figure 9.6

developing countries that have traditionally been capital-poor. We now assume that production of XM depends on capital stock in the manufacturing sector: (9.48)

where ßM is the fixed output-capital ratio in the manufacturing sector. We also assume that there is a paucity of savings in the economy, hence investment is determined by available savings. We now have three state variables, θ, KA and KM, instead of the previous two, θ and KA. This may complicate the dynamics—hence we make the simplifying assumption that the real wage is rigid: (9.49)

It is being assumed that the real wage in terms of the agricultural good stays the same, i.e. the adjustment of the real wage is instantaneous. This is the reason that this version of the model is being called neoclassical. Now we are left with two state variables KA and KM. An immediate task now is to construct the investment function for the manufacturing sector. In line with our hypothesis for the agricultural sector we postulate that the desired capital stock in sector M depends on the terms of trade between agriculture and industry: (9.50)

where is the desired stock of capital and µ1, and µ2 are positive parameters. The higher the PR, the lower is the profitability in sector M and the lower the desired stock of capital.

226


Investment in the manufacturing sector is a partial adjustment to the gap between the actual and desired capital stock: (9.51)

is a positive parameter. The new temporary equilibrium determines values of XA, XM, CA, CM, IA, IM, I, D, θ and PR from equations (9.29)–(9.34), (9.48), (9.49) and (9.51). The excess demand equation can now be written as (9.52)

and (9.53)

In these two equations KA and KM are treated as predetermined (state) variables, given values of PR and IA as functions of KA and KM. The rest are parameters. We will assume the following functional relation between KA, KM and PR and IA. (9.54)

(9.55)

A justification for the above sign patterns runs as follows. When KA goes up, so does agricultural output. Hence the relative price PR falls. When KM goes up, so does XM and, therefore, so does income in the M sector. This will tend to increase demand for food and, therefore, PR. Further, we assume that an increase in KM and, therefore, XM will mean higher supply of investment goods to agriculture so that investment in agriculture, IA, goes up. With an increase in the capital stock in agriculture the gap between the actual and desired capital stock in agriculture goes down and so, therefore, does IA. We now write (9.56) (9.57)

We can depict the equilibrium and 9.7. Totally differentiating equation (9.56) gives

schedules as in Figure


227

Figure 9.7

(9.58)

Similarly, upon totally differentiating (9.57) we have (9.59)

From equation (9.58) we have

Thus in the KM, KA plane this schedule is upward sloping. From equation (9.59) we have

Thus in the KMKA plane this schedule is also upward sloping. If

then the model will be stable. In Figure 9.7 we have assumed this to be the case. Starting from Q1 where we have less than optimal capital stock we can have balanced growth to Z1. The equilibrium capital stocks are labelled

228


and Now, since the wage rate is rigid it is not possible for the wage rate to overshoot its target, unlike the case in Figure 9.4. COMPARATIVE STATICS WITH THE NEOCLASSICAL MODEL With the neoclassical model we conduct the same comparative static exercises that we have attempted with the Keynesian version. Consider, first, an increase in investment. In the present case investment in both sector A and sector M are important. We shall here interpret an increase in investment demand as an increase in investment in the manufacturing sector. Increased investment in M sector means that capital stock and output in this sector rise. Shifts in output in favour of the manufacturing sector mean that terms of trade move in favour of agriculture. The real wage in terms of M goods also goes up. Both these factors put the brakes on increased investment in the M sector. Hence, in the short run, the two sectors are in competition with each other. The effect of the increase in non-agricultural investment is shown as an upward shift to the schedule as in Figure 9.8. This conclusion need not hold when we move to the medium term. Suppose that the increased investment is financed by additional savings so that there is an equal drop in private consumption. This will tend to reduce demand for food, and therefore, relative price will start to move against agriculture. In response to the fall in relative price, investment in agriculture starts to drop. In Figure 9.8 this means that the schedule shifts to the left. The dynamics are as plotted in the diagram.

Figure 9.8


229

The dynamics studied in Figure 9.8 constitute the philosophy behind some planning models for developing countries—particularly the Mahalanobis model applied to Indian five-year plans. An increase in investment in the manufacturing sector allows us to both increase manufacturing sector capital stock and transfer agricultural surplus to the manufacturing sector through a change in the terms of trade. Since the agricultural sector is supposed to be (ultimately) subject to diminishing returns to scale while the manufacturing sector is productive, an obvious policy prescription would be to direct more and more investment away from agriculture into industry. This line of reasoning has its limitations, of course. Some of these may be mentioned here. First, as non-agricultural incomes rise, the demand for food may go up. This will ultimately start to exert upward pressure on the relative price of food. Moreover, it is wrong to presume that the productivity of the manufacturing sector will keep rising. A cursory glance at the Indian experience with the Mahalanobis model should convince us about this. Year after year the capital-output ratio in Indian manufacturing kept rising during the years of adherence to this model—pointing to severe inefficiencies. This is particularly true because an important assumption of the Mahalanobis model is that the economy has very limited trade possibilities and that production in the manufacturing sector is best organised in the public sector. The shielding of the manufacturing sector from both internal as well as international competition has had the effect of severely affecting the efficiency of operation of this sector. See, for instance, Jha and Sahni (1993) for details. Let us now consider the effect of a change in the wage target. In the Keynesian version of the model an increase in the real wage target was seen to be expansionary. In the present version, however, a higher real wage shifts the distribution of income in favour of workers. Since workers have a lower propensity to save than profit earners, savings and therefore investment decline. In Figure 9.9 the schedule shifts to the left and the schedule shifts downwards. The economy moves to a lower value of both capital stocks and, therefore, outputs. In stark contrast to the Keynesian version, an increase in real wage no longer stimulates output and growth. One can indeed make the more general statement that demand factors do not play a significant role in the growth process. This, the careful reader will recognise, is an essential message of the neoclassical theory of economic growth beginning with Solow (1956). The analysis in this chapter underlines the differences between demandand supply-side effects in determining output in a dualistic set-up. In the low growth ‘African’ model demand factors were paramount. In the ‘semiindustrialised dual economy South Asian’ model both demand and supply factors could be important. It is probably the case that demand factors are important in the short run but over the longer run supply considerations become significant.

230


Figure 9.9

Dualistic models of output determination are potentially relevant in countries where the rural sector is fairly traditional and where a conscious attempt has been made (principally by the state) to embark on a path of rapid industrialisation. A significant difference between such industrialisation and that in the present industrialised countries is that, whereas the latter enjoyed the benefits of relatively free competition, the former are beset by a number of imperfections in commodity and factor markets. Politically aligned trade unions can exert considerable pressure on wages and indeed affect the choice of technology in the industrial sector. Given the limited size of the market in most LDCs and the complexity of modern technology, producing units are very large and product markets oligopolistic. It should be appreciated that this experience is not characteristic of some LDCs, in particular of some Latin American countries. Compared with such countries most African and Asian LDCs attained independence relatively recently. This meant that the urge to industrialise quickly was more pronounced in them and, given the state of the world economy in the immediate post-Second World War period, the appeal of centralised planning was also more significant. In some important sense, then, the degree of dualism is more severe in African and Asian countries than in Latin American countries. It is in this sense that the macro model of Chapter 8 is presented as a prototype of a macro model for Latin American countries, and we make the claim that the macro models of this chapter are more relevant ‘alternative’ macro models for the developing countries of Asia and Africa.


231


2

Neoclassical macroeconomic theory experienced a resurgence in the 1980s. Not surprisingly during this same period, the governments of most of the industrialised nations were headed by conservative political parties. Neoclassical economic theory dismissed neo-Marxist theory as flawed and unrealistic. It also rejected structuralists’ claims that developing countries’ problems were due to structural impediments in the international economy and that domestic structural flaws required significant state intervention in the economy. To neoclassical economists, economic stagnation in developing countries was a by-product of poorly designed economic policies and excessive state interference in the economy. They argued that in order to stimulate the domestic economy and promote the creation of an efficient market, developing country governments had to eliminate market restrictions and limit government intervention. This was to be accomplished through the privatisation of state-owned enterprises, promotion of free trade, reduction or elimination of restrictions on foreign investment and a reduction or elimination of government regulations affecting the market. In sum, market forces, not government intervention in the economy, would bring about development in stagnating economies. This framework became the basis for the massive economic changes that occurred in Latin America after the onset of the debt crisis in the early 1980s and for the equally profound transformation of socialist economies after the fall of the Soviet Union. Pick any country from Latin America or the former Soviet Union and list the impediments to such a neoclassical transformation of the economy. Do you think that the neoclassical transformations of the country you have chosen have been successful? Hint: It would be useful to look at the economic profile over at least a five-year period of the country you have chosen on the World Bank’s or IMF’s websites. Do you agree with the view, held by some economists, that the structuralist viewpoint as espoused in this chapter is applicable to shortrun macroeconomic phenomena whereas neoclassical macro models are more suitable for the long run?

10


INTRODUCTION Growth economics as a modern discipline became popular after the Second World War when economists became concerned about the growth of war ravaged economies. If Keynes’ General Theory of Employment, Interest and Money was the answer to the problems of depression of the interwar years, the Keynesian growth theories of Harrod and Domar formed the intellectual basis for reconstruction and growth of the war ravaged world economy. Harrod and Domar emphasised the need to raise the rate of saving and investment to raise the rate of growth of the economy. The equilibrium growth rate of real output in such models was simply equal to the rate of saving divided by the capital-output ratio. Higher growth could be achieved only by either raising the savings rate or lowering the capital-output ratio, i.e., improving the efficiency of capital. For given technology, raising the rate of savings was the only feasible way of raising the growth rate. However, there was a catch. In true Keynesian spirit, investment and savings could not automatically equal each other. This would happen only when the actual rate of growth equalled the ‘warranted’ rate of growth. Any gap between the two would show up as even increasing instability. This was Harrod’s growth theory. Domar added to this the natural rate of growth—that of population. Unless the actual, warranted and the natural rates of growth were equal to each other, the economy would experience instability. The basic attraction of the Harrod-Domar model was that it used the, by then, highly popular theories of Keynes to a study of long-run economic growth. As such, these theories became very popular and their use spread rapidly, even to developing countries. Thus, the early planning models of India were based on two- and then four-sector disaggregations of the HarrodDomar model. These were then popularly known as the FeldmanMahalonobis model. Such theorising provided a rationale for government intervention to raise the rate of saving and investment and to invest in heavy industry. This also supported the then quite prevalent view that since capitalist

234


economies would find it difficult to make such interventions, economic growth would be higher in socialist economies. A good early treatment of the HarrodDomar model is available in the reader by Sen (1964). The late 1950s and 1960s saw a sharp improvement in the global economic outlook with healthy economic growth, particularly in Europe and North America. Employment and prosperity were at historically high levels and it appeared that the scourge of unemployment-related instability could finally be removed from theorising about economic growth. Long-term economic growth could formally be separated from short-term concerns such as unemployment and medium-term concerns such as business cycles. This led to the birth of the neoclassical model of economic growth. The origin of this model is traced to Solow (1956) but there were other important contributions, most notably by Swan (1956). Economic growth is now a major area within the discipline of economics. There are not only excellent surveys of this literature but also surveys of surveys. In this chapter we have the modest task of reviewing some key elements of the theory of economic growth insofar as they are relevant to the experience of the developing countries. We first study the Solow model, then various versions of the ‘new’ growth theory. We also consider salient aspects of the growth experience internationally and review the literature on convergence in closed and open economies. THE NEOCLASSICAL GROWTH MODEL Solow (1956) considered a growing economy as one in which demand plays no role, since we are concerned about the long run. A related purpose of the Solow model is to discover how much of economic growth is due to factor accumulation (something that the Harrod-Domar model was also concerned about) and how much is due to technical improvements—a factor essentially ruled out by the Harrod-Domar model. The basic structure of the Solow model is as follows: (i) Real national output is given by supply conditions alone. Output supply, in turn, is given by a production function of the form: (10.1)

where Y, K, L, A are, respectively, real national output, the capital stock labour and a factor augmenting the productivity of labour.1 Thus we work with a one-sector economy with capital and labour as the only inputs. (AL) is sometimes referred to as effective labour supply. (ii) This production function has the following important characteristics. (a) It exhibits constant returns to scale or is linearly homogeneous. In other words, when capital and effective labour rise by a factor λ > 0, output also goes up by the same factor λ. Thus:


235

Constant returns to scale is a crucial assumption for this model. This implies that the scale of production is large enough that the economy can operate at the nadir of the aggregate cost curve. Large economies would certainly satisfy this assumption but not small economies. (b) The marginal product of each factor of production is positive but diminishing. In other words

(c) The production function satisfies the Inada conditions, which are:

In other words, as a factor becomes increasingly scarce, its marginal product becomes arbitrarily higher and as the factor becomes increasingly abundant, its marginal product becomes smaller and smaller. Since the production function displays constant returns to scale we can divide through by AL to write the production function in its ‘intensive’ form: (10.2)

where is the capital-labour ratio. Since F is characterised by constant returns to scale it must be the case that f has diminishing returns so that and where f′ denotes the first derivative of f with respect to k and f″ denotes the second derivative. To simplify matters and without loss of generality we work with a CobbDouglas production function, which we write as: (10.3)

with As is well known, under conditions of perfect competition, α represents the share of capital in total output and refers to the share of effective labour. The intensive production function can be written as: (10.4)

Since it follows that this production function shows diminishing returns. Remunerations to capital are, under competitive conditions, f′ and the remuneration to effective labour, from the Euler equation,2 is It is clear that the production function in (10.4) satisfies the Inada conditions. Now, let us introduce economic growth into the picture. People save a constant fraction s of their income. Thus savings, S, are equal to sF(K,AL). This is equal to gross capital formation. Net capital accumulation is This is equal to gross capital formation less depreciation:

236


where δ is the rate of depreciation of capital. The fundamental variable in this model is the capital-labour ratio, k. We now proceed to define a dynamic equation for it. The population growth rate is n so that

Also assume that the growth rate of the productivity of labour is g:

Now realise that (10.5)

Hence (10.6)

Substituting appropriately in (10.6) we have: (10.7)

where the last equality follows from our using the Cobb-Douglas production function. This equation is represented in Figure 10.1.

Figure 10.1


237

In Figure 10.1 we plot the intensive production function kα as well as skα. Realise that at the point where will be zero so that k will be stationary, k* is called the steady state capital-labour ratio, in the sense that at k* there is no tendency for the capital-labour ratio to rise or fall. Realise also that for values of we must have k>0 from equation (10.7). Similarly for it must be the case that k>0. Hence, in the neighbourhood of k*, there is a tendency for the capital-labour ratio to move back to k* if it is displaced from it. This is called the local stability3 of the steady state equilibrium. In steady state equilibrium capital and effective labour grow at the same rate. Since k* is constant, f′, the rate of return to capital, is also constant. However, realise that the return to effective labour is constant. In other words, the return to actual labour is rising at the rate of growth of productivity of labour, g. Hence, in steady state equilibrium, the rate of return to capital is constant and that to labour is rising. The rate of growth of per capita income is, therefore, g and the rate of growth of total output is It is interesting to pose to this model the question of whether a higher savings rate will lead to a higher growth rate. Remember that in the HarrodDomar model so long as the stability condition was satisfied a higher savings rate would lead to higher growth rate. In the Solow model this will not happen in steady state. This is illustrated in Figure 10.2.

Figure 10.2

238


Figure 10.2 redraws Figure 10.1 for two different savings rates s1 and s2 with The associated steady state capital-labour ratios are and with Realise that in either steady state the rate of growth of output as well as output per capita are exactly the same. So, in steady state, higher savings rates do not make any difference to the growth rate. However, they do have an effect in transitions to steady state. As the savings rate rises workers have more capital to work with than before so that This raises the transition growth rate. (Remember that the rate of growth of output is

) As workers have more capital to work with, the wage

rate rises relative to the return to capital. But now remember that the savings rate is constant at the new higher level so ultimately in the new steady state the rate of growth of output and the rate of growth of per capita output go back to their steady state levels. Hence, a rise in the savings rate has a transitory effect on growth rates and factor incomes (Solow, 1988). The important point is to check how long this transition period is going to be. Before we come to that, another point is worth noting. Empirical studies conducted by Solow and others confirmed that factor accumulation—the growth of capital and labour—accounted for less than 20 per cent of the growth in major OECD economies. This ‘growth accounting’ confirmed the view that the major contributory factor toward economic growth has been the growth of the technical progress coefficient. This estimation was done using conventional measures of capital. When the measure of capital was expanded to include human capital, then the role of the technical progress coefficient was reduced. We will have more to say on that later. THE QUESTION OF CONVERGENCE AND THE SPEED OF ADJUSTMENT Although we know the Solow model to be stable, a question of interest is: how long does k take to reach k*? We proceed as follows. We know that depends upon k so that we can write Also we know that when Suppose we are interested in ascertaining how long it would take to reach steady state within a neighbourhood of this steady state. In that case we can take a Taylor expansion to write: (10.8)

Now we know that

(10.9)


239

The first part of the expression within brackets on the right-hand side follows from the steady state equilibrium condition and substitution for s from it. Now is the share of capital in total output so that we can write (10.10)

Substituting (10.10) into (10.8) we get: (10.11)

Equation (10.11) implies that in the vicinity of the steady state the capitallabour ratio adjustment is in proportion to the gap between the current and the steady state capital-labour ratio. This adjustment factor depends upon the share of capital in total output. To take a simple numerical example, suppose is about 6 per cent with labour growing at 1 to 2 per cent, labour productivity growing at 1 to 2 per cent and depreciation at 4 per cent. If capital’s share in output is a quarter then the adjustment is of the order of i.e. 4.5 per cent per year. This adjustment is quite small and clearly reaching steady state will take a long time in the Solow model. The Solow model identifies two sources of variation—either spatially or temporally—in output per worker. These two factors are capital per worker and the productivity growth coefficient A. However, we have also seen that steady state output per worker will grow at the rate g. Hence long-run rates of growth of per capita income across the world or over time will differ only because of differences in g. In particular, differences in the capital stock across countries or across time do not account for much of the vast differences in output per capita that are seen. To see this consider the fact that output per worker in the US today is ten times what it was 100 years ago. To explain this kind of difference purely through differences in the capital stock would require the capital stock to be higher by a factor of 101/ßk. If this would require that capital stock today should be 104 (= 10,000) times higher today than it was 100 years ago. This is simply not the case. In fact capitaloutput ratios have been more or less constant over time, so that capital stock today is ten times larger than it was 100 years ago—not 10,000 times. A further difficulty with adducing differences in output per capita to differences in capital stock is that this would imply vast differences in the return to capital. If markets are competitive then the return to capital is simply For the Cobb-Douglas production function, for instance, If and the difference in y is a factor of 10, the implied difference in the marginal product of capital is huge! Such differences in the marginal product of capital—especially across countries—cannot be expected to persist for long, if capital is mobile.

240


Even, over time, for the same country there are no such differences in the return to capital. Thus the major driving force behind differences in output per capita in the Solow model is the difference in labour productivity. However, by keeping this exogenous, the Solow model seems to be assuming rather than explaining it. To be more charitable, we can say that the labour-productivity term is simply a catch-all term, reflecting factors other than capital augmentation and labour growth that affect output growth. Some effort at explicitly modelling this labour productivity term came from the endogenous growth and human capital branches of the growth literature, which we will study shortly. Before that we will examine another important implication of the Solow model. A SIMPLIFIED VERSION OF THE ‘NEW’ GROWTH MODEL The ‘new’ growth model as an alternative to the neoclassical growth model has many versions and their predictions vary. Several useful expositions exist. See, for instance, Romer (1996). In this chapter we outline two simple versions. In the first version emphasis is placed on the generation of new technology through investment of capital and labour resources. The second model studied is a simple version of the human capital version of the ‘new’ growth theory. Capital and labour investment in augmenting technology This model emphasises a process by which devoting capital and labour resources to the production of new technology plays an important role. It is but natural that assumptions about the way the new technology gets generated would be a crucial factor in economic growth. A particularly simple form of a growth model embodying these features is outlined below. Aggregate output is given by: (10.12)

As before, this production function displays constant returns to scale. αK and αL are, respectively, the fraction of capital and labour resources that are devoted to the production of new technology. The generation of technological progress is given by:

(10.13)

The last expression is a Cobb-Douglas representation of the technology generation function. ß, γ and θ are parameters (fractions between zero and one) and B is a shift parameter. Although the production function for output has constant returns to scale, there is no presumption that the function depicting a generation of technology


241

also displays constant returns to scale. The constant returns to scale argument says that doubling capital and labour will be able to produce the same output twice over but this will not lead to the generation of new technology—just a replication of the old. This opens up the possibility of having diminishing returns to scale in technology growth. At the same time, it may be the case that interactions among researchers, the presence of fixed set-up costs, etc., may be important enough in research and development (R&D) so that doubling capital and labour more than doubles technological growth. Thus we may admit the possibility of increasing returns to scale. If the growth of technology is proportional to A; if the effect is stronger than proportional and it is weaker if Labour grows at the fixed exponential rate n. A constant fraction, s, of output is saved. Thus the dynamics of capital accumulation can be defined as: (10.14)

Now defining gx as the rate of growth of any variable X, we have: (10.15)

where is a constant. It is clear from equation (10.14) that the behaviour of the growth rate of capital depends upon the growth of the ratio AL/K. This growth rate is, simply, Hence gK is rising if and is falling if this expression is negative. If this expression is zero then gK is steady. We can depict this relation as in Figure 10.3.

Figure 10.3

242


Figure 10.4

Figure 10.5

To understand fully the dynamics of capital stock growth, we have to model the behaviour of gA as well. As discussed above, there can be three cases of this depending upon whether we have diminishing, constant or increasing returns in the production of knowledge. When diminishing returns prevail we can visualise the situation as in Figure 10.4. In this case there exists a locally stable steady state such that both gK and gA are constant. This is depicted in Figure 10.5. Figure 10.5 has the steady gA schedule steeper than the steady gK schedule. This is needed for the stability of the model. As depicted in the figure, not only does a steady state equilibrium exist, it is also unique. There is only one value (each) of gK and gA that yields this steady state equilibrium. At the equilibrium point E we must have and satisfy the following:


243

Figure 10.6

(10.16)

and (10.17)

Solving (10.16) and (10.17) simultaneously we have:

and that (10.18)

Clearly the dynamics in Figure 10.5 apply to the case when the denominator of (10.18) is positive, i.e., If this inequality does not hold, no unique steady state equilibrium is possible in the model. When the two schedules will look like those in Figure 10.6. The dynamic arrows indicate that the economy gravitates toward the middle range where gK and gA keep growing. Hence in this case output keeps growing at higher and higher rates. The intuition behind this is straightforward. As capital and the stock of knowledge grow the returns are more than proportionate so that output grows even faster, which leads to an even faster growth in K and A and so on. When the two schedules are parallel and the dynamics of the economy are similar to the case when the situation is depicted in Figure 10.7. The second case occurs when and the first when In the second case there is convergence to steady state growth; however, the growth rate is not unique.

244


Figure 10.7

A GROWTH MODEL WITH HUMAN CAPITAL This model emphasises the importance of investment in human capital as a source of economic growth. In this model people make two savings decisions—to invest in physical capital (savings propensity SK) or human capital (savings propensity SH). The interplay of these savings decisions is crucial to the ensuing growth. Denote the amount of human capital at time t as H(t). Output is given by: (10.19)


245

This is a standard Cobb-Douglas type of production function and distinguishes between three inputs: physical capital K(t), human capital H(t) and ‘effective’ labour A(t)L(t). This production function, like the Solow neoclassical model, has constant returns to scale. Labour force growth is as per the Solow model: (10.20)

The growth of the physical capital stock is given by: (10.21)

whereas the growth of human capital is given by: (10.22)

and the growth rate of A is a constant: g per period. Thus

Since the production function shows constant returns to scale, we can divide through by A/L to obtain the production function in its intensive form with lower case letters representing respective per effective labour magnitudes: (10.23)

Since the intensive production function shows diminishing returns to scale. The dynamic equations of the system is as follows:

(10.24)

. For k=0 we must have, then, This defines an equilibrium relation between k and h such that the effective physical capital-labour ratio remains constant. This is depicted in Figure 10.8. For the human capital per unit of effective labour to be constant we must have: (10.25)

This locus can be traced as in Figure 10.9.

246


Figure 10.8

Figure 10.9

Movements in h, k outside of equilibrium are depicted with the arrows. Combining these two figures together we get the dynamic representation of this economy as in Figure 10.10. At the equilibrium point, E, the physical capital per unit effective labour and human capital per unit effective labour are constant. Let these values be k* and h* respectively. Hence, output per unit effective labour remains unchanged. In steady state we must have:


247

Figure 10.10

(10.25a) (10.26)

Taking logs we have: (10.27) (10.28)

Solving simultaneously, the (logs of) the equilibrium values of k and h are:

(10.29) (10.30)

The log of the intensive production function is

and substituting from (10.30) and (10.31) we have: (10.31)

In the Solow model SH drops out so that the corresponding expression would have been written:

248


(10.32)

The addition of the human capital term has important implications for explaining cross-sectional differences in the levels of per capita incomes across countries. Kendrick (1976) estimated (for the US) that the stock of human capital is higher than the stock of physical capital. For the sake of a numerical illustration let us take If this is approximately the case, then the elasticity of y with respect to SK is much lower in the human capital version of the model than the Solow model itself. This is because the elasticity with respect to SH is higher than with respect to SK. Furthermore, this model predicts that if there are sharp differences in savings rates across countries, output per capita would vary more than what is predicted by the Solow model. If the parameters α, ß, n and g are similar across countries then differences in per capita income across countries can be ascribed only to SK in the Solow model, whereas in the human capital version of the model these differences can be ascribed to SK as well as SH. Hence, the human capital model has greater explanatory power than the Solow model. For instance, suppose and then if savings rates in country 1 are twice as high as in country 2, the difference in their per capita incomes would be of the order of:

Now Hence, in this case incomes per capita would differ by eight times. The difference would be much smaller if we only had differences in sK (the Solow model). If the country with higher savings propensities also had higher lower population growth, the differences in per capita income would have been even higher. Another point to note about the human capital version of the model is that it assumes diminishing marginal product of capital. Hence, rates of return to capital are higher in poor countries as opposed to rich countries. A natural corollary would then be that capital should flow out from rich to poor countries. The model then fails to predict the empirically observed fact that most capital movements are among rich countries. With only a few exceptions, developing countries do not get much capital flows. THE NATURE OF THE GROWTH EXPERIENCE There are several important aspects to the experience of economic growth over time as well as across space that we should keep in mind. Economists have been uneasy about several implications of the Solow model. In particular,


249

they have singled out three important considerations. First, is the question of exogeneity of technical change. Technical progress in the Solow model augments the productivity of labour but is completely exogenous to the economy. An economy can do nothing to boost its long-run rate of economic growth. A second factor that has given concern is that since technical progress would occur anyway in the Solow model why would anyone—firms, the governments, universities and research institutions—want to contribute to technical progress through R&D? There are no incentives as such for such endeavours. The final factor affecting the applicability of the Solow model is its predictions about convergence. Surely, as the world becomes increasingly integrated through capital flows technologies worldwide would converge. This should lead to a convergence in levels of output per capita across rich and poor countries. However, what has actually happened is that there seems to be convergence in the output per capita across rich countries but not across rich and poor countries. In fact there has been divergence across rich and poor countries. Table 10.1 illustrates some key aspects of the growth experience of key developed countries. However, the growth experience of developed and developing countries has not been similar. Most of the world’s economies are small. In 1960 the largest 5 per cent of the world’s economies contained 59.0 per cent of the world’s population; the largest 10 per cent contained 70.9 per cent. Twentyfive years later the largest 5 per cent of economies held 58.3 per cent of the population; the largest 10 per cent, 70.2 per cent. In both periods, the lower 50 per cent of the world’s economies ranked by population held in total less than 12.5 per cent of the world’s population. Table 10.1

Average is unweighted average. Source: Robert Summers and Alan Heston, Penn World Tables (1995) Philadelphia, PA: University of Pennsylvania Press.

250


Furthermore, the cross-country distributions of populations have been relatively stable. For the last 35 years, the percentiles associated with the distribution of population across countries have been remarkably stable. This does not mean that those countries now highly populated have always been highly populated, but rather that the distribution of cross-section differences has changed little. Growth models have been used to understand the behaviour of per capita income or per worker output (labour productivity). The facts about this are as follows. From 1960 to 1989, world income per capita grew at an annual average rate of 2.25 per cent. However, the time paths of per capita incomes in individual economies varied widely around that of the world average. In 1960 about 10 per cent of the world’s national economies each had per capita income less than 0.22 times the world average. But these economies contained 26 per cent of the world’s population. It appears that poor economies are mainly large, although it is actually China alone accounting for most of that population. By contrast, the richest 10 per cent of national economies each had per capita incomes exceeding 2.7 times the world average, while all together containing 12.5 per cent of the world’s population. By 1985 the 10th percentile per capita income level had declined to only 0.15 times the world average and the economies in that poorest 10 per cent then held only 3.3 per cent of the world’s population as China had become relatively richer and was no longer a member of this group. At the same time, the top 10th percentile per capita income had increased to 3.08 times the world average and the share of the world’s population in these economies fell to 9.3 per cent. Hence although the distribution of population across the world has been quite stable, the cross-country distributions of per capita incomes seem quite volatile. The extremes appear to be diverging away from each other with the poor becoming poorer, and the rich richer. In the middle ranges the picture is more complex. In 1960 the difference between the per capita income of the 15th and 25th percentiles was 0.13 times the world average income, which in 1999 had fallen to 0.06 times the world average income. The corresponding gap between the 85th and 95th percentiles fell from 0.98 times the world’s average per capita income in 1960 to 0.59 in 1989. Thus, while the overall spread of incomes across countries increased over this period, that rise was far from uniform. Many studies argue that additional control variables affect the transition to steady state. But the choice of such variables is rather ad hoc and sometimes the number of variables used can be very large. In this sprit Barro and Sala-i-Martin (1995) assert that with the right conditioning variables, a rate of convergence of 2 per cent per year is uniformly obtained across a broad range of samples. They draw two implications. First, in a Cobb-Douglas production function for aggregate output, physical capital’s coefficient is over 0.9, appreciably larger than the 0.4 implied by


251

factor shares in national income accounts. Second, convergence occurs; the poor do catch up with the rich. Mankiw, Romer and Weil (1992) provide an essentially equivalent-convergence analysis when they add human capital investment as an additional control. Their analysis differs from several such studies in that their modification of the basic growth regression is justified by an explicit economic model; namely, they estimate the exact law of motion generated by the Solow model with Cobb-Douglas technology. Barro and Sala-i-Martin (1995) rightly argue that the speed of convergence is also important. They hypothesise that poorer countries are catching up with the rich faster if the cross-section standard deviations of per capita incomes diminish over time. However, as Durlauf and Quah (1999) argue, this concept also has its problems. Thus, to the extent that there are differences in per capita output because countries have not reached their steady states, such differences will tend to grow narrow over time. Moreover, countries with high capital-labour ratios will have lower returns to capital than countries with low capital-labour ratios. Hence, in a world of internationally mobile capital, capital will flow from capital rich to capital poor countries. Thus, there will be a convergence of capital-labour ratios and, hence, steady state output per unit labour as well. Thus openness would appear to encourage convergence. If openness does indeed lead to convergence, is it convergence in growth rates or convergence in levels? Convergence in growth rates is the natural outcome if the open-economy extension means that growth in technology/ knowledge becomes a global phenomenon that affects all ‘open’ economies alike. In this case, the result is very similar to the neoclassical model. However, if the poorer countries’ growth rate is bounded from above by the leading country’s growth rate, initial differences will persist. In other words, convergence in levels requires that greater openness or globalisation leads to initially faster growth in poor countries. Consequently, convergence in levels is a much stronger result. However, endogenous growth models usually have little to say about level differences. As soon as the growth rate becomes endogenous, ‘initial conditions’, such as the timing of a growth-enhancing policy, can lead to permanent gaps between countries. To obtain convergence in levels, one would have to include a ‘catching-up’ mechanism in the model, for instance via copying, that ensures that the poorer country is not permanently left behind. Some authors have argued that one reason why convergence in growth levels or rates has not occurred across developed and developing countries is because of policy impediments in developing countries. Such authors have argued that many developing countries have been slow in reforming their policy environments with the consequence that growth has stagnated. This point is examined in a recent paper by Easterly (2001). He asks why median per capita growth in developing countries was higher (2.5 per cent) during 1960–79 when the policy regime was ‘unfavorable’. During the 39-year period

252


from 1960 to 1999 all factors that are termed ‘determinants of growth’ such as financial development, black market premiums, real over-valuation, improved steadily. Yet the median growth rate of per capita income in developing countries during 1990–99 was 0 per cent. Some authors have, therefore, termed the 1990s as a lost decade for developing countries. Indeed starting around 1980 there was a sea change in developing countries toward market-friendly economic policies and getting the prices right. International institutions aided this effort. Thus, the World Bank began ‘adjustment lending’ to developing countries in 1980. Such lending was conditional on implementing the new ‘market-friendly’ consensus on economic policies. At about the same time the IMF also expanded its portfolio of lending. Between the two of them these two Bretton Woods institutions made 958 adjustment loans to developing countries during 1980–98. But as Krugman (1995) notes: ‘the real economic performance of countries that had recently adopted Washington consensus policies…was distinctly disappointing.’ Possible paths of convergence and divergence of per capita output are noted in Figures 10.11(a) and 10.11(b) respectively. In Figure 10.11(a), Tr represents the steady state growth path of the rich country. The slope of this line represents its rate of growth. For the poor country a number of options are outlined. A steady state growth path in which the rich and poor countries grow at the same rate is Tp. A favourable shock at t0 that will lead to convergence of output per capita in rich and poor countries is shown by An unfavourable shock that lowers the rate of growth of the poor country in the short run but leads to the same steady state growth as in Tp is Dotted lines represent movement outside of steady state. A situation in which there is divergence between the rich and poor countries is represented by Figure 10.11(b). Tr and Tp represent the steady state growth paths of the rich and poor country. As can be seen, per capita outputs are going to diverge over time across the two countries. Irrespective of a favourable shock or an unfavourable shock the steady state growth rate is the same as in Tp, and long-run income per capita in the rich country will increasingly diverge from that in the poor country. The solid lines in both diagrams are steady state paths whereas the dotted lines represent transition to equilibrium in response to a shock. The literature generally distinguishes between two types of convergence. The first form as depicted by the path in Figure 10.11(a) is one in which ‘absolute’ convergence occurs. In this case convergence occurs not just in the sense of the same growth rates but the same growth paths. A weaker form of convergence is ‘conditional’ convergence. Depicted by the paths and Tp this phenomenon underscores the possibility that countries may have the same growth rates but different growth paths. The vertical distance of the LDC curve from the rich country curve represents differences due to differences in underlying parameters, such as savings rates and population growth.


253

Figure 10.11

To summarise then, the Solow model predicts that the only way income per capita can vary across countries is if there are differences in their rates of technical progress. If all countries have access to the same technology then their incomes per capita will not diverge in the long run. However, even if countries have access to the same technology, their incomes per capita can

254


vary in the short run. Another message we get from the Solow model is that the ‘long run’ is a very long period of time. Adjustment to steady state is slow. An important question that arises at this stage is whether greater openness would lead to slower or faster convergence. In the Solow model if capital is free to move across countries then it will follow from countries with low rates of return to capital (rich countries) to countries with higher rates of return (poor countries). This would imply that most capital would flow from rich to poor nations. However, the real world experience has been just the opposite. Most capital flows are between rich countries. New growth theory and convergence in open economies Open-economy extensions of new growth theory are less easily discussed than in the case of the neoclassical model. The main reason is that there are many different variants of endogenous growth models. There is no core model that could be extended step-by-step as in the previous section. There is also no easy way of selecting one variant over another. Hence there is only limited scope for empirical testing of new growth models. For these reasons it is useful to discuss a number of important questions that need to be asked of any specific open-economy new growth model in order to determine its potential effect on growth and convergence in a poor country. We take as maintained the hypothesis that ‘openness’ and economic integration are good for poor countries. However, there are models where this is not the case. Thus, in contrast to neoclassical growth theory, there is no unambiguous case for greater openness that can be derived from endogenous growth theory. A key feature of all new growth models is that the rate of technical progress is determined endogenously. This means that policies can alter the growth rate and not just the level of per capita income, which makes them much more potent in the long run than in the neoclassical model. Thus, depending on the specific model of the economy, policies regarding openness can have very powerful effects. Since increased globalisation acts in a similar fashion, i.e. as a parametric shift in an underlying fundamental, it has the same effects as policies that increase openness. Similarly, trade restrictions now have not only level but also growth effects. If the latter are negative, the case for free trade and greater integration is even stronger than in the neoclassical model. But if the growth effects are negative, they may potentially outweigh the static losses from trade restrictions. One explanation for slow convergence has been provided by Barro, Mankiw and Sala-i-Martin (1995). They develop an open-economy version of the Solow model with human capital to explain why there might be slow convergence with capital flows. In their model, countries can borrow and lend abroad. But potential borrowers can obtain loans from abroad only to


255

finance increases in physical not human capital. Human capital accumulation has to be generated out of domestic savings (i.e. foregone consumption), this slows down the adjustment process to the steady state. In this model it can be shown that under a number of realistic assumptions, it is possible to obtain convergence speeds roughly in line with the cross-country regression evidence. But in this approach there is a somewhat artificial distinction between the level of credit constraints affecting physical and human capital (Obstfeld and Rogoff, 1996). In practice it is neither possible to fully borrow against physical capital, nor is it impossible to borrow abroad to finance human capital accumulation. So it may be better to allow for capital market imperfections or adjustment costs to generate delays in the convergence process. As Gundlach (1999) demonstrates, the per capita output growth rates on the way to the steady state implied by the Barro et al. (1995) model are much higher than what one observes in reality. It is also important to note that the tendency for capital to flow from rich to poor countries depends on both countries sharing a common rate of underlying growth, i.e. rate of technical change. However, if capital flows are the only form of interaction then the two countries will grow at different rates in the long run. In this case capital flows may increase the rate of conditional convergence, i.e. tendency for countries to return to their own steady state paths, but these long-run paths themselves may diverge. Moreover, capital may also flow in the ‘wrong’ direction, i.e. from poor to rich countries. In this context it is interesting to note the result of Gundlach (1999) who shows that the Barro et al. (1995) model is only compatible with actual rates of growth if one gives up exactly this one key assumption, i.e. if one allows the rate of technological progress to differ across countries. Finally, it needs to be realised that when part of the domestic capital stock can be owned by foreigners, a distinction has to be made between GNP and GDP. If one compares a capitalimporting poor country with a capital-exporting rich country, GDP per capita differences will understate the true differences in standard of living. Thus, the equalisation of domestic rates of return to capital does at best ensure conditional convergence in GDP. However, since part of the poor country’s output is now owned by rich country residents, there will be a permanent gap between their respective incomes per capita. Technology and knowledge flows What is normally considered the closed-economy version of the neoclassical model in fact contains a strong open-economy assumption, i.e. that all countries have access to the same technologies and that knowledge flows are instantaneous and without costs. Without this assumption one can no longer predict that there will be a common long-run growth rate across countries in a closed-economy setting. Non-divergence would then be purely accidental.

256


In the normal state of affairs countries should diverge if there are no technology or knowledge flows. A recent model by Lucas (2000b) illustrates this point very well. He simulates a model world economy in which all countries asymptotically grow at the same exogenous rate. However, countries move from an initial state of stagnation to modern growth at different points in time, thus there is a leader-follower situation. Whether a stagnating economy starts to grow depends stochastically on the income gap between the follower and the countries that have already started to grow. When it starts to grow, it grows more quickly than the leading countries until it has caught up. In the end all countries reach the same equilibrium level of income. Lucas’s model is not intended to explain growth as such. What he shows is that one can get a situation where the world distribution of income initially widens before narrowing again just through the fact that some countries start to grow later than others. However, the result that all countries finally converge is due to the built-in assumption of perfect diffusion of technology and knowledge in the long run, even if there are temporary delays for the lagging countries. If diffusion is less than perfect, then we may get divergence between countries even if there is some exchange of technology and knowledge between countries. To see this consider a set of rich and poor countries with diffusion taking the simple form: with where A is the technology coefficient. Both countries will grow at the same rate since the rate of growth of technological progress in the rich country will be the same as in the poor country. However, if the link between the rates of technical progress is then the rich country will grow at a faster rate than the poor country unless ß happens to be equal to one. This simple example illustrates yet again that some very specific assumptions are needed to ensure that countries do not diverge in the neoclassical model, even when knowledge and technology are transmitted between countries. International trade has often been viewed as a key channel for the transmission of knowledge and technology between countries. If this is indeed the role that trade plays, then clearly the more countries trade with each other the faster will be the rate of convergence between them. But trade may also perform a subtler role, as shown in a recent paper by Ventura (1997) who analyses the neoclassical model for a small open economy. The motivation for his paper comes from the observation, due to Young (1995), that the high post-war growth rates in the East Asian miracle economies were largely due to factor accumulation and not rapid productivity growth. This is difficult to square with the evidence, at least in the context of a closedeconomy neoclassical growth model, since diminishing returns to capital would ensure that growth slows down fairly quickly. Of course, this is not what has been observed in East Asia where a number of countries, such as South Korea and Hong Kong, have grown very rapidly over a long period of time.


257

In Ventura’s model where integration into the world economy ensures that a weak form of factor price equalisation (FPE) holds. Although final goods are non-traded, there is trade in intermediate goods. Returns to capital are still diminishing, but now only at the global level. That is, world averages behave just like the closed economy. But as long as a given economy remains ‘small’, factor returns are exogenously given and thus independent of the domestic capital stock. This has the surprising result that a change in the savings rate can have a permanent effect on growth, even in the context of the neoclassical model. The reason is that, as the capital stock rises, resources are moved into more capital-intensive industries. Normally this would lead to a fall in capital returns over time, but here FPE ensures that the value of the marginal product of capital remains unchanged. In one sense, this open-economy extension transforms the neoclassical model into something very close to an endogenous growth model. This is also the key implication for our purposes. It means that a poor, small open economy can grow at a faster rate than the richer countries (at least for some time) and thus catch up with them. Growth is export-led, and as the country becomes richer its production and exports become increasingly more capitalintensive. The Ventura model thus makes a very strong case for an outwardoriented growth strategy. Summary All three types of open-economy extensions to the neoclassical growth model considered here have positive implications for convergence, and thus also for growth in a developing country. Poor countries benefit from greater integration into the world economy either through technology or capital imports, or indirectly via the possibility to export. Therefore, if one accepts the neoclassical growth theory framework, ‘openness’ is always beneficial. In fact, without the open-economy assumption of perfect knowledge and technology flows, there will be divergence even in the neoclassical model. The determinants of growth literature If there are regular patterns to the growth experience of countries over time and space, then there would be important lessons for developing countries. Needless to say there have been several efforts to unravel such patterns. We have already considered one aspect—of whether greater openness is conducive to economic growth. Barro (1991) started this literature by trying to explain cross-country differences in growth rates in terms of differences in initial income, primary and secondary enrolment, political instability and deviations from purchasing power parity. In subsequent work, Barro (1997) added

258


fertility and life expectancy to the list of initial conditions affecting crosscountry growth. Fischer (1993) added macroeconomic variables like the budget deficit, black market premium and the inflation rate. King and Levine (1993a,b), Levine and Zervos (1993), and Levine, Loayza and Beck (2000) have stressed financial development (most often measured by the ratio of M2 to GDP) as a robust causal determinant of economic growth. Dollar (1992) stressed a measure of real exchange rate overvaluation as a proxy for outward orientation and thus a determinant of growth. Easterly and Levine (1997) added a measure of infrastructure development (telephone lines per capita) to the list of initial conditions affecting growth, and Knack and Keefer (1995) highlighted the importance of well-developed institutions for growth performance. Some authors have emphasised the deleterious impact of excessive initial inequality on growth. Examples of this literature include Alesina and Rodrik (1994), Deininger and Squire (1998), and Persson and Tabellini (1994). Gylfason (2000), Sachs and Warner (1995), and Tornell and Lane (1999) have argued that dependence on natural resource exports deprives such countries of the externalities that are associated with manufacturing exports and further lead to considerable rent-seeking. If such factors were the only ones responsible for economic growth then we should have had acceleration in the rate of growth in developing countries in the 1990s. However, the opposite seems to have happened. Easterly (2001) argues that the most important reason for this is the slowdown in the industrialised economies during this period. Hence, factors other than its characteristics can influence the pattern of growth of a country. A secondary factor also emanating from industrial countries was the rise in world interest rates that increased the debt burden of developing countries. Easterly, Kremer, Pritchett and Summers (1993) made an earlier argument for the importance of random shocks relative to national economic policies, based on the weak cross-period persistence of growth rates contrasted with the strong persistence of policies. Pritchett (1997) also describes patterns of growth in developing countries reaching ‘plateaus’ or remaining on ‘plains’, but again his emphasis is on cross-country variation rather than the aggregate performance of developing countries. Another interesting contribution in this area is by Rodrik (1999) who asked the question ‘where did all the growth go?’ He found that countries that are characterised by ethnolinguistic fragmentation and had poor institutions suffer a particularly strong negative impact from terms of trade losses. He explained the variance around the mean growth decline, showing why some countries collapsed much more than others. Since regressions explain variation around the mean rather than the mean itself, his regression does not explain the mean growth decline itself: 2 percentage points in his data. Even countries with zero social conflict on his measures had a significant decline in growth.


259

However, Kenny and Williams (2001), in an interesting piece on what empirical facts are known about economic growth, argue that empirical evidence is hardly unanimous in support of a particular view of the growth process—be it the Solow type exogenous growth theory, various versions of the endogenous growth model or the human capital version of the growth model. It is also the case that the empirical evidence is unable to firmly pin down a reasonably parsimonious list of variables that affect economic growth. Kenny and Williams (2001) cite the work of Sherden (1998) which presents fairly convincing evidence that the short-term forecasts of economic growth are quite poor. On average, it appears that the best forecast of this year’s growth rate is simply last year’s growth rate. Further, the accuracy of economic predictions has not improved over time. Modern cross-country growth regressions have listed well over 100 variables that might affect the rate of economic growth. These include economic, cultural, sociological and geographical factors. However, most of these variables were mentioned in the classic work on economic growth by Arthur Lewis (1954). Yet there does not seem to be any homogeneity in this pattern of dependence across time or countries. In any case, the number of variables that are to be finally included is so large and their quantitative impact on the rate of economic growth so uncertain that there is hardly any meaningful guide to policy. In fact Pritchett (1996b) goes one step further when he argues that policy changes over time and policy differences across countries within the OECD, as well as wider samples of countries, have had almost no effect on very long-run rates of economic growth rates. As Kenny and Williams (2001) put it, ‘policies vary too slowly to explain the volatility of shorter-term growth rates, and too fast to explain the stability of longer-term growth rates’. Kenny and Williams (2001) point out that two kinds of universalism have led to the disappointing results on the determinants of economic growth. They question the extant literature’s assumption that all economies are subject to similar laws in the same manner as physical experiment conducted anywhere would give the same result. However, social processes are complex and it becomes difficult to describe the complex behaviour of humans in the same way as one can describe the behaviour of particles of matter. Thus growth processes may be different in different parts of the world and one should allow for the possibility that the same policy measure may have different implications in different parts of the world. For example, economic reforms led to a doubling of the GDP in China but Russia’s GDP fell by 50 per cent. Kenny and Williams (2001) further challenge the view that components of economies and economic processes everywhere are comparable and that these components interact in much the same way in different countries. These two assumptions are not satisfied in reality making predictions about whether a growth strategy that has worked in one country will necessarily work in others difficult to sustain.

260


CONCLUDING REMARKS The study of economic growth is fascinating. So much so that Lucas (1988) noted that once one starts thinking of economic growth it is difficult to think of anything else. This chapter has sketched key aspects of this exciting area especially in regard to implications for developing countries. Apparently there exist models that can explain the vast gaps in per capita output across rich and poor countries. However, there seems to be little agreement about the factors responsible for long-run economic growth. Hence there is little unanimity about the policy initiatives that need to be taken to boost longterm economic growth in developing countries. DISCUSSION QUESTIONS 1

2

3

4

In the simple Solow model write down an expression for per capita consumption. Write down the condition necessary to attain the maximum per capita consumption. This has been known in the literature as the ‘Golden Rule of Economic Growth’. See Phelps (1966). A vast literature on the determinants of economic growth exists and this chapter has surveyed some of this. A problem with this literature is that it gives rather elaborate and conditional explanations for why countries grow at different rates without really being able to provide effective policy proposals for accelerating the rate of growth. Do you agree with this statement? Why? This chapter has presented essentially the real theory of economic growth; in particular, it has admitted no monetary influences on economic growth. The basic model of ‘money and growth’ was elaborated by Sidrauski (1967a, b) and Tobin (1965). They consider wealth to consist of capital and real cash balances. Output is defined in the standard way but income includes output as well as the accumulation of real cash balances. Hence there are two dynamic equations—one for capital per unit labour (as in the Solow model) and the other for real money stock per unit labour. Analyse the dynamic properties of such a model. Hint: See Burmeister and Dobell (1970, Chapter 6). Consider the one-sector Solow growth model with two classes of people— workers and capitalists. Capitalists earn only interest income whereas workers earn only wage income. Only capitalists save and, therefore, all capital is owned by capitalists. Analyse the steady state equilibrium of this economy. How would your answer change if workers saved a constant fraction of their output but their savings propensity was lower than that of the capitalists? Hint: These are the famous two-class growth models of Kaldor and Passinetti. For an exposition see Burmeister and Dobell (1970, Chapter 2).


261

NOTES 1

2 3

The technical productivity factor A here is considered to be augmenting the supply of labour. This is called labour-augmenting ‘neutral’ technical progress. This technical progress is neutral since it is not embodied in new machines. To attain stability the only form of neutral technical progress that can be admitted is the labour-augmenting type (see Burmeister and Dobell, 1970). When the production function is such that the elasticity of substitution between capital and labour is one, it does not matter whether this neutral technical progress is capital augmenting or labour augmenting. This will be seen below. The Euler equation is applied as follows. Output per unit effective labour is f and the payment to capital (per unit effective labour) is kf′. The remainder must be left over for effective labour. The steady state is also globally stable in the sense that even with a large displacement from the steady state, the economy will tend to revert back to k*. For a proof see Burmeister and Dobell (1970). What is important here is, of course, the time taken to revert back to steady state. This is the question of convergence which will be taken up later in this chapter.

Part III

Policy Dilemmas Faced by Developing Countries

11

An evaluation of the IMF programmes in developing countries

INTRODUCTION The effects of IMF conditionalities have been subjects of some controversy. In Chapter 7 we have examined the IMF’s approach to economic stabilisation and discussed some empirical evidence presented to evaluate the effects of these policies. Whereas the evidence presented by Gylfason (1987) and others questioning the effectiveness of the IMF is cast in terms of the short-term programmes of the IMF, this controversy seems especially heated for countries facing acute balance of payments problems and currency crises, as witnessed in 1997 in Korea, Indonesia and Thailand and elsewhere. Stiglitz (2000), for example, supports critics of the IMF who argue ‘…the IMF’s economic “remedies” often make things worse—turning slowdowns into recessions and recessions into depressions’. Some academic work also reaches this conclusion. Bordo and Schwartz (2000), for example, conclude, ‘…the recent spate of [IMF] rescues may be the case of the medicine doing more harm than good’ (p. 60). Similar statements by Krugman and other leading economists can be cited. How would one empirically evaluate the effects of the IMF? Most empirical studies using panel data sets and regression techniques find that IMFsupported programmes improve the balance of payments and current account (e.g. Bordo and Schwartz, 2000; Khan, 1990). This should hardly be surprising since the IMF gives out a loan in foreign currency to these countries. As a matter of fact a key purpose of the IMF is ‘…to give confidence to members by making the Fund’s resources temporarily available to them under adequate safeguards, thus providing them with the opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity’ (IMF Articles of Agreement, Article I (v)). While there is little room for argument on the question of the effects of IMF programmes on balance of payments, the question of the real output effects of these programmes is mired in controversy. At a superficial level it

266


would seem to be the case that countries would choose to be part of an IMF programme only if such a move was in their interest. In other words, participation in a programme would presumably be unlikely if the output costs were perceived to be particularly large, outweighing the benefits arising from improvement in the balance of payments, continued access to credit markets and so on. Stiglitz (2000) and others argue, however, that while officially the IMF does not force countries to participate in programmes and negotiate conditions, ‘In practice, it undermines the democratic process by imposing policies.’ A number of studies have investigated the output costs of joining IMF programmes. But there is no agreement among the various contributors. The analyses range from Przeworski and Vreeland (2000) who argue that the output effects are strongly negative to Dicks-Mireaux, Mecagni and Schadler (2000) who discover positive output effects of IMF programmes. These conflicting results arise from several sources, including differences in the types of IMF programmes that are investigated; differences in the groups of countries that are investigated (e.g. poor developing versus emerging market economies); differences in the methodologies that are employed; and, perhaps most important, how other factors influencing output growth are taken into account. In this context Hutchinson (2002) sheds useful light on the role of severe currency and/or balance of payments crises on output growth and how these events interact with subsequent participation in IMF programmes. His approach is to measure the output cost of participation in an IMF programme, and investigate whether there are feedback effects that make implementation of programmes especially problematic in the immediate aftermath or concurrent with an ongoing balance of payments/currency crisis. Thus three questions are addressed: (i) given that a country is already facing a severe currency crisis, does participation in an IMF-supported stabilisation programme tend to make real GDP growth weaker? (ii) Is it possible to identify channels (policy instruments) through which IMF-supported programmes can affect GDP growth? (iii) Is it possible to quantify the output effects of the IMF-supported programme in East Asia during 1997–98? To address the first question Hutchinson controls for the effect of a currency crisis on real GDP, and isolates the additional effect arising from IMFprogramme participation at this time. Further, beyond providing countries with access to substantial lines of credit, IMF programmes are generally associated with conditions on the future conduct of fiscal, credit and other policies. Thus it is important to understand whether the adoption of IMF programmes makes it difficult for countries to follow growth-enhancing policies ex post facto. If the critics of the IMF are right, conditionality leads to overly restrictive macroeconomic policies and poor output performance. Because of inertia in rapidly changing policies, these restrictive policies would be hard to change immediately after an IMF programme is over, with

Evaluation of IMF programmes in developing countries

267

deleterious consequences for economic growth. Finally, it would be useful to understand in the context of the East Asian crisis whether those countries that followed IMF programmes (Korea, Thailand, the Philippines and Indonesia) had a worse economic performance compared to the country that did not participate (Malaysia). THE IMF’S RESPONSE The IMF has of late become ever more responsive to such criticisms1 and even invites feedback from interested parties. It, however, professes confidence on the value of recent IMF programmes to the real side of LDC economies. To provide empirical support, it cites evidence of a panel data nature provided by Bordo and Schwartz (2000) and Khan (1990). These studies find that IMF-supported programmes improve the balance of payments and current account. This is not surprising since a key purpose of the IMF is ‘…to give confidence to members by making the Fund’s resources temporarily available to them under adequate safeguards, thus providing them with the opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity’ (IMF Articles of Agreement, Article I (v)). The rationale for IMF financing At present, the IMF’s balance of payments financing is supplementary to what is available from the private sector (as well as from other official lenders). However, in the original Bretton Woods scheme of things, private capital flows were not expected to be a major factor. This was primarily because capital account convertibility was expected to be the exception rather than the rule, although widespread current account convertibility was expected. In such a context, the financing role of the IMF was clearly intended to be central to the international monetary system. In practice, however, there were few drawings on IMF resources in the early postwar years, and, in recent years, the vast expansion of private capital flows has reduced the need for IMF financing for those countries with secure access to international capital markets. This includes most of the industrial countries and some developing countries as well. IMF financing was very important for several developing countries after the oil price crisis and, very recently, it has become important in the context of the Mexican, Southeast Asian and Russian crises. One important question that arises here is that, in the changed circumstances after the immediate impacts of the contagion have been fought off, whether the IMF’s role as a supplier of balance of payments financing can continue to be justified. If such a rationale can be shown to exist, a second question then arises: are IMF lendings consistent with these objectives? As in other areas of government intervention, the general rationale for the IMF’s financing activities needs to be grounded in the existence of market

268


imperfections or in the inadequate provision of international public goods (Camdessus, 1995). The argument would then be that IMF lending would be able to help remedy this state of affairs. At this level of generality, the rationale for the IMF’s financing activities remains much the same today as it was at the time of the Bretton Woods conference; unimpaired access to private financing of payments imbalances is generally available to some important countries in the world economy, but not to all countries and not under all circumstances. Accordingly, in the absence of IMF financing (accessible under appropriate conditionality), many countries facing balance of payments difficulties might need to adopt policies that would be unnecessarily destructive of national and international prosperity. By helping to avoid such outcomes, IMF financing benefits both those countries experiencing balance of payments difficulties and their partners in international trade and finance. This then provides the basic rationale now as it did at the time of the foundation of the IMF. We now elaborate this rationale. Market imperfections Market discipline may typically speed up the recognition that certain policies (such as uncontrolled expansion in public expenditure) are unsustainable. This would then necessitate the adoption of remedial measures and some course correction. But if the market does not function perfectly, the recognition that certain policies are unsustainable might come too late. The change in market sentiment at this time can lead to a violent outflow of capital, causing a balance of payments crisis. This has happened much too often for comfort in the recent past. In the absence of IMF funding, governments would have to adopt very harsh measures with large and uncertain costs in terms of output, employment and growth. The IMF can, thus, perform a useful role at this stage. Four important factors affecting private lenders can cause these market imperfections to occur. (i) Imperfect information about a country’s policies and the economic circumstances facing it can lead to shifts in sentiment that are not justified by fundamentals. A recent example is the large flows to most emerging markets of Southeast Asia in the early 1990s and their sharp interruption or reversal after the recognition that exchange rates were overvalued and foreign debt positions were unsustainable. (ii) Coordination problems among lenders can lead to bandwagon effects or ‘free rider’ behaviour. The Latin American debt crisis of the 1980s illustrates both aspects of the problem. Lack of concern for potential risks and competition among banks to recycle the revenues of oilproducing countries produced an initial period of overlending in the late


269

1970s. The onset of the debt crisis in 1982 led to a protracted period of negotiation of debt relief, during which creditors attempted to position themselves favourably with respect to both the borrowing country and other creditors. (iii) Problems in enforcing loan contracts on sovereign borrowers can produce a reluctance to lend. The inability of borrowers to credibly commit to repay loans, together with the absence of clear recourse of creditors in the event of default, may lead to suboptimal lending (Eaton and Fernández, 1995). (iv) Multiple equilibria can exist. For instance, there may be one equilibrium in which lenders believe the authorities intend to do what they say and in which money is available on reasonable terms that allow policy commitments to be fulfilled, and another equilibrium in which lenders do not believe the authorities and in which investors demand a high risk premium that, in turn, provokes a balance of payments crisis. There is no assurance that the market, left to itself, will choose the better of the two equilibria. Although empirical work in this area is still in its infancy, an examination of speculative attacks on ERM currencies suggests that these episodes are not distinguished from non-attack periods by different economic fundamentals (Jovanovic, 1989). The IMF can alleviate these market failures through both its surveillance and its lending activities. Through stronger surveillance, including encouragement of countries to make more timely data publicly available, the IMF would help to provide financial markets with the information that they need to form judgements concerning a country’s creditworthiness, alleviating the first problem described above. The IMF has also assisted members in their relations with commercial banks and official creditors through strengthened surveillance, thereby creating confidence in the countries’ adjustment policies and facilitating debt-restructuring agreements, and thus helping to correct the second market imperfection (see, for instance, Chapter 3 of Watson et al., 1986). However, it is through financing and the accompanying exercise of conditionality and provision of policy advice that the IMF thinks it can have the most impact. The commitment of resources in support of a programme signals the IMF’s confidence in a member’s policies. Such a signal is more likely to convince private lenders of the creditworthiness of a borrower than mere announcement of support. In this way, the IMF can catalyse other sources of financing and help extract commitments from other lenders. The IMF must also safeguard its resources and take account of the repayment of resources to the IMF in its conditionality. The exercise of conditionality makes the third problem cited above less severe. Thus, the IMF may be willing to provide support in some circumstances when private lenders are not, as the latter have less influence on the policies of borrowing countries. Moreover, the exercise of conditionality would be expected to

270


have externalities on other lenders, who would also reap the benefit of sounder policies in borrowing countries and their increased ability to repay. In this way, as well as by signalling confidence in a member’s policies, the IMF may help avoid bad equilibria—the fourth type of market imperfection. International public goods A second major area where, in the IMF’s view, the IMF can supplement private capital flows is the provision of international public goods. In particular, the IMF, by being prepared to make resources available in times of adverse shocks, can encourage countries to greater openness to trade and capital flows in the medium and long terms. Because openness is a public good (gains from trade accrue to all parties), it is in the global interest to encourage it (see Frenkel, Goldstein and Masson, 1988); indeed, the growth of trade and the elimination of foreign exchange restrictions are explicitly listed as purposes of the IMF in Article I. The IMF can encourage provision of these public goods by supplying a form of insurance against adverse shocks for countries (for example, Mexico, Argentina, the Czech Republic and Poland, among many others in recent years) that are liberalising or have liberalised their economies. This insurance involves access to IMF credit on reasonable terms should the balance of payments deteriorate. Private financial markets are unlikely to provide such insurance in adequate amounts; if they do, it would be at a risk premium that would be too high because lenders could not fully appropriate the benefits to global welfare. This would be a standard case of underprovision of public goods when left to private parties (Chapter 6 of Jha, 1998). The availability of IMF resources provides the assurance necessary for countries to engage in desirable policies that have favourable externalities and thus enhances the ability of other organisations (for example, the World Trade Organisation) to fulfil their mandates in this area. It is true that, just as in other insurance activities, such as life insurance, there is a potential moral hazard problem: the insured party may be induced to take on more exposure to adverse shocks than is desirable. If the IMF is likely to bail out countries when they run into trouble, won’t they undertake inadvisable policies, in particular, unnecessarily risky ones? In examining this moral hazard problem, three points have to be kept in mind (Diamond and Dybvig, 1983). First, as in the case of life insurance, some risk taking is desirable; the risk-minimising solution, which might prevail in the absence of a safety net, would be to close the economy to foreign trade and capital. Second, excessive risk-taking is limited by co-insurance, so that the insured party can expect to suffer a significant part of any loss. In particular, a country that chooses inappropriate policies that would lead to balance of payments difficulties typically pays an important cost in the subsequent adjustment process. In addition, IMF financing does not consist of grants, and the recipient of such financing must pay interest and promptly


271

repay the principal. Finally, as an extension of the analogy with life insurance, the IMF can be said to play the role of an investigating doctor through its exercise of surveillance and conditionality, which reduces the scope for excessive risk taking. Nevertheless, on balance, the IMF by essentially becoming an International Lender of Last Resort, may be encouraging moral hazard (Uzan, 1998). However, the insurance it offers may be critical in preventing the occurrence of bad equilibria in the Diamond and Dybvig (1983) framework. Another international public good is an exchange rate system that facilitates smooth adjustment and promotes stability; as Article I (iii) indicates, its promotion is one of the purposes of the IMF. Of course, the context of infrequently adjusted pegged rates has changed radically with the replacement of the Bretton Woods system by generalised floating. In large part, this change reflects the infeasibility of defending fixed exchange rates at a global level in the context of expanded capital flows and the preference of the largest industrial countries to devote their monetary policies to promoting domestic stability objectives (as described in Mussa, 1994). As these countries have been generally successful in recent years in delivering low inflation, it can be argued that it would be a mistake to distract their attention from keeping their own domestic houses in order (see Frenkel, Goldstein and Masson, 1991). These economies are providing an important stabilising influence on the world economy. Nevertheless, the IMF retains an important responsibility to help ensure the smooth functioning of the flexible exchange rate system with a minimum of volatility and misalignments. RECENT IMF PROGRAMMES—A QUICK OVERVIEW In Table 11.1 we present some evidence on IMF programmes in developing countries. What is most noteworthy about the experience of the developing countries is that the number and average size of short-term programmes (as a percentage of GDP) has shown little fluctuation since, except for going up marginally during the height of the debt crisis following the 1979 oil price hike. In sharp contrast since their introduction in the mid-1980s and particularly since the emergence of transition economies the number and average size of longer-term programmes has been going up fast. A relatively simple way of characterising the evolution of IMF interventions in recent times would be the following. In the aftermath of the second oil price shock (of 1979) soon after the first shock of 1973, developing countries were faced with mounting international payment obligations. Industrial countries reacted to this deterioration in their terms of trade with raises in the price of their own exportables. Hence, developing countries (particularly oil-importing developing countries) were faced with even worse external payments conditions as prices of their importables from both OPEC as well as the industrial countries increased, whereas the prices of their exportables

272


Table 11.1 IMF programmes in developing countries—approvals by time (1970–99) Number of programmes approved (average size of programme in million SDRs) [average size of programme relative to GDP]

were sluggish. As a consequence, several of them borrowed heavily from multinational banks. This increasing debt burden was further aggravated because of three factors: (i) dramatic increases in real interest rates consequent upon high global inflation; (ii) poor returns on money borrowed; and (iii) deterioration in the terms of trade. This burden soon erupted into a fullscale crisis over the global economy (during 1982) with the statement by Mexico that it was unable to meet its international debt obligations. This crisis of inability to repay international debt in time rapidly spread to many developing countries. The IMF’s response to this was the realisation that the heavily indebted poor countries (HIPC) would not be in a position to follow the strict discipline of its short-term programmes: Long-term Structural Adjustment Facility (SAF), Extended Structural Adjustment Facility (ESAF) and the like which allowed longer payback times. In doing so, it was sometimes alleged that the IMF was stepping on the shoes of the World Bank. After all, it was the World Bank whose job it was to provide long-term concessional aid to developing countries! In any event, particularly after the break up of the Soviet Union


273

and the emergence of the so-called transition economies, the role of longterm programmes of the IMF expanded considerably. With the onset of the Mexican crisis of 1994 and the East Asia crisis in 1997–98 the role of these long-term programmes only expanded. A major difference between the two sets of programmes from the IMF’s point of view is the application of different sets of conditionalities in the two cases. Conditionality is the link between the approval of a country’s request for funding or continuation of already approved funding and the adoption of specified policy programmes by the country to which the loans are being made. What are these conditionalities? In the case of short-term programmes, typically the IMF has asked for devaluation of the home currency as a precondition for granting loans. Then the country in question may be asked to commit itself to a path of credit and public deficit control. Continuation of the loan is contingent upon this country’s progress on this path. We have discussed the IMF’s rationale for such policies in the context of the IMF model in Chapter 7. IMF conditionalities pertain to crises that originate in or impact upon the current account of the balance of payments. The country in question, thus, requires only short-term corrective help. With the longer-term programmes, the IMF has in recent times included what it calls structural conditionalities. An example of this would be to meet conditions for fiscal sustainability. Another example could be that in case a country has opted for a currency board it must carry out some banking reforms in order to ensure financial stability. Similarly during financial crises reforms in the banking system may be necessary. Structural conditionalities are typically applied in situations where the crisis does not originate in the current account—in the case of transition countries and the countries affected by the currency crises. Typically structural conditionalities are seen to be the more obtrusive form of conditionalities. Hence, they have caused greater disenchantment. EVOLUTION OF IMF CONDITIONALITIES IMF conditionality has evolved substantially over time. Some element of conditionality was embedded in the Fund’s loan programmes as far back as the 1950s. However, with the two oil price shocks of the 1970s (in 1973 and 1979) the number of applications by developing countries for IMF funding increased dramatically. To the IMF’s credit it must be said that it responded to these criticisms. In 1979 it announced Guidelines on Conditionality,2 which emphasised the need to limit the criteria by which the performance of the country would be evaluated and attempted to make IMF programmes more sensitive to country-specific requirements, possibilities and policy constraints. However, before these Guidelines could be fully translated into policy initiatives, IMF programmes expanded considerably with the developing

274


country debt crisis of the 1980s. IMF programmes became more frequent and criticisms of the IMF conditionalities became ever more strident. The IMF did respond to some of these criticisms by announcing Structural Adjustment Programmes (SAF) and Extended Structural Adjustment Facility (ESAF). This was particularly in response to the fact that many very poor countries became the IMF’s clients as a consequence of the debt crisis. The explicit purpose of these facilities was ‘the alleviation of structural imbalances and rigidities’ in low-income developing countries, ‘many of which [had] suffered for many years from low rates of economic growth and declining per capita incomes’. In the 1990s IMF programmes expanded considerably in the aftermath of the breakup of the Soviet Union and the emergence of the ‘transition’ economies. Structural measures became much more important as a consequence. The East Asian crises of 1997 further strengthened this tendency. This was natural since the nature of the problems faced and the kind of conditionality used went through a sea change. The expansion of structural conditionality was also reflected in a sharp increase in the numbers of performance criteria, structural benchmarks and prior actions. There are several reasons for such changes. First, the Fund has been emphasising increasing the rate of growth through appropriate supply-side policies. For this it emphasised structural reforms. To support these the IMF set up the Extended Fund Facility (EFF). This facility was designed to address the situation of ‘an economy experiencing serious payments imbalance relating to structural maladjustments…or an economy characterized by slow growth and an inherently weak balance of payments position which prevents pursuit of an active development policy’. All major international organisations, the IMF included, emphasised economic growth as an objective in the 1980s, especially because of the poor growth record of the HIPC. At the same time, IMF lending to LDCs particularly through the Structural Adjustment Facility (SAF) and later the Enhanced Structural Adjustment Facility (ESAF) in the 1980s increased sharply. The IMF’s involvement with the transition economies grew sharply after the breakup of the Soviet Union. Both these groups of countries needed to grow fast and the IMF felt that the large-scale transfer of resources to these countries without removing some important distortions that remained in place would be futile. Hence the notion of conditionalities gained additional significance. In both these groups of countries, both external adjustment and growth were impeded by massive structural distortions, and it was believed—especially in light of early experience—that attempting to transfer resources to these countries without tackling these distortions would be largely futile. The IMF also emphasised concurrent reforms in the financial sectors of these countries. It was also felt that the success of other policy measures—particularly monetary and fiscal policy—is contingent upon appropriate structural conditions. With a poorly developed bond market, for instance, one cannot


275

expect open market operations to have much impact. Inadequate appreciation of this fact had resulted in macroeconomic policy having unintended outcomes. In Chapter 12 we will see that inadequate development of the fiscal infrastructure could lead to adverse impacts of apparently well-planned economic policy. The IMF has also initiated programmes of periodic reviews of their programmes. This reflects the need to address greater uncertainty about key macroeconomic relationships in a world of high capital mobility. This also emphasises the growing importance of structural reforms, which are often difficult to characterise quantitatively or even qualitatively with sufficient precision to be suitable as performance criteria. Furthermore, such reviews permit the overseeing of policies that have often unpredictable lags in design and implementation. However, although there have been substantial changes in the approach to monitoring, conditionality has become somewhat vague. Conditionality is typically that part of the IMF programme that is monitored as an indicator of the success of the policy. But, by the IMF’s own admission, in many cases, conditionality has been applied (or has been construed as applying) to reforms that are not really critical to the Fund’s decision on whether to continue its financing. Such ambiguity tends to make conditionalities appear to have broader scope than they actually do. Since this has a tendency to obfuscate the dimensions of the ensuing debate this is a problem the IMF is concerned about. Some others have argued that the IMF advice takes little cognisance of the policy constraints faced by developing countries. In particular, the IMF is alleged to underestimate the difficulties in achieving the policy consensus required for such reforms. Finally, there are concerns that overly pervasive conditionality may detract from implementation of desirable policies by undermining the authorities’ ownership of the programme. It is the authorities who must implement policies, and it is mainly they and their citizens who live with the consequences of either action or inaction. Policies are not likely to be implemented in a sustainable way unless the authorities accept them as their own and unless the policies command sufficiently broad support within the country. Conditionality that is too pervasive may galvanise domestic opposition to the programme as well as blurring the authorities’ focus on what is essential. Experience suggests that the link between conditionality and programme ownership is a complex one; in particular, there are cases in which high programme ownership has gone hand in hand with extensive conditionality, as a government wishes to use a Fund programme to strengthen its commitment to an ambitious set of domestically-owned reforms. Nonetheless, the strong likelihood that a more focused approach to conditionality will enhance country ownership is an important motivation for streamlining.

276


CURRENCY CRISES AND IMF PROGRAMMES Particularly for countries that have recently undergone a phase of economic crises there is considerable debate about the output effects of IMF stabilisation programmes. A number of prominent economists including the Nobel laureate Joseph Stiglitz have criticised the approach of the IMF and suggest that the IMF’s economic remedies often make things worse as they turn slowdowns into recessions and recessions into depressions. But there are voices in support of the IMF’s approach and no consensus has emerged among economists on this issue. In this context, Hutchinson (2002) has attempted a useful analysis of the effects of these stabilisation programmes in economies going through a crisis. He attempts to measure the output cost of participation in an IMF programme, and tries to assess whether there are feedback effects that make implementation of programmes especially problematic in the immediate aftermath or concurrent with an ongoing balance of payments/currency crisis. Hutchinson examines three specific questions: (i) Given that a country is already facing a severe currency crisis, does participation in an IMF-supported stabilisation programme tend to make real GDP growth weaker? (ii) Can one identify the channels (policy instruments) through which participation in IMF-supported programmes affect real GDP? (iii) How much of the downturn in East Asia following the 1997 currency crisis may be attributed to participation in IMF programmes? Table 11.2 lists the major currency crises that have occurred in the developing world. Hutchinson follows a panel data approach. To address the first question he factors out the effect of a currency crisis on real GDP, and asks whether there is a separate effect over and above that considering whether there is an additional effect arising from IMF-programme participation. Answering the second question necessitates identifying the policy channel or policy mechanism through which IMF-programme participation affects real GDP growth. Beyond providing countries with access to substantial lines of credit, IMF programmes are generally associated with conditions on the future conduct of fiscal, credit and other policies. Identifying the way IMF conditionality affects the formulation of policy in practice (ex post)—as opposed to the agreements themselves (ex ante)—is an important step in determining how participation in programmes might affect GDP. If the critics of the IMF are right— conditionality would lead to overly restrictive macroeconomic policies and poor output performance. This should then show up in the data. In answering the third question we need to compare the macroeconomic performance of the East Asian countries that faced the 1997 crisis and undertook IMF stabilisation programmes (Korea, Thailand, the Philippines and Indonesia) with those crisis countries that did not participate in the programme (Malaysia). To investigate these issues Hutchinson focuses on short-run IMF stabilisation programmes (Standby Agreements and Extended Fund Facilities)

Table 11.2 Occurrence of currency crises and IMF programme participation

Table 11.2 (continued)

IMF programmes: a. Standby and Extended Standby Agreements (SBA) b. Extended Fund Facility (EFF) c. Structural Adjustment Facility and Enhanced Structural Adjustment Facility (ESAF) d. Poverty Reduction and Growth Facility (PRGF) e. Supplemental Reserve Facility (SRF) f. Contingency and Compensatory Fund Facility (CCFF) Source: Hutchinson (2002). Reprinted with the permission of University of Chicago Press, Chicago, IL: University of Chicago Press.

280


that are explicitly focused on balance of payments adjustment, rather than programmes directed primarily toward structural reform and poverty reduction. He considers the broadest spectrum of developing and emergingmarket countries possible (Table 11.2). The estimation method used permits an examination of the question whether the adverse output effects of a currency crisis are made worse when the IMF steps in with a stabilisation package. Hutchinson focuses on short-run IMF stabilisation programmes (Standby Agreements and Extended Fund Facilities) that are explicitly focused on balance of payments adjustment, rather than programmes directed primarily toward structural reform and poverty reduction. He comes to the conclusion that the estimated cost of an IMF stabilisation programme, in terms of foregone output growth, is about 0.6–0.8 per cent during each year of programme participation. Further, currency crises have the additional effect of reducing output growth over a two-year period by about 2 per cent. This is the answer to Hutchinson’s first question. However, the answer to his second question is more ambiguous. Participation in an IMF-supported programme following a balance of payments or currency crisis does not appear to hurt or help output loss, even if the programme countries follow tight credit policies even after the IMF programme is over. So far as his third question is concerned, there is some evidence that the decline in GDP growth generally precedes the approval of an IMF programme and may not be attributable to programme participation per se. The fall in output that accompanied the 1997–98 currency crises were very large—much larger than those that would be predicted by historical evidence of earlier currency crises. Since Malaysia, which did not participate in an IMF programme, also suffered huge output losses, it seems to be the case that merely participation in an IMF programme does not exacerbate output losses during a currency crisis. Hence, the presumption that participation in an IMF programme would exacerbate output losses associated with a currency crisis seems to be misplaced. CONCLUSIONS There is considerable merit in the argument that countries resort to IMF programmes voluntarily. Hence they should be prepared for the output consequences of such participation. However, some of the same arguments that make the presence of an institution such as the IMF a necessity make it difficult for a perfectly satisfactory solution to emerge. If international credit markets are imperfect and the IMF steps in to fill this gap, it does not follow that the IMF’s actions will facilitate the emergence of a solution consistent with perfect credit markets. As an international lender of last resort and a global regulator of international finance, the IMF certainly plays an important role. However, as the global financial system throws up new challenges, the design of IMF programmes should undergo a change as the IMF responds to


281

changing and volatile economic conditions. The analysis in this chapter suggests that there is considerable room for such improvement in design. In Chapter 17 we will understand that the institutional structure of the IMF needs a considerable revamp. DISCUSSION QUESTIONS 1

One of the official histories of the IMF by de Vries (1987) noted that there is an understandable perception of asymmetry between developing and industrial country members in that the conditionality applied to the use of the Fund’s resources has significantly affected developing members, while surveillance under Article IV…seems to have had little practical effect on the large industrial members.

2

Why do you think this has been the case? Do you think this arrangement is efficient? Equitable? The IMF’s poverty reduction and growth strategy is outlined in a document available on the internet at http://www.imf.org/external/np/ exr/facts/prgf.htm. This seems to suggest that integrating macroeconomic policies with a poverty reduction strategy is an important precondition for rapid gains to the poor. Read this document and comment on the IMF’s view of the impact of short-run and medium-term macroeconomic adjustment on poverty.

IMF conditionalities Although the IMF’s assistance is usually referred to as ‘lending’ or ‘loans’, a member country actually ‘purchases’ SDRs or other currencies from the Fund in exchange for its own currency and agrees to ‘repurchase’ (buy back) its own currency at a later date. Because the member is charged for this transaction, the purchase or ‘drawing’ looks like a loan. The loans from the Fund’s General Resources Account can be used for any purpose relating to general balance of payments support, such as restoring reserves in the country’s central bank or selling the acquired currency in the foreign exchange markets to stabilise exchange rates. Member countries have automatic access to a portion of the Fund’s resources, called the ‘reserve tranche’. Member countries seeking an IMF loan beyond the reserve tranche must convince the Fund they have a balance of payments need. As the amount of the loan increases beyond the reserve tranche (technically, moving into the ‘first credit’ followed by ‘upper credit’ tranches), the IMF imposes conditions on the use of the funds—known around the world as ‘conditionality’. Conditionality refers to the explicit commitment by the member country to implement remedial measures in return for IMF assistance.

282


Conditions may range from general commitments to cooperate with the IMF with respect to establishing domestic economic policy to presenting the Fund with specific measures the country intends to implement at specific points in time. Those measures typically have related to the domestic money supply, budget deficits, international reserves, external debt, exchange rates and interest rates. However, the IMF’s conditionality has spread into other areas. Drawings in the upper credit tranches typically are associated with 12–18-month ‘standby arrangements’ (similar to pre-approved lines of credit) or ‘extended arrangements’ under the Extended Fund Facility, which provide assistance for longer periods of time (3–4 years) and for larger amounts than standby arrangements in order to address structural problems in the economy. Upper credit tranche drawings are usually made in instalments and released only after economic performance targets have been met. The country’s specific plans are set forth in a ‘Letter of Intent’ presented to the IMF. Although formally the member country formulates and ‘presents’ the Letter of Intent to the Fund, practically speaking the Fund wields a great deal of influence regarding the content of the Letter. In addition to drawings from the General Resources Account, the Fund has established a number of other lending ‘facilities’ designed for specific problems, such as the Enhanced Structural Adjustment Facility (concessional loans for poor countries experiencing protracted balance of payments problems—now called the Poverty Reduction and Growth Facility), the Compensatory and Contingency Financing Facility (shortfall in export earnings or rise in costs of cereal imports), the Systemic Transformation Facility (balance of payments difficulties experienced by countries changing from non-market to market economies), and the Supplemental Reserve Facility (exceptional balance of payments difficulties because of sudden and disruptive loss of market confidence). Except for the Supplemental Reserve Facility, the IMF uses quotas (discussed in Chapter 17) to determine how much a member country may borrow under a facility. The Fund’s Executive Board reviews the access limits periodically, taking into account payment problems of its members, the need to safeguard IMF resources, and developments in the Fund’s liquidity. The IMF also engages in surveillance over the exchange rates of member countries. After the collapse of the Bretton Woods System based on fixed exchange rates, the IMF’s charter was amended to allow member countries to choose their own exchange rate systems. However, under Article IV of the charter, members agreed to collaborate with the Fund and other members to assure orderly exchange arrangements and to promote a stable system of exchange rates. Moreover, Article IV gives the IMF broad powers of surveillance over exchange rate policies of members, and gives it the authority to adopt principles that will guide members


283

with respect to those policies. Surveillance is carried out in part through annual consultations with member countries, known as ‘Article IV consultations’. During these meetings with member country officials, IMF staff members analyse the country’s economic developments and policies. In addition to the traditional economic variables we mentioned above in connection with conditionality, the Fund recently has begun to look at social, environmental, industrial, labour market and governance issues that influence economic management of the country. The IMF has encouraged member countries to make their domestic currencies convertible on the current and capital accounts—although this enthusiasm has mellowed after the East Asian and subsequent currency crises. The IMF pursues its alleged primary goal—to facilitate the expansion and balanced growth of trade—by establishing a multilateral system of payments for current account—e.g. trade transactions. Convertibility is important for international trade. Under the Bretton Woods System, the convertible dollar was the key currency. Trade was conducted in dollars; importers and exporters typically held dollar accounts for their business transactions. By accepting the obligations of Article VIII of the IMF’s charter, member countries agree not to impose restrictions on payments and transfers relating to the current account, and they also agree to refrain from engaging in discriminatory currency arrangements or multiple currency arrangements without the approval of the IMF. Although historically member countries have been slow to accept Article VIII obligations, the number of countries adhering to these rules is rising more rapidly today (in 2001 it was approximately 147). The IMF has several ways of assuring compliance with its rules and policies. Technically, letters of intent, standby arrangements and other IMF loans are not contracts or any other type of legal obligation. But, there are other ways the IMF can apply pressure on countries to comply. First, conditionality enables the IMF to withhold funds if a member country does not comply with the conditions of the loan. Second, members know that by violating the rules and policies of the IMF they may be shut out of the international capital markets. Third, the Fund can prohibit a member country from using the General Resources Account. Fourth, the IMF can expel a country (as it did to Czechoslovakia in 1954). Finally, through collaboration and consultation with member countries, the Fund tries to persuade and cajole countries into complying with rules and policies. The IMF Provides Countries with Assistance Regarding Technical Issues. Another function of the IMF is to provide technical assistance. The IMF began to give countries technical assistance in 1964 when ex-colonies wanted help in setting up their own central banks and ministries of finance. The IMF also provides assistance to countries to understand and adopt its views on fiscal policy, monetary policy,

284


banking, institution-building, financial legislation and statistics (which helps IMF surveillance). When the IMF is providing assistance in areas not directly related to economics, it still links the assistance to economics. For instance, the IMF provides technical assistance in the area of law-financial legislation, but the emphasis is on enacting laws that support a free market.

NOTES 1

2

Part of the criticism against the IMF is that it contributes to moral hazard by creating the expectation of bailouts (implicit debt guarantees) whenever countries face balance of payments problems. Empirical evidence on this point is mixed. For example, Dreher and Vaubel (2001) find support for moral hazard associated with IMF programmes, while Lane and Phillips (2000) do not. See Willett (2000) for a recent review and evaluation of the literature on the debate surrounding the role of the IMF. There is a large literature reviewing the effects of IMFsupported stabilisation programmes. See, for example, Bird, 1995; Edwards, 1989a; Gylafson, 1987; Pastor, 1987; Santaella, 1996; and Willett, 2001. For updates on IMF conditionalities see the IMF website www.imf.org.

12


INTRODUCTION In this chapter we purport to examine the macroeconomic effects of fiscal policy, particularly deficits, in developing countries. The roles of the fiscal authority in developed and developing countries vis-à-vis developed countries are markedly different. In both developed and developing countries there is a concern for raising living standards over time, but this need is much more pronounced in developing countries, given the extent and depth of poverty in these countries. In the relative absence or perpetual weakness of institutions to mobilise and direct savings, the role of the state is crucial in harnessing the resources for development. Since the regulatory apparatus is weak and market signals imperfect, the state has an important role to play in allocating investment funds. Further, with widespread poverty, there is the expectation that fiscal expenditures would play a major role in anti-poverty programmes. Pressures for populism through price controls and the like are considerable. At the same time, and for some of these same reasons, the state in developing countries is handicapped in its ability to play an activist role. First, the state in a developing country is a weaker entity politically than in most developed countries. This means that there is often very little consensus on the contours of a tax and expenditure programme (see Heady, 2000). Second, the resources available with the government are meagre since tax bases are small and tax administration weak. Much of tax revenue comes from inefficient and distortionary indirect taxes such as excise duties. International trade is heavily taxed. Effective personal income taxes are low and easily evaded and corporate taxes are high. Even so, expenditures routinely, and ever increasingly, outpace revenues. With poor credit and bond markets and fiscal expenditures that are inflexible in the downward direction, some of the financing of the resultant deficit spills over onto the external sector and the central bank. Even within the developing country group there is considerable heterogeneity in experience with respect to the fiscal deficit. The differences

286


are more pronounced between the middle and low-income country categories but they exist even within the low-income category of countries. For example, low-income countries differ sharply in regard to the depths of their financial markets. Thus in 1999 in Burkina Faso net foreign assets as a percentage of broad money were -1.9 per cent. The corresponding figure for India1 was 14.3 per cent. Hence there are sharp differences within countries in the developing country group with respect to the options available for public borrowing in the event of there being a large fiscal deficit. Indeed the poorest among the LDCs are caught in an insidious resource trap (UNCTAD, 2000). The relation between per capita income and savings appears no different in these countries than in the presently developed nations. However, because of low per capita incomes, savings are low and because of this economic growth is low. In addition, as UNCTAD (2000) estimates, external shocks have a far more serious effect in the least developed countries than other developing countries. To quote this report, ‘The average least developed country economy has, since the 1970s, been exposed to adverse external trade shocks with an impact, in the worst years, more or less double the average of other developing countries.’ The fact that the least developed countries can spare far fewer resources to combat the effects of these external shocks, then, exacerbates this problem. External finance is the obvious way to get around this sharp resource crunch. However, such supplies are meagre. Large, stable market economies such as India and South Africa attract considerable capital inflows whereas most poor economies of the sub-Saharan region get poor inflows. Thus in 1997 South Africa had FDI flows of US$27,483 million and portfolio investment assets of US$7,817.77 million2 whereas the corresponding figures for Botswana were US$404 million and US$204 million. Both FDI as well as portfolio flows are poor. FDI flows are concentrated largely in resource extraction.3 The reasons for such poor flows are not hard to discover. These are related to ‘costs of asset development, risks which are rooted in the vulnerability of the least developed countries to shocks, lack of business support services, weak physical, social and administrative infrastructure, and the small scale of projects’ (UNCTAD, 2000). With rapidly diminishing official aid and poor private equity flows, external financing of the fiscal deficit in the poorest countries has to rely increasingly on private loans. These are available at increasingly difficult terms, as Harberger (1985) has noted, as the domestic resource cost of servicing these goes up with additional borrowing. Furthermore, typically, this resource cost is underestimated. Other reasons for differences across countries include continuity and stability of policy regimes. Thus Zambia, which has had a history of policy reversals, would be associated with higher risk than Botswana and Mauritius, which have had credible and stable policy regimes. The costs of borrowing abroad will typically be higher for countries with frequent reversals of policy stance as the risks associated with lending will go up sharply.


287

A consequence of these factors for financing the fiscal deficit is that many of these countries have to rely, to a considerable extent, on non-bond (monetary) financing of the deficit. When this occurs, the distinction between fiscal policy and the monetary base of the central bank is blurred and the independence of monetary from fiscal policy is compromised. In this chapter we will address several issues. First, we review the notion of the fiscal deficit in the particular context of developing countries. Then we articulate the notion of sustainable fiscal deficits and sustainable external debt. We then turn our attention to the notion of sustainable ‘twin deficits’ and report some results for these on a cross-section of low and medium income countries for the period since 1950. We also consider the implications of collecting additional revenue through inflation tax on the debt. Subsequently, we consider some problems associated with international capital flows and indicate the difficulty of pursuing stabilisation policies under these conditions and the role that fiscal policy can play to ameliorate this difficulty. THE FISCAL DEFICIT: WHAT DOES IT MEASURE? The conventional measure of the fiscal deficit as the difference between total government expenditure and current government revenue while being clear as an accounting concept, is not above controversy as an economic entity. Tanzi (1993), for example, mentions the following three difficulties: (i) the conventional measure of the deficit fails to recognise that different tax and expenditure categories have different types of effects on aggregate demand. For example, an excess of expenditure on the infrastructure creates productive capacity and will have a different impact than an excess of expenditure due to consumption subsidies.4 (ii) A second problem arises because tax revenues are not exogenous of expenditures. The level of public expenditures determines national income, which then determines tax revenue, at least in part. It is in this context that during the Kennedy-Johnson presidencies in the US—a period of full employment—the notion of a full employment budget surplus was defined and used. In the context of developing countries, such a notion would be suspect since the binding constraint on output in their cases is not the supply of labour. A more relevant constraint would be the availability of credit in terms of hard currencies.5 (iii) Finally, there is the problem of sources of financing the deficit. In developing countries several sources of financing have been used: e.g., central bank financing, commercial bank financing, domestic sale of government bonds to cover the deficit and foreign financing. Each of these has different macroeconomic consequences. Central bank financing raises the monetary base and the money supply, thereby blurring the distinction between monetary and fiscal policies. Foreign financing will raise the cost of servicing external debt whereas domestic bond issues will raise interest rates. An additional difficulty is that some sources of finance are available only under certain circumstances. For example, a country

288


with a thin bond market (which is the case in almost all of sub-Saharan Africa except South Africa) can hardly afford to issue bonds to cover the fiscal deficit and may have to rely on central bank financing or some such measure. A country with large external debt would, in all probability, be able to finance its deficit externally only by borrowing short term at high rates of interest. This would make it difficult to finance the external debt and may put pressure on the currency. Some authors such as Buiter (1985, 1993) have gone much further ahead and argued that even cyclically adjusted and inflation adjusted measures of the deficit are only imprecise indicators of the true deficit. For example, the capital gains/losses on government assets and liabilities are not included in the conventional flow of funds accounts. Examples of these would include changes in relative prices (say changes in mineral prices) and changes in the real value of nominal debt during an episode of inflation.6 Particularly in those developing countries where public investment has played a significant role in the economy, a distinction is made between revenue or current deficit and capital deficit. The former is the deficit on expenses of a recurrent nature after netting out investment expenditure. Surely, if a country is running a large and growing deficit on such current transactions it is a reason for worry. However, it should be pointed out that the distinction between capital and current expenditures is often an artificial one, for example, in aid dependent economies, large amounts of aid-financed current expenditure connected with aid-financed projects are placed in the capital expenditure category—the case of Mozambique, for example. It would be meaningful only if it was clear that all capital expenditures were productive in nature. Although capital expenditures are associated with capacity building activity, it might be the case that some such expenditures are wasteful in nature. This would typically be the case if the project in question has not been evaluated carefully and/or involves equipment that is highly capital intensive or has an unduly import intensive or is, in some other way, inappropriate for the economy. For example, a substantial part of Nigeria’s public investments have low social rates of return, reflecting inappropriate technology choices and long-standing overvaluation of the currency that cheapened the cost of capital equipment. Furthermore, this current or revenue deficit says nothing about the impact of the deficit on the balance of payments or on economic growth. There are some other problems associated with the measurement of deficits that are worth mentioning. First, is the problem of arrears. This becomes particularly relevant in the case of repayment of foreign debt. If, for example, the interest payment on the foreign debt is rescheduled—which is the case in many heavily indebted poor countries (HIPCs) such as Tanzania—should we say that the deficit has gone down? Similarly, if the government delays some payments (often the public wage bill) whereas it takes in all its revenues does fiscal deficit go down? These questions become particularly relevant


289

during a period of high inflation when delaying payments denominated in nominal terms can have a significant impact on the real value of such payments. For example, Angola’s government receives large resource revenues from oil but in the recent past it has nevertheless accumulated large wage arrears, the real value of which would have declined rapidly with hyperinflation (see Aguilar, 2000). A further important problem arises when the fiscal deficit reported is only that of the central government. In countries with different levels of government this would be an inappropriate indicator of the deficit. Even if the deficits of state and lower levels of government are included, there may be other government agencies that are running a deficit, which does not get reflected, in the measured fiscal deficit. Examples would include the deficits of central banks in some Latin American countries, the deficits of local (particularly municipal) governments in India and the so-called ‘oil pool’ deficit in India where excesses of payments for petroleum imports over what is collected from consumers is recorded. The oil pool deficit, for example, can become large during a period of rises in the international prices of petroleum products and difficulties in raising the domestic prices of such products.7 In the Indian8 case, for example, part of the burden of this adjustment has been shifted to the future by issuing ‘oil bonds’. Such transactions do not get reflected in the fiscal deficit. All this then tends to make the fiscal deficit a not entirely satisfactory measure of the excess of government expenditure over revenues. The difficulties in measuring and interpreting the deficit notwithstanding, it is useful to understand whether the underlying fiscal stance is sustainable. The literature on this issue is growing rapidly. In the next section I develop the extant notion of internal debt sustainability and apply it to select developing countries for which continuous data are available. Subsequently the notion of external debt sustainability is discussed. SUSTAINABILITY OF THE DOMESTIC FISCAL DEBT Rapid accumulation of domestic debt can lead to severe macroeconomic problems, and can impede control of the fiscal deficit itself. To take only one example, Zimbabwe’s fiscal deficit is estimated to be close to 20 per cent of GDP in 2000. ‘Public debt is rising rapidly, with new debt being issued to meet interest payments (the so-called “Ponzi” game). Interest payments on domestic public debt are expected to exceed 50 per cent of total government revenues by end-2000, thereby squeezing development and social spending’ (EIU, 2000). This section develops the simple analytics of sustainability for the domestic as well as the external deficit and applies it to a spectrum of developing countries.

290


The government intertemporal budget constraint The most straightforward way to assess the sustainability of a public debt situation is to start from the governmental intertemporal budget constraint. This is written in nominal terms as: (12.1)

where Gt is the value of government expenditures (purchases of goods and services plus transfer payments); Bt is the government debt at the end of period t, Tt is the government’s tax revenue and rt is the one-period rate of interest payable on the government debt. Equation (12.1) states that in the absence of money finance, the government budget deficit must be financed by new debt creation.9 Hence, expressing (12.1) in terms of ratios to gross GDP we will have: (12.2)

where the lower case letters denote the ratio of the corresponding upper case variables to nominal GDP: and with P and Y being the price level and real GDP respectively. is the rate of inflation and is the rate of growth of real GDP. In the derivation of (12.2) we have used the relation that: where is the primary deficit expressed as a percentage of GDP. We have the following cases: Case 1: In this case in (12.2) the debt ratio will stabilise and the economy will remain solvent if:

If the initial debt to GDP ratio (b0) is strictly positive, this requires two conditions: for all t so that the debt ratio stabilises rather than explodes. This is the so-called sustainability condition and makes any stable path of the primary deficit consistent with a stable public debt to GDP ratio. In addition we have condition (b) that

on average, if not in every period, so that the debt burden is ultimately liquidated. These two conditions are necessary and sufficient and ensure that the debt, no matter how large, can be paid off through tax increases or expenditure cuts or inflation. Thus the government is solvent. The steady state (finite) value of the debt-GDP ratio is


291

(12.3)

Equation (12.3) emphasises a strong link between the government’s indebtedness and its primary deficit. Case 2: In this case the debt is unsustainable and the debt stock will become infinite no matter what sequence of primary deficits are chosen unless the debt stock itself can be offset by matching the sequence of increasing but discounted primary surplus in the future. To consider sustainability further here transform (12.2) to get: (12.4)

Where we have used the fact that:

is the real interest rate minus the rate of growth of real GDP. Equation (12.4) will always hold ex post facto. Looking forward we can write the identity in (12.4) for time period t+1 as: (12.5)

where bt is known in period t. For this one period constraint to hold in expectational terms, this must equal the expected discounted net debt-GDP ratio in period t+1 conditional on information at time t. For fiscal policy to be sustainable for one time period (12.5) must hold. Writing the budget constraint of (12.5) for subsequent time periods t+2, t+3 etc. and solving forward we get (12.6)

It is apparent that

is a time-varying real discount factor adjusted for the growth of real GDP with θ >0. A necessary and sufficient condition for sustainability is that as the discounted value of the expected debt-GDP ratio converges to zero. This is a transversality condition and can be expressed as: (12.7)

Equation (12.7) implies that a government is solvent if the transversality condition guarantees the non-explosiveness of the public debt and when no

292


Ponzi games are allowed, i.e., no new debt is issued by the government to meet interest payments. Hence it follows that the current debt is offset by the sum of the current and expected future discounted surpluses, implying that the budget constraint holds in present value terms with: (12.8)

The critical value of the debt-GDP ratio Given (12.8) and using as a definition of the maximum level of the government’s primary surplus we can determine the critical value of the public debt ratio (bc), which will satisfy the sustainability condition: (12.9)

We can also determine the necessary primary surplus, given the initial debt ratio, b0, the real interest rate and the growth rate of real GDP, to stabilise the future debt-GDP ratio. When we can use (12.2) to define the finite value (b0) to which b converges as: (12.10)

The gap between the stabilising primary surplus (z**) and the actual primary surplus may be used as a sustainability indicator. This indicator gives the magnitude by which either revenue must be increased or expenditure must be cut relative to income to stop the debt ratio from growing. Equation (12.10) seems helpful when assessing empirically the fiscal policies and the debt situation for developing countries. For example, it would be useful to compare this with the debt ceiling the IMF may have prescribed for these countries at the times of their agreeing to stabilisation programmes.10 From the above analysis it is clear that sustainability of the public debt is essentially an intertemporal question. In particular, every temporary deficit can be sustainable so long as it is matched by an adequate future surplus. Most empirical tests on sustainability apply time series methods and ask whether the observed characteristics of debt-related variables satisfy the solvency condition in (12.7). This solvency condition can be tested in a variety of ways depending on the processes postulated for the primary deficit and the real interest rate adjusted for output growth (θt). Hamilton and Flavin (1986) and Trehan and Walsh (1991), among others, examine the case where is strictly exogenous and θt is constant. Wilcox (1989) considers the case with exogenous but variable θt Uctum and Wickens (1997) consider the case where θt is stochastic and could be exogenous or endogenous. For the sake of simplicity, we will assume that the real interest rate adjusted for output growth, θt, is constant with an unconditional mean.


293

To proceed further now take the first difference of (12.6), substitute for ∆bt using (12.4) and simplify to get:

(12.11)

where with defined as total government expenditure inclusive of expenditure on goods and services, transfer payments and interest on the debt. If the government satisfies its intertemporal budget constraint then the expected limit term in (12.11) is zero so that the sum of the current budget surplus and the expected present discounted value of future surplus will equal the amount needed to repay the principal and the interest on the initial debt. When this condition holds, it can be said that the current expected paths of government spending and taxation are sustainable. As Papadopoulos and Sidirpoulos (1999) demonstrate, if the limit terms on the right-hand side of (12.11) are zero, then a certain cointegrating relationship emerges. Hence cointegration is a necessary condition for the intertemporal budget constraint to hold. To see this assume that dt+s and τt+s follow random walks with drift. Thus these variables follow these time series processes: (12.12) (12.13)

where αd and ατ are constants and vd and vτ are zero-mean stationary processes. Hence (12.11) can be rewritten as: (12.14)

with

Given that dt and τt are I(1) and given that (12.12) and (12.13) imply stationarity on the right-hand side of (12.11), the left-hand side of (12.11) must also be stationary, for which a necessary condition is that (12.14) be stationary, which will be the case when dt and τt are cointegrated. Thus a test for sustainability of the debt would check for the cointegration of

294


these two variables if they are I(1). This cointegrating regression would take the form: (12.15)

Formally, then, if dt and τt are I(1), the null hypothesis is that dt and τt are cointegrated and that If this null hypothesis is not rejected then the public debt is sustainable. also According to Hakkio and Rush (1991) the condition guarantees sustainability. To see this, substitute for τt in (12.2) to get with By iterating forward we will get:

(12.16)

Using (12.16) the limit term

in (12.14) must equal

Using this relation and a similar expression for

it can be shown that the expected limit terms on the right-hand side of (12.14) are equal to zero so long as However, although the limit terms equal zero, the limit of the undiscounted value of b equals infinity when Thus this notion of sustainability is not as unambiguous as that which requires that Tests for sustainability of the fiscal deficit were conducted for the sample of low- and middle-income countries. The data set used was that provided by the IFS CD ROM. Annual data for the period 1950–99 was considered. Time series analysis of the sort considered in this chapter requires long data sets. For several of the countries such data were not available. In Table 12.1 we report results on the internal deficit sustainability tests for all countries for which a reasonably long data set was available.11 The results in Table 12.1 are quite striking. In almost all countries both government expenditure as well as government revenue are non-stationary. In the case of Sri Lanka both expenditure as well as revenue are stationary whereas in the case of Zambia, revenue is stationary whereas expenditure is


295

Table 12.1 Sustainability of internal deficit of low- and middle-income countries

not. Thus Sri Lanka appears to have a stable fiscal situation whereas the Zambian deficit situation is worrisome. In the rest of the countries only in the case of Paraguay and South Africa is the deficit sustainability condition satisfied. In all the other countries, either government expenditures and revenues are not cointegrated or, if cointegrated, the sustainability restriction does not hold. Thus, the fiscal deficit in these countries is not sustainable in the long run. SUSTAINABILITY OF THE EXTERNAL DEBT In line with the arguments developing the notion of a sustainable internal deficit, a case can be made to ascertain the sustainability of the external deficit (typically the current account balance or trade balance) of a country. Just as a government cannot borrow in the domestic market indefinitely to finance its budgetary deficit, it cannot borrow indefinitely in global capital

296


markets to finance its trade account deficit. This notion can be formalised in a manner similar to that expressed above. It should be noted that it is only rather recently that the notion that large current account deficits may cause problems has achieved acceptance. As late as 1994, Max Corden espoused the virtues of the Lawson doctrine.12 Thus Corden (1994) wrote: The current account is the net result of savings and investment, private and public. Decentralized optimal decisions on private saving and investment will lead to a net balance—the current account— which will also be optimal. There is no reason to presume that governments or outside observers know better how much private agents should invest and save than these private agents themselves, unless there are government-imposed distortions. It follows that an increase in a current account deficit that results from a shift in private sector behavior should not be a matter of concern at all. On the other hand, the public budget balance is a matter of public policy concern and the focus should be on this. Notwithstanding the Lawson doctrine large current account imbalances were associated with currency attacks in Chile (in the early 1980s), in the UK and Nordic countries (late 1980s); in Mexico and Argentina (mid-1990s) and in several Asian countries and Russia in 1997 and thereafter.13 Hence, there seem to be some problems with the Lawson doctrine. The literature lists the following five reasons for the failure of this doctrine: 1

2

3

When there is Ricardian equivalence,14 a public sector deficit will be interpreted by rational, forward-looking agents as implying higher taxes in the future. In response, current consumption will fall to pay for these higher taxes. This will then impact on the current account deficit. By assuming that the current account deficit will be impervious to changes in the fiscal deficit, the advocates of the Lawson doctrine assume that Ricardian Equivalence does not hold at all. Typically in the case of developing countries, many private sector (foreign) liabilities are contingent liabilities, which, in a crisis, can be changed into public liabilities under pressure from external creditors. This would then have impacts on the fiscal deficit. This has happened explicitly in the case of private banking liabilities in the case of the Asian countries and implicitly in the case of so-called counter-guarantees given to independent producers of electricity in India. As Kaminsky and Reinhart (1996) observe, the distinction between private liability and public liability seems to get blurred when the financial institution concerned is ‘too large to fail’. This follows from Harberger (1985). He argues that there is an externality in that as developing countries (particularly those with poor credit rating)


4

5

297

borrow more, they face an upward sloping schedule for foreign credit. Furthermore, as McKinnon and Pil (1995) argue, developing countries may have overly optimistic estimates of the increase in their permanent income in response to a policy shift. This over-optimism may lead to higher foreign borrowings, which may exacerbate current account and fiscal deficits. As the current account deficit worsens there would arise the need to attract more capital inflows. This would then lead to an appreciation of the real exchange rate. In a globalising world in which export promotion is an integral part of the development strategy, such appreciation would lead to a drop in growth rates and, therefore, potential tax revenues. Thus Agosin (1994) finds large swings in the real exchange rate because of temporary capital flows to significantly depress machinery and equipment investment, and thus long-run growth. This then translates into lower tax revenues. Finally, markets typically look at a country’s total debt and just its internal public debt. Once the current account deficit exceeds some level, at given exchange rates, it becomes ever more attractive, from the perspective of an individual borrower to borrow abroad. This might lead to a speculative bubble.

We now discuss the standard approach to determining the sustainability of the external deficit. Let Ft be the foreign liabilities of a country. This is typically defined as external debt less foreign assets including international reserves denominated in real foreign currency terms. Further, let TBt denote the real trade balance expressed in domestic currency. Foreign liabilities (expressed in domestic currency) will then evolve as: (12.17)

where et is the reciprocal of the average real exchange rate, and is the world interest rate. Clearly a positive trade balance would reduce this country’s external indebtedness whereas an increase in the world interest rate would worsen it. To simplify the algebra define as To obtain this country’s external constraint along the lines of Chalk and Hemming (2000) and Hakkio and Rush (1991) we solve (12.17) forward to get (12.18)

where

298


Equation (12.18) is to be interpreted as this country’s intertemporal external sustainability condition. Meaningful statistical tests of sustainability have to be derived from equation (12.18). For external sustainability we must rule out Ponzi games with external debt. In other words, net foreign liabilities cannot grow faster than the interest rate, i.e.

is a necessary condition for external sustainability. This requires that (12.19)

Equation (12.19) makes explicit the fact that external debt is sustainable only insofar as it can be financed by future trade surpluses. In the extant literature this requirement has been interpreted to imply stationarity of the current account deficit or the cointegration of exports and imports inclusive of net interest payments, if exports and imports are I(1). A simple way of interpreting the sustainability of equation (12.19) is to divide both sides by output to get (12.20)

where lower case letters denote values as proportions of GDP and qt is the real appreciation of the domestic currency. Net foreign liabilities as a ratio of GDP are reduced by a positive trade balance, an appreciating currency or by faster economic growth. Now, if (12.21)

then, when

net foreign liabilities will increase relative to output over time. This can be regarded as unsustainable. THE TWIN DEFICITS It is straightforward to connect sustainability on the domestic fiscal and the external side—the so-called twin deficits. To see this write the national income identity as: where Yt is national income, Ct is private consumption, I is investment, and X and M are exports and imports respectively. Defining private saving as income minus consumption we have: 12.22)


299

where D is the government budgetary deficit. Using covered interest parity t and summing (12.22) over all time periods in net present value terms we have

(12.23)

where Using the internal and external sustainability conditions, write equation (12.23) as

(12.24)

With both fiscal and external sustainability the ‘lim’ terms in equation (12.24) will disappear and we will have: (12.25)

If there is domestic fiscal sustainability but the external deficit is unsustainable then the R* term gets added to equation (12.25). Thus: (12.26)

If the external deficit is sustainable but the domestic deficit is not, then we will have (12.27)

Equations (12.26) and (12.27) make it explicit that both the internal and external deficits figure in the government budget constraint. We report some results on sustainability of the external deficit in Table 12.2. Time series properties of exports, imports and the trade balance are reported. Some countries like Nepal for which the internal sustainability result could be worked out do not have long enough external data series so that they are not reported. Even middle income countries such as South Africa and low-income with well developed markets such as India, do not perform well in the external sustainability test. Paraguay seems to satisfy external as well as internal deficit sustainability. In addition, Ecuador, Egypt, Ghana, Iran, Thailand and Tunisia have stationary trade balances so their external deficits would appear sustainable on this count. However, in the case of Tunisia and Iran exports and imports are not cointegrated so external

300


Table 12.2 Sustainability of external deficit of low- and middle-income countries

sustainability becomes suspect. Even in the case of Sri Lanka the coefficient on imports is much higher than that on exports. In the case of Zambia this coefficient is fifty times higher.15 It is quite apparent, then, that developing countries have considerable difficulties in meeting internal and external deficit sustainability conditions. The fact that external sustainability conditions16 are hard to meet would imply the need for continual capital inflow in order to keep the balance of payments in equilibrium. This would necessitate the maintenance of a substantial rate of return wedge between domestic and foreign rates of return. In particular, this would translate into substantially higher domestic rates of interest as compared to global interest rates. This acts as a drag on higher growth and makes the problem of debt servicing harder, which, in turn, exacerbates the problem of internal fiscal deficit.17


301

The higher internal fiscal deficit in many of these countries also contributes towards higher interest rates and the crowding out of private investment. In addition, in several of these countries the fact that the domestic bond market is rather thin implies that the proportion of the deficit that is covered by money or non-bond finance as opposed to bond finance remains disturbingly high. This is reported in Table 12.3 for select countries for the 1990s. This table compares non-bond financing in the developing countries used in this study with that in a few OECD countries. To the extent that non-bond finance is relied upon, there exists an automatic link between the fiscal deficit and the credit side of the central bank’s balance sheet. The government issues bonds to be ‘bought’ by a captive central bank. This changes the monetary base and, hence, the money supply in the economy. The independence of monetary and fiscal policy is compromised and the ability of the central bank to pursue stabilisation policies18 is reduced.19 Some have argued that, if this effect was quantitatively strong, there would be a link between the fiscal deficit and inflation—particularly if developing countries wish to use seignorage revenues to close the budgetary gap.20 However, in developing countries this association is weak. Thus, Easterly and Schmidt-Hebbel (1993) and de Hann and Zelhorst (1990) find a positive correlation between inflation and the fiscal deficit in developing countries only when the inflation rate is high and there is a clear seignorage motive to get additional revenue from money creation. However, Buffie (1999) argues that this result can be ascribed to the behaviour of the public sector wage cycle and that the relation between the fiscal deficit and inflation remains intact once this is factored out. Buffie’s argument is that once public expenditure is restrained, perhaps as part of an IMF stabilisation programme, there is the clear expectation that any cut in the real wage in the public sector that this involves would not be expected to last long. It would be expected that once the strictures of the programme are lifted, the real wage would climb back up. The disinflationary programme, therefore, lacks credibility. Buffie considers two possible cases: (i) in which the low wage phase is followed by a high wage phase of equal length so as to leave the average wage remains unchanged over the wage cycle, (ii) The low real wage phase is followed by a return of the real wage to its prestabilisation level. In the first case, since the market expects the real wage rate to rise, inflation picks up even as the deficit falls. It is indeed possible that deficit and inflation will be inversely correlated with high inflation prevailing through the low deficit phase. In the second case this result is weaker and depends upon money and consumption being Edgeworth substitutes. The upshot of this argument that the links between fiscal deficit and inflation remain intact even when there is little observed correlation between the two.21

302


Table 12.3 Forms of financing the government deficit

N.B. NBF=Non-bond Finance DF=Domestic Financing (Central Government) F=Total Financing (Central Government) Ratio=NBF/DF Ratiof=NBF/F NBF=DF minus short- and long-term bonds financing plus local financing *=Provisional figures. Source: Government Finance Statistics Yearbook, IMF, 1999.


303

304


FISCAL POLICY, CAPITAL FLOWS AND THE MONEY SUPPLY PROCESS In the presently rapidly globalising world economy, a liberal capital regime is often promoted in the interest of optimal international allocation of capital. This change has been assisted by an emerging consensus, both in academic and policy making circles, on the type of policies that promote growth with equity. The new development model emphasises macroeconomic stability, competitive market structures, globalisation (integration into the world economy) and a role for the government and fiscal policy that emphasises facilitating the growth of the market and the private sector. To achieve these objectives public finances must be put in order and this chapter has spelt out some of the conditions under which this will be possible. Exchange rate policy should emphasise a reduction in the rate of inflation. Further, stabilisation policy packages advocated by the IMF and the dire necessity of attracting international capital flows has tempted many developing countries to pursue fiscal policies to attract capital inflows. These include maintaining interest rate differentials, providing favourable tax treatment to foreign capital and the like. Starting from the late 1980s, developing countries have experienced surges in capital flows. According to the World Economic Outlook (IMF, 1995), net capital flows into developing countries, as measured by the capital account surplus (inclusive of errors and omissions) increased from $18 billion in 1988 to $164 billion in 1993 and $250 billion in 1995. Most of this increase has occurred in the private sector. Net private capital flows to developing countries consisting of FDI, portfolio equity, bond issues, loans and other liabilities, rose from $33 billion in 1988 to $167 billion in 1995. FDI and portfolio investments account for over 70 per cent of these flows. Some information on these flows is provided in Table 12.4. Some aspects of the regional distribution of such flows are portrayed in Table 12.5. Table 12.4 Net capital flows to developing countries, 1990–98 (in billions of US dollars)

Source: Global Development Finance, The World Bank (1999).


305

Table 12.5 Regional distribution of capital flows to developing countries

Source: IMF World Economic Outlook (1995) for figures up to 1989 and IMF World Economic Outlook (1999) for other years.

Such large capital flows have turned out to be a major welcome as well as worrisome macroeconomic development for these countries. Some of the obvious positive features are: (i) acquiring capital for higher economic growth; (ii) smoothing out consumption over time; and (iii) acquiring new technology and expertise through FDI. At the same time, these inflows have also posed major challenges for the conduct of monetary policy. Thus there is a tendency for the real exchange rate to appreciate as a result of the build-up of foreign exchange reserves and associated expansion of monetary base, greater speculative activity on the part of the domestic asset markets and possible disruption associated with sudden reversal of flows. Capital inflows lead to increased expenditures some of which will spill over onto tradable goods. This will increase the size of the trade deficit. If this were the only adjustment required, there would be no grounds for concern because the higher trade deficit will be financed directly by the capital inflow with no disequilibrium in the market for nontradable goods. Thus the principal reason to worry about the trade deficit would be to understand whether the implied external debt was sustainable. However, some of the additional expenditures would spill over onto the non-tradable goods sector. The size of the tradable goods sector will, therefore, shrink and that of the non-tradable sector rise. Further, this sudden increase in the balance sheet on both the asset and liability sides for the banking system may make it possible to finance investment/consumption decisions of agents that are not sustainable over the long run. This would lead to a further external current account deficit, increasing the private sector’s indebtedness and the emergence of nonperforming assets. Unless sterilised, these additions to the debit and credit sides of the financial sector’s balance sheet will lead to inflationary consequences. Further, the buildup of international debt through recurrent current account deficits can undermine the country’s creditworthiness. If the

306


domestic monetary authority attempts to sterilise the capital flows, the domestic monetary base becomes endogenous. However, sterilisation may not insulate the economy from the effects of capital inflows if these inflows are triggered by an increase in the domestic money demand that raises domestic interest rates. In addition, it might have quasi-fiscal costs for the central bank to the extent that monetary authorities purchase low-yielding foreign assets and sell high-yielding domestic assets. Moreover, if sterilisation were achieved through increased reserve requirements, it would act as a tax on the banking system and may promote disintermediation. Hence, sterilisation is at best a partial solution. It follows, therefore, that money supply becomes endogenous or, at least, hard to target because of the presence of capital flows. Hence, it is imperative from the point of view of attaining macroeconomic stability that forces of endogeneity of money supply emerging from the fiscal side be abated. If the money supply has to bear the burden of adjustment from the fiscal side as well as that of international capital flows, it would be difficult to pursue stabilisation policy. THE CARDOSO MODEL OF THE EFFECTS OF THE FINANCING OF THE FISCAL DEFICIT Cardoso (1992) considers the impact of the financing of the fiscal deficit. Consider a small open economy that finances its current account deficit either by commercial loans or by changes in foreign exchange reserves. Only the government can borrow externally. The government finances its budget deficit by borrowing abroad and by adding to the monetary base that leads to monetary expansion. He writes the government’s budget constraint as:

where is the world interest rate, Dt is the external debt and et is the exchange rate. ∆Dt is external borrowing and ∆Ct is domestic credit creation defined as the change in monetary base, H, minus the change in foreign reserves, The balance of payments in dollars can be written as: where NXt represents net exports. Substituting into the government’s budget identity and defining the inflation rate as and Rt as the real exchange rate we get:

Dividing both sides of this expression by real national income, yt, we have:

We define


307

h stands for the ratio of the real monetary base to income and g is the ratio of the primary budget deficit to output and θ is the ratio of net exports to output. This expression then shows the link between deficits and the monetary base. Nominal interest rate is related negatively to h. It is assumed that the interest rate adjusts continuously to clear the money market. Thus we have: To model the inflation dynamics it is assumed that in the goods market inflation increases whenever the real interest rate is below the interest rate consistent with full employment. Thus inflation dynamics are defined by: This equation implies inflation inertia, which indicates that there is scope for monetary policy to affect inflation. The steady state equilibrium of this model is then expressed as a system of two dynamical equations. The budget constraint in steady state is defined as

This defines a downward sloping—in fact a rectangular hyperbola type— relation between the two endogenous variables: πt and ht. The second such relation comes from the inflation-interest rate link and can be written as: πt = κ (ht)–r, i.e. This is also a downward sloping relation. Hence there is the possibility of multiple steady state equilibrium. This is shown in Figure 12.1. To the right of the schedule, inflation is rising, whereas inflation is declining to the left of this schedule. Points to the right of the constant ht schedule involve falling real cash balances, whereas points to the left indicate rising cash balances. The disequilibrium zones are as labelled in Figure 12.1. There are two steady state equilibria: E1 and E2. E1 can be a stable equilibrium whereas E2 is an unstable equilibrium. Cardoso (1992) shows that the economy will converge to E1 if the rate of inflation is low, the interest elasticity of real cash balances is not too large and inflation responds sluggishly to differences between current and full employment interest rate, i.e., γ is small. We can then address the question of what happens to this economy as the budget deficit gets worse. Intuition suggests that this would ultimately lead to a shift in the stable equilibrium toward higher inflation and lower real cash balances. An increase in the budget deficits requires a higher full employment real interest rate. Thus the schedule will shift to the left. But it will shift the constant ht schedule to the right since the monetary finance of the increased deficit would put higher real cash balances with the public. As money increases in response to the higher budgetary deficit the nominal interest rate falls, which stimulates output and pushes up the inflation rate. Ultimately inflation catches up with money supply growth and will then exceed it. This will reduce real cash balances and increase the real rate of interest. Thus the stable equilibrium in Figure 12.1 moves in a northwesterly

308


Figure 12.1

direction with higher and higher deficits. It is conceivable that with higher and higher fiscal deficits, monetisation will lead to a situation in which the two schedules in Figure 12.1 do not intersect and no viable stable equilibrium exists. This could have been the case with the high inflation episodes in Latin America, particularly Brazil and Mexico, studied by Cardoso (1992). CONCLUSIONS This chapter has considered some aspects of the effects of fiscal policy on macroeconomic adjustment in developing countries. Broadly, two areas of concern are delineated. The first spells out the conditions under which the internal and external debts are sustainable and pointing out the role of the ‘twin deficits’. In this context this chapter presents some evidence on the sustainability of the internal and external deficits in the context of some developing countries. The second broad theme of this chapter is tracing the sources of endogeneity in money supply to fiscal policy and international capital flows pointing out the difficulties faced by stabilisation policy under these conditions. APPENDIX: RICARDIAN EQUIVALENCE The government budget constraint and all that This note articulates the basic mechanics of the government budget constraint. Let DFt be the real government budget deficit in year t. It is straightforward to write the government budget deficit as:


309

(A12.1)

All variables are in real terms, r is the real interest rate on government debt, Bt-1 is the stock of debt from the previous period so that rBt-1 is the interest payment on the government deficit. This is one item of expenditure. The second item of expenditure is Gt—real government expenses. Tt is the real value of tax revenue (net of transfers) collected by the government. Hence the budget deficit equals total government expenditure less tax receipts. The expression in (A12.1) happens to be the correct value of the deficit. Official publications, however, do not make a distinction between real and nominal magnitudes and report current figures. The official (nominal) measure of the deficit (NDFt) can be written as: (A12.2)

where i is the nominal rate of interest. Now suppose that we are in a situation where NDFt is zero. If the economy is experiencing zero inflation then (A12.1) and (A12.2) are equivalent methods of expressing the deficit. However, suppose that there is positive inflation going on. In that case, the real value of the government debt will be going down (at the current rate of inflation πt) so that (A12.2) would overstate the amount of public debt outstanding. The correct measure of the deficit would then be:

This is sometimes also called the inflation-adjusted deficit. From this the government budget deficit can be written simply as: (A12.3)

This expression simply states that the issue of new debt today finances the deficit of the government. Whenever there is a deficit the debt goes up; when there is a surplus the debt goes down. This expression can be written in an alternative way: (A12.4)

The first term on the right-hand side represents interest payments and the second term is called the primary deficit, i.e. one without interest payments. Moving Bt-1 to the right-hand side we have (A12.5)

so that the current debt represents the sum of the amortised value of the past period debt and the primary deficit. It gives an intertemporal relationship between debts in successive periods. Let us now conduct the following thought experiment. Suppose the debt and the primary deficit are both equal to zero in period 0. The government increases its expenditure by $1 and does not increase taxes. Debt will go up

310


by $1 in the first year. If the primary deficit is still equal to zero, debt will now be In year 2 with the primary deficit still equal to zero, the debt will be More generally then the debt in year t will be If the debt is to be cleared up in year t then the primary surplus that the government must run in that year must exactly equal This is the amount of additional taxes that must be raised in year t. It is now useful to introduce the notion of Ricardian Equivalence. This concept was first floated by the nineteenth-century British economist David Ricardo but later developed in its modern form by Robert Barro (1974) in a now famous paper. The Ricardo-Barro Proposition states that if the government finances its expenditure by issuing debt which will have to be paid off by raising taxes (running a primary surplus) at a later date, rational economic agents will see this as a substitution of taxes later for taxes now. Hence the issue of government debt will not cause rational economic agents to change their behaviour and government bonds will not be considered net wealth. Government bonds are, then, fundamentally different from private bonds. Private bonds are wealth for the household sector and a liability for the firms that issue such bonds. Government bonds are neither an asset nor a liability for the household sector since they realise that the amortised value of the expenditures financed by these bonds will exactly equal the value of additional taxation imposed on households to retire the debt. A similar argument to the above can be made in respect of savings. Suppose that the government announces a tax cut of $1 today (in order to stimulate the economy) but no commensurate cut in government expenditures. Will consumption go up today? If Ricardian Equivalence holds, rational economic agents will realise that they will have to shoulder an additional tax burden of in year t. The rational strategy for them would then be to reduce current consumption by $1 and earn interest on it at the rate r. So that in year t they will have exactly to pay the additional tax. Hence, the rational response to the tax cut is not to increase consumption but to increase savings. One possible way out of the Ricardian Equivalence outcome is to have governments increase taxes to pay off the additional debt at a very late date in the future and hope that people will discount the future sufficiently so as to increase consumption now in response to a tax cut. Typically the date for payment of the debt may be postponed until the next generation is making the crucial decisions and the current generation is either dead or retired. If people do not care enough about future generations then it may be possible to have new government bonds be considered as net wealth by the current generation. In essence, then, whether the Ricardo-Barro Proposition holds or not becomes an empirical question. How valid is the Ricardian Equivalence Proposition in developing countries? The answer to this is complex, except that very rarely, if at all,


311

does this proposition hold exactly. At issue here is the extent to which the private sector internalises the public sector’s budget constraint. In the case of India, Muhleisen (1997) presents evidence that the Ricardian equivalence proposition is only of minor significance. A 1 per cent rise in public savings is met by only a 0.25 per cent decrease in private savings. Corbo and SchmidtHebbel (1991) and Haque (1988) estimate this figure to be 0 per cent and 0.5 per cent respectively for a cross-section of developing countries. Hence there is considerable evidence that the Ricardian Equivalence Proposition, although in the right direction, does not hold exactly in most developing countries. DISCUSSION QUESTIONS 1

2

3

4

5

Debt sustainability, while it refers to the long run, has important implications in the short run as well. In a developing country context list the possible effects of excessive government borrowing on (a) ability of the central bank to conduct an independent monetary policy; (b) interest rates, (c) exchange rates; and (d) external borrowings. Governments in developing countries often combine high borrowing with repressed financial markets—in other words they control the interest rates at which they borrow. Most of this debt sits in the vaults of the central banks of these countries since the market-determined interest rates are much higher. Consider a reform programme that involves the central bank offloading some of this debt onto the free market. What effects do you expect this to have on interest rates, the fiscal deficit? In some countries like Zambia in recent times, much of government borrowing done is externally. In fact the gap between even minimal government expenditures and government revenues is so wide, and domestic financial markets are so poorly developed, that only external flows can fill this gap. Using EIU reports or World Bank/IMF country reports on Zambia comment on the reasons why Zambia finds itself in this position. Some economists have made the argument that developing countries that have considerable forest cover can ask for debt forgiveness and/or rescheduling in exchange for them not cutting down their forests. This ‘debt for nature’ swap has been pushed forward as a method of getting the double dividend of both a better environment as well as lower international debt for developing countries. For a review of these arguments see Jha and Schatan (2001). Write a short critique of this point of view. Some economists have argued that by charging a positive interest rate on their loans, the World Bank and the IMF take back more from developing countries than they give them as loans. Even the US Treasury has now advocated that the World Bank should give grants rather than loans to developing countries. Evaluate this point of view.

312


NOTES 1 The Burkina Faso figure is taken from IMF (2000c) whereas the figure for India comes from the Reserve Bank of India Bulletin (2000). 2 The data are from the IFS CD ROM. 3 Most of the least developed countries effectively face vertical supply curves of external capital inflows. Various reasons can be attributed toward this. Macro policy disasters such as those in Tanzania have played an important role. Others such as Guinea are just too small. Countries such as Angola do manage to get some FDI in resource extraction. Angola, for example, borrows externally by mortgaging its potential oil revenues but there are limits to this when debt rises to high levels and additional loans can be obtained only at very high interest rates. 4 For a further elaboration of this point see Gemmell (2000). 5 The distinction is sharply illustrated by the recent experience of the US. For the past five years or so, tax revenues in the US have been very buoyant following high employment conditions. As a result a stubbornly high US budgetary deficit has turned into a surplus. The US economy, of course, does not face a foreign exchange constraint so that this transformation was possible. Recent recessionary trends coupled with a substantial tax cut have now placed this surplus at risk and large deficits are not likely to appear. 6 Recent arguments along the lines suggested by Buiter include those by Easterly (1999) who argues that fiscal adjustment can be illusionary. In particular, this would be the case when such adjustment lowers the public debt but leaves unchanged the net worth of the government. In other words, governments may find ways of maintaining their consumption even when they are actually involved in a process of reducing public debt. 7 On February 28, 2001 India’s oil pool deficit alone, the Economic Times, March 1, 2001 reported, stood at 0.5 per cent of GDP. 8 In India the prices of petroleum products are administered by the government through an ‘oil pool’ mechanism. The stated purpose of this is to provide uniform and stable prices of such products to consumers within the country and reasonable profit margins for the oil companies. The oil pool accounts are supposed to be self-financing. The inflows into the pool account are from collection of surcharge on the sale of petroleum products while the outflow is for meeting the variation in the elements of standard cost. The difference between inflows and outflows represents the surplus/deficit position of the oil pool account, and is not included in the budgetary deficit of the central or state governments. In recent years this account has persistently been in deficit since revisions to consumer prices of petroleum products have been far from complete. This tendency has been exacerbated with the recent sharp rises in world petroleum prices. Some of the deficit in the oil pool account is covered by the issue of ‘oil bonds’ issued by the government and bought by oil companies, which can redeem these bonds at specified future dates (see Ministry of Finance, 1997). 9 The left-hand side of the government’s budget constraint must also include exchange rate changes and privatisation which reflect changes in government net debt due to revaluation of the government’s financial assets. In the empirical work reported in this chapter this broader definition is used. 10 The task of sustaining a debt ceiling is easier said than done, particularly in lowincome countries with poor tax institutions. This is primarily because investment and productivity of tax institutions would probably grow at a lower rate than the public debt itself. Thus countries trying to cut the debt by raising more revenues would probably need to raise taxes—particularly distortionary taxes such as


11

12 13

14 15

16

313

trade and indirect taxes. This may cut the debt/GDP ratio in the short term but may end up hurting growth and, therefore, the tax potential in the medium run. Hence the bulk of adjustment must come from cuts in expenditure. This is typically associated with cuts in social expenditures and spending on infrastructure. If international donors are willing to provide more concessional loans (or outright grants) where the interest rate charged is lower than that prevailing, then some of the debt can be retired using these foreign funds and cut its domestic interest bill. This policy has yet to be tried in a significant way anywhere. A relevant aspect of the tests for sustainability is to ascertain whether the time series involved have a break. Tests along the lines of Gregory, Hansen and Bruce (1996) were conducted but are not reported here since the data set involved is not long enough to give robust results. However, results of these tests are available from the author. After Nigel Lawson, UK Chancellor of the Exchequer, who in 1988 dismissed the possibility of any causal link between internal deficit and the current account deficit. It is important to underscore the point that there is a subtle difference between low- and middle-income countries with respect to the links between external sustainability and currency crises. Surely, large current account deficits financed by capital flows are possible only in countries that are successful in attracting such flows. This would typically be possible primarily in middle-income countries. The banking systems in most low-income countries, with the possible exception of India, are simply not robust enough to attract and retain such flows. Even in India major capital flows, particularly in crucial infrastructural areas, are attracted by the central government providing counter-guarantees to the investing corporation. This then creates difficulties for Indian balance of payments similar to those for middle-income countries. These problems may become more common with more low-income countries exploring the counter-guarantee route to attracting capital flows. Hence supplementing domestic savings with foreign savings is a strategy open only to a few countries and even then this is fraught with some difficulty. For evidence on Ricardian Equivalence in the developed country context see, for example, Bayoumi and Masson (1998). One largely unresolved issue is whether net FDI flows should be included in the calculations for sustainable current account deficits. Frankel and Rose (1996) find in a panel of 100 developing countries from 1971 to 1991 that a high ratio of FDI to debt is associated with a low likelihood of a currency crash. Between 1970 and 1982, for example, Singapore ran a current account deficit between 12.1 per cent and 20 per cent of GDP. Almost one half of the gap consisted of FDI. The savings rate within this period doubled from 21 per cent to 40 per cent, the economy grew at an average rate of 8 per cent and there was no currency crash. Hence, there is some presumption that since FDI is determined by longterm considerations, generates positive externalities and does not exert pressures on the real exchange rate if it is the outcome of a privatisation programme, is in a different category when reckoning sustainability of current account deficit. However, if capital is fungible this optimism may be misplaced as Reisen (1999) argues. There is an argument that financial bailouts by international bodies such as the IMF lead to an adverse selection problem in that countries know that since they will be bailed out they may follow imprudent macroeconomic policies. The International Institute of Finance (IIF) (1999) however, reports that the evidence on this is weak—particularly in the case that has often been cited as the most

314

17

18

19

20

21


important example of such imprudent macroeconomic behaviour—the Mexico financial support package of 1995. It has been argued that this package caused moral hazard that was a force behind the large volume of lending at low spreads that occurred in emerging markets in 1996 through mid-1997. On the contrary, the study finds that the empirical evidence does not support the moral hazard diagnosis. Instead, the decline in emerging market spreads can be explained statistically by buoyant international liquidity conditions (as measured by highyield US corporate spreads) and (to a much smaller extent) by improving economic fundamentals in borrowing economies. A good illustration of this point comes from South Africa. The government has been quite prudent in terms of domestic financing, despite facing large demands for public spending to rectify the inequalities of apartheid and accelerate employment growth. But the need to defend the rand against speculative attack has led to domestic interest rates significantly above world rates for the last three to four years with damaging effects on private sector investment and growth. For aid-dependent economies such as Tanzania this problem may be less severe insofar as they rely on aid inflows to finance the current account deficit on concessional or grant terms. In the case of Mozambique, for example, much of the external deficit—which widened during reconstruction—as essential imports exceeded war-damaged exports is aid financed. In this case, then, the focus of attention should ideally be the internal and not the external deficit. Schuknecht (1999) argues that using a nominal anchor like a fixed exchange rate regime may not be sufficient to pursue stabilisation objectives. He models electoral business cycles in the case of 25 developing countries and shows that governments indeed try to improve electoral prospects by pursuing expansionary fiscal policies around election time. This is more likely to occur, ceteris paribus, in an economy with a nominal anchor such as a fixed nominal exchange rate. These episodes would then be followed by higher inflation and exacerbation of fiscal deficit pressures. In this context, Sargent (1999) argues that monetary policy can be constrained by fiscal policy whenever fiscal deficits grow large enough to require monetisation. This can happen in developed as well as developing countries and irrespective of whether the central bank is independent or not. In such situations the central bank loses its ability to influence income or interest rates and can influence only the time path of prices. Issler and Lima (2000) illustrate the importance of this phenomenon in the case of Brazil. They show that not only is Ricardian Equivalence important (so that the fiscal deficit and the current account deficit are linked), but also that without seignorage revenues the Brazilian deficit would not be sustainable. To be sure, Buffie’s argument is valid principally for the non-sub-Saharan Africa region countries. In low-income countries of this region such as Zambia, low real wage phases have not typically been followed by high real wage phases.

13


WHAT IS SEIGNORAGE? As has been discussed, governments of many developing countries finance public sector deficits through printing of money. This has revenue effects that we discuss in this chapter. We also discuss the problem of inflation control in general with a postscript for developing countries. Seignorage is the revenue that the government earns by virtue of the fact that it has monopoly on the printing of money. It is related to but different from the concept of an inflation tax. In this section we will go over both these concepts in some detail. Consider the demand function for real cash balances: (13.1)

where i is the nominal rate of interest and Y is real GDP. Using the Fisher equation we can write the nominal interest rate as the sum of the real interest rate, r, and the rate of expected inflation (πe). Thus we can write (13.1) as (13.2)

where and M should be interpreted as high-powered money (currency plus reserves). Consider, first, the steady state of this model. Under this condition it would be reasonable to suppose that output and real interest rates are unaffected by the rate of money growth. Further we may expect actual inflation and expected inflation to be equal to each other. In steady state the rate of inflation would be equal to the rate of growth of money supply, gM, where

For the sake of simplicity we can further assume that the real rate of interest and the real GDP are constant in steady state so that we can write: (13.3)

316

Policy Dilemmas Faced By Developing Countries

The seignorage revenue of the government is equal to the real value of additional money stock because this is the additional resource it can command through money creation. Thus we can write seignorage revenue Rs as: (13.4)

Thus in steady state real seignorage revenue equals the rate of growth of the money supply multiplied by the real cash balance. Steady states associated with higher rates of growth of money supply will have higher gM but, by virtue of the higher inflation, lower M/P. Hence there is a tradeoff. Substituting money demand for M/P we have (13.5)

Equation (13.5) shows that an increase in gM increases seignorage by raising the rate at which money balances are taxed, but decreases it by reducing the tax base, M/P. Thus we have (13.6)

where L1 denotes the derivative of L with respect to its first argument. The first term on the right-hand side is positive and the second term is negative. Hence this expression has an upper bound which can be solved for by setting This would then give the value of the expansion of the monetary stock that would maximise the revenue from seignorage. Seignorage revenue, while not inconsiderable under most circumstances, is especially important under conditions of hyperinflation. Hyperinflation is typically defined as periods of inflation above 50 per cent per month. Under such extraordinary conditions, tax systems often break down and the government has to increasingly rely on seignorage revenues. The phenomenon of hyperinflation was studied in a classic paper by Cagan (1956).1 He wrote a specific form of the money demand equation as: (13.7)

where b is the interest semi-elasticity2 of the demand for money. Units have been so chosen in (13.7) that the income elasticity of the demand for money is unity. For this money demand equation seignorage revenue would be: (13.8)

where C represents the constants in the first term on the right-hand side of (13.8). Differentiating with respect to gM Cagan gets the condition for maximal seignorage revenue as: (13.9)


317

Figure 13.1

which will be positive for and negative otherwise. Thus maximal seignorage revenue in steady state is achieved when For most countries Cagan reported estimates of b to be between 1/3 and 1/2 so that the compound rate of growth of money stock is between 200 and 300 per cent. The annual rate of increase of the stock of money can thus be between and per year. In a subsequent contribution Romer (1996) calculates that seignorage at the peak can fetch about 10 per cent of GDP. This is a significant source of revenue, but woefully inadequate if it is to be the most important source of revenue, since government expenditures are rarely below 15 per cent of GDP. The seignorage revenue as a function of the rate of monetary growth is depicted in Figure 13.1. Maximal seignorage revenue is obtained at the rate of monetary growth as shown in the diagram. If the government wants high seignorage revenue it can easily push the economy into a high inflation phase. To see this let us take a numerical example. Suppose, as per the previous example, that revenue to the extent of 10 per cent of GDP can, at most, come from seignorage and that If that is the case then C must equal 9 per cent of GDP. If the government actually collects revenue to the extent of 2 per cent of GDP then gM would be required to be 24 per cent, for 5 per cent revenue gM would need to be 70 per cent whereas for 8 per cent it would have to be 142 per cent. Remember that we are talking of the steady state where the rate of inflation is equal to the rate of monetary growth. Thus even with moderate revenue requirements from seignorage the inflationary implications are quite serious. We have to keep in mind once again that we are talking about the steady state. In particular in this state cash balances have been adjusted to their

318


equilibrium value. However, if that is not the case and cash balances are slow to adjust then it is possible that at any point in time cash balances are too high to be optimal for the prevailing rate of inflation. In such a situation seignorage revenue can be higher than that given by Figure 13.1. To see the effects of this we need to distinguish between desired and actual real cash balances. With constant real interest rate and constant real national income, desired real cash balance (m*) can be written as: (13.10)

Cagan now brings in the notion of adaptive expectations—that adjustment in real cash balances are a constant fraction of the gap between actual and desired cash balances. Thus we have: (13.11)

with

Substituting for ln m* from (13.10) we have (13.12)

Seignorage revenue, as before, equals (13.13)

Suppose that initially the economy is in steady state with government revenue requirements from seignorage less than the maximum possible, Cagan says that hyperinflations occur when government requires an amount of revenue greater than and thinks it can get this because cash balances are slow to adjust. This indeed is possible in the short run. Now we know that the real money stock grows at the rate gM less the rate of inflation. Thus (13.14)

Hence (13.15)

Revenue requirement is

Substituting from (13.11) we have (13.16)

We can now solve (13.16) to get

(13.17)


319

Figure 13.2

We have assumed that Hence it follows that the expression in the curly brackets {.} is negative. This follows from the fact that desired money stock is Ce-bπ. Taking logs and rearranging we will get In steady state seignorage is πm. Hence the sustainable level of seignorage when money balances are m is Since we are assuming that it follows that the expression in the curly brackets is negative. Further since the right-hand side of (13.17) is everywhere negative so that the money stock continually falls. Thus if the government has to get its revenue it has to inflate at ever higher rates. This then is the genesis of hyperinflation. Thus this is a reasonable explanation of the phenomenon of hyperinflation. An examination of equation (13.4) should convince us that seignorage revenue can be positive even in the absence of inflation. If monetary expansion is just enough to satisfy the increase in the demand for money following economic growth, the price level will be stationary but seignorage revenue will be positive. The inflation tax, on the other hand, refers essentially to the change in revenue from taxation of money balances in response to a change in the inflation rate. It is seignorage viewed from the vantage point of the household except that inflation tax is defined only for positive rates of inflation whereas seignorage can occur at zero rate of inflation. The basic idea involved is depicted in Figure 13.3. Figure 13.3 depicts a demand for money schedule in nominal interest ratereal cash balances space. Initially the nominal interest rate is i0 corresponding to which the money demand is (M/P)0 with rate of inflation π0. Suppose now that the rate of inflation goes up to π1 so that the interest rate climbs up to i1 and money demand falls to (M/P)1. The area i1i0AB represents the inflation

320


Figure 13.3

tax due to the higher inflation rate. The opportunity cost difference for holding real money balance is which is the rate of the tax and the base of the tax is the existing stock of real cash balances, (M/P)1. Now, suppose that the real rate of interest is constant then This multiplied by cash balances is the same as what we get from the formula for seignorage— however, remember that seignorage is defined for zero inflation whereas the inflation tax is not. COSTS OF INFLATION The costs of inflation stem from the fact that the basic purposes of using money are compromised during (at least high) inflation. Money essentially serves three purposes. First, it is a medium of exchange. Goods and services are bought and sold for money. Second, money is a unit of account. Third, money is a store of value. People assess their assets and liabilities in monetary units. During high inflation money serves all three purposes badly. When inflation is high, relative prices are more variable and it is difficult to use money values to assess relative prices. Second, money’s role as a unit of account is compromised since the relation between money value and real value becomes tenuous during periods of high inflation. Similarly, people are less inclined to store their wealth in money form since the rate of return on holding money in this form becomes highly negative and is more likely to move into less liquid assets like stocks and real estate which, nevertheless, provide some rate of return. As a result of this, doing transactions becomes more costly. Since money keeps losing its value people have to make more trips to banks. This time


321

could have been better spent in other activities. The increasing costs of doing transactions during high inflation have been termed ‘shoe leather’ costs in the literature. An additional cost of high inflation is the distortion introduced in the tax system. Inflation interacts with the tax system in a variety of ways to change real tax liability even when real earnings have not changed. For example, suppose that with zero inflation a person is earning $10,000 a year and pays income tax at 10 per cent, i.e. $1,000. Suppose now that the rate of inflation rises 20 per cent and that his nominal income maintains pace so that it rises to $12,000. Suppose, however, that with an unindexed or only incompletely indexed tax system this lands the person in the 15 per cent income tax bracket. The tax liability rises to $1,800 whereas with a fully indexed system it would have risen by only 10 per cent to $1,200. Hence the after-tax real disposable income of this person is lower even though the real pretax income is unchanged. Such arbitrary changes in tax liabilities will have drastic implications for factor supply. Inflation also interacts arbitrarily with other taxes including capital gains taxes (see Jha, 1998). Another important cost of high inflation is the erosion of real incomes of fixed income earners such as pensioners, as well as those people whose incomes are only incompletely indexed to inflation. It may well be argued that these costs are not really the work of inflation but the result of incompletely indexed tax and salary structures. However, complete indexation of tax and salary structures is rarely found even in developed countries. WE must remember we are talking of developing countries here. These costs of inflation occur even when this inflation is fully anticipated.3 If inflation is unanticipated, it imposes additional costs. Since this inflation was unanticipated people are not ready for it, by definition, and it has not been factored into contracts. Thus the inflation imposes an unplanned erosion of real values. The real value of debt goes down so that debtors gain and creditors lose. The greatest debtor of all in most economies (including developing economies) is the government. So governments with large debt overhangs sometimes contemplate a sharp bout of unanticipated inflation to reduce the real value of their liabilities. This comes at considerable cost to the rest of the economy and involves a sharp redistribution from the private to the public sector. While there is considerable appreciation of the costs of high inflation in the literature, costs of low inflation are less well articulated. It is generally recognised that most countries prefer some inflation to complete price stability. However, depending upon their economic structures, countries with low inflation can also at some times be incurring economic costs.4 The economic costs could be more severe in the short run for high costs of inflation and subsequent disinflation. Both cross-section and time-series empirical evidence indicate a negative relation between economic growth and high inflation.5 The relation between growth and low rates of inflation is less clear (Sarel,

322


1996). This gets more complex when an economy starts to disinflate and must incur the associated costs.6 A country that has successfully disinflated has paid the price both of inflation as well as disinflation. For good reason, then, most economists oppose high inflation. However, there is less agreement on what constitutes an optimal rate of inflation apart from stating that it should be a low rate of inflation. Some economists believe that inflation is a major evil, and argue that monetary policy (or monetary reform) should be geared toward its outright elimination (see, e.g., Gavin, 1990; Gavin and Stockman, 1988; Hoskins, 1990, 1992; Howitt, 1990a; and Selody, 1990a, b). Many others argue that eliminating inflation would reduce output and employment, and the cost of the lost output and employment would more than offset the gains from establishing price stability (see, e.g., Fortin, 1990; Lucas, 1990; Peters, 1990; and Scarth, 1990). Still others argue that the costs of low inflation are small anyway, and could be dealt with by other means (e.g., indexing the fiscal system, as in Aiyagari, 1991). It is helpful if we compare this figure with those implied by studies published earlier. Foster (1972) and Garfinkel (1989) provided estimates of the cost of a fully anticipated inflation rate of 4 per cent, which Foster estimated to be less than one-twentieth of 1 per cent of GNP, and Garfinkel estimated to be about 0.3 per cent of national income. Similarly, Fischer (1986:46), Lucas (1981:43–4) and McCallum (1989:127) estimated the cost of a fully anticipated inflation rate of 10 per cent, the first two putting it at 0.3 per cent of national income, and Lucas putting it at 0.9 per cent. However, one problem with such estimation is that they refer to losses borne in the current period and ignore losses in the future. To get an estimate of the complete loss from inflation one should compute the present discounted value of the costs in future time periods, and allow for growth of real cash balances. If this is done, costs of inflation are likely to be higher as suggested by Feldstein (1976), Tatom (1976) and others. Other authors such as Leach (1983) make the point that if inflation is a tax on real balances, then it functions in some ways like a general commodity tax which has the effect of reducing the supply of labour and, hence, total output. The costs of this distortion to the labour market must then be added to the earlier real balance ‘triangle’ loss. In this same spirit Eckstein and Leiderman (1992) show that many earlier estimates of the welfare loss depend in an important way on the maintained assumption of a Cagan-style semi-log specification of the demand for real balances (equation (13.7)). They develop an alternative approach based on a Sidrauski monetary model and provide simulations of the (steady state) welfare loss by estimating the model on Israeli data. Their results suggest that an inflation rate of 10 per cent produces a current period welfare loss of about 1 per cent of national income. These figures are considerably bigger than traditional estimates, and one’s confidence in them is reinforced by the fact that the Eckstein-Leiderman specification produces seignorage and national income ratios much closer to observed experience than the traditional


323

Cagan specification does. De Haan (1990), uses a Sidrauski model to estimate that the current period welfare loss of a 5 per cent inflation relative to zero inflation was about 3.12 per cent of output. This figure is much bigger than figures from traditional exercises. Sidrauski models thus generate considerably bigger losses than traditional models based on the Cagan specification of the demand for real balances. An additional point to remember is that high inflation is also more variable inflation. Variable inflation has its own costs in that prices become more difficult to predict. Scarce resources have to be expended on predicting prices and to that extent welfare is lower. One of the most important messages from the above analysis is that inflation, particularly high inflation, is to be avoided. Even low levels of inflation have their costs. The question then is how does one minimise the rate of inflation? Several issues are involved here and the literature on this topic is rich and expanding rapidly. Although this literature is typically cast in the framework of advanced countries, many of the issues are of relevance to developing countries and need to be considered seriously. 1

2

3

4

The first point to be understood is that typically the control of the money supply is with the central banker of the country. He may be interested in pursuing his own agenda—in particular he may be interested in exploiting any potential inflation-unemployment tradeoff. However, rational economic agents know that and will respond appropriately. Will the central banker be successful in this effort then? A second question is what kind of central banker would be able to better address the fact that inflation has considerable welfare costs and opt for a low inflation strategy? Yet another question is, given that economic agents know that the central banker has this option to exploit the inflation-unemployment tradeoff, and therefore will find any stated intention of driving down the rate of inflation to zero credible? In other words, is a zero rate of inflation unattainable? Finally, we will comment briefly upon recent efforts to design incentive mechanisms for central bankers to follow the public’s preferences with respect to inflation—so-called inflation targeting. We will also examine recent literature on the usefulness of this policy for developing countries.

These are some of the important issues in central banking that we will tackle in the remainder of this chapter. THE TIME INCONSISTENCY OF OPTIMAL MONETARY POLICY Although optimal monetary policy can be defined it should not be pursued. This is the simple but powerful message of an important branch of the central

324


banking literature. This issue was first broached in a classic paper by Kydland and Prescott (1977). They develop this argument as follows. Suppose that there is a short-run tradeoff between inflation and unemployment of the sort given by the expectations augmented Phillips curve or the Lucas supply curve. This is given as: (13.18)

where y is the log of real output, y is the log of the natural rate of output and π and πe are, respectively, the actual and expected inflation. This is the output supply equation from the standard Phillips curve. Because of rigidities in the system like the existence of imperfect competition the flexible price level of output is less than the socially optimal level of (the log of) output y*. The central banker would like to minimise the gap between y and y*. Kydland and Prescott also assume the existence of an optimal rate of inflation (π*). For reasons outlined above this optimal rate of inflation is presumably quite low—it could even be zero. The central banker’s objective is to minimise the following social loss function: (13.19)

This is a simple quadratic loss function. That gaps in output from desired output are different from gaps in actual inflation from desired inflation are captured by the term It is also implicit that The central banker controls the growth of the money supply that then goes on to determine the level of aggregate demand. There is no uncertainty in the model so that we can think of the central banker as choosing the rate of inflation directly subject to the constraint that inflation and output have to be related to each other as per the supply equation (13.18). Suppose that the central banker could make a binding commitment about what the rate of inflation would be before expectations are formed. Since the commitment is binding and there is no uncertainty, people will expect the rate of inflation to be exactly that announced by the central banker. From the supply equation (13.18) output will equal the natural rate of output. The central banker knows that this will happen. Hence in choosing the rate of inflation the central banker minimises

In this equation π is the

only variable of choice—the others are all parameters. The optimal solution obviously is to set Hence with a ‘rules’ type policy actual inflation is always equal to the desired level, although output is less than optimal. Thus social loss is


325

Now suppose that the central bank does not commit itself to this policy or, equivalently, economic agents do not believe that the central bank will commit itself to this policy. In this case, the central banker will take economic agents’ expectations of inflation as given and try to minimise social loss. In doing so, it is effectively ignoring the effects of its policy on the economic agents. Thus the central banker chooses π to minimise

The first-order condition is, simply, One can solve this expression for the optimal π to get: (13.20)

In equation (13.20) realise that since π is an increasing function of πe. Furthermore, if people have rational expectations and, since there is no uncertainty, in equilibrium π must equal πe. To solve for this equilibrium rate of inflation set in (13.20) and solve to get: (13.21)

It is easy to visualise this situation graphically as in Figure 13.4. In Figure 13.4 along the 45° ray from the origin, The other line is a graph of equation (13.20). πEQ is the only rate of inflation that is consistent

Figure 13.4

326


with central banker’s loss minimisation and for which agents’ expectations are satisfied. Realise now the implications of this. From equation (13.18) when y must equal y. However, inflation is higher than π*. The policy in which the central bank simply tried to set and did not attempt any minimisation of the loss function was also associated with output equal to the natural rate but, at least, In the case of explicit minimisation, on the other hand, output is still equal to the natural rate but inflation is higher. Hence explicit loss minimisation involves a greater loss in welfare. It is better to pursue policy such that a simple rule of the form is practised and is believed. The basic difficulty with the optimising arrangement is, however, that economic agents realise that the central banker has the temptation to minimise loss and would form expectations under this assumption. Economic agents know that even if the policymaker announces a rules policy of and people form expectations of π such that the central banker will then have a temptation to renege on his promise by exploiting the inflationunemployment tradeoff to raise output. Hence it is not optimal from the point of view of economic agents to set irrespective of what the central banker announces. This phenomenon has been termed dynamic inconsistency of policy. Another way of characterising it is to say that the central banker’s strategy is not subgame perfect. The policy dilemma is, then, not whether we should pursue a rules policy or a discretionary policy but how to convince economic agents that a rules policy is being followed. At a superficial level the solution to this problem would appear to be simple. All one has to do is to enact a constitutional provision that only a rules type monetary policy would be followed. However, this would be unduly restrictive and would not permit the central banks any flexibility. Consider a situation where, for example, there has been a credit collapse in the economy for whatever reason. Suppose the central bank is not allowed the flexibility of responding to this by loosening monetary policy. Furthermore, one has also seen that in recent times inflation in many countries including some developing countries has been low despite policy not being made according to rules. Hence, it would appear that it should be possible to address the dynamic inconsistency problem without enacting strict rules. We now discuss some of the approaches taken in the literature to address the dynamic inconsistency problem. DELEGATING MONETARY POLICY TO A CONSERVATIVE CENTRAL BANKER This idea was proposed in an important paper by Rogoff (1985). His solution to the problem is disarmingly simple! Entrust monetary policy design to a person of eminence who is known to be particularly averse to inflation. In


327

Figure 13.5

words, the central banker has a loss function L’ different from that in equation (13.19). The form of the central banker’s loss function is similar to (13.19) except that the central banker has a higher aversion to inflation. (13.22)

The public knows that Going through the same loss minimisation exercise as earlier, the equilibrium fully-anticipated rate of inflation is now (13.23)

Since this equilibrium rate of inflation is now lower than previously. The implications of the Rogoff conservative central banker are shown in Figure 13.5. The increase in inflation aversion shifts the relation downwards and clockwise as shown in Figure 13.5. The equilibrium rate of inflation for Rogoff’s central banker is lower at This is an improvement upon the previous situation. REPUTATION MODELS OF MONETARY POLICY DESIGN Reputation models of monetary policy design as a way out of the time inconsistency problem have been suggested by Barro (1986) and others. Suppose the central banker holds office for two time periods. It is possible that they use the first time period to establish a reputation for themselves as followers of the rules policy. This helps dampen expectations of inflation.

328


They can then use this fact to exploit the inflation-unemployment tradeoff in the second time period. To simplify the algebra suppose that the social welfare function is linear in output and that Social welfare in any time period t is given by: (13.24)

Suppose that there are two types of central bankers and the public does not know which one the current central banker is. Type I policy maker has the same social welfare function as the public’s and, therefore, maximises (13.25)

where Wi is the welfare in time period i and δ is the rate of discount. The Type II central banker cares only about inflation and sets in each time period. The public does not know which type the current central banker is. The decision-making problem of the Type II problem is relatively straightforward since he sets in each time period. We need to concentrate on the Type I planner. Even for this planner the second period problem is straightforward since all he has to do is to choose π2 to maximise the second period welfare. In other words, the second period maximisation problem of the Type I planner is:

where is the expectations of inflation for the second period. The solution to this problem is to simply set The intuition behind this result is relatively straightforward—since the game ends in the second time period for the central banker, he should exploit the inflation-unemployment tradeoff to the limit. The important point is, however, that his actions in the first time period will determine what is. So the first period problem assumes considerable significance. If the central banker chooses any inflation rate different from zero in the first time period, economic agents know that they are facing a Type I central banker and set Thus in the second time period and output would equal the natural rate of output, y The value of social welfare for the central banker must then be:

(13.26)


329

Here ‘IT’ stands for Type I central banker revealing the truth about his identity. Let q be the probability that a Type I central banker chooses Suppose the central banker indeed sets For the public this poses a problem of identification. Suppose that the central banker is either Type I with probability p or a Type II central banker masquerading as a Type I central banker. The probability that a Type I central banker chooses zero inflation is pq—the probability that we have a Type I central banker multiplied by the probability that he deliberately chooses The public can then use Bayes’ Rule to make a deduction about the probability that the central banker is indeed a Type I central banker. The estimate of this probability is This is simply the probability that a type I central banker chooses zero inflation divided by the sum of the probability and the probability of a Type II banker. Its expectation of inflation for the second time period would be this probability multiplied by inflation if the central banker is actually of Type I, i.e. Thus when the central banker chooses

welfare is:

(13.27)

Realise that as q rises W0(q) falls. The intuition behind this is as follows. Suppose people observe zero inflation. If, in their assessment the probability that this is a Type I central banker masquerading as a Type II central banker is higher, i.e., q rises. This will naturally increase their expected inflation. This, of course, lowers welfare since the ideal rate of inflation is modelled to be zero. There are three possible equilibria in this model. Suppose that even when q is zero In this case the central banker would not find it fruitful to masquerade as a Type I banker. It makes more sense to simply set Clearly this will be the case for some values of the rate of discount. To find out what values these are simply consider

or, simply,

This means that welfare derived from the second period is quite small or that people discount the future heavily.

330


Another possibility is that the future is discounted very lightly and thus carries a large weight. In an extreme case of this there is a large payoff to lowering the expected rate of inflation. This will happen when This happens when

or

In this case, a Type I central banker would make every effort to convince the public that he is a Type II central banker. An intermediate situation is one in which Our analysis suggests that this will happen when In such cases the central banker would randomise. If the public believed he would choose zero first period inflation then he would choose α/a, and he would choose zero inflation if the people expected him to choose positive inflation. The equilibrium for this will occur when the central banker is indifferent between the two positions. q will adjust so that he is indifferent between the two positions. We can get this value from equations (13.26) and (13.27) and solving for q. This requires that The basic intuition behind the analysis is quite straightforward. The people are uncertain about the type of central banker they have. The lower the inflation they observe the more they are convinced that they have a Type I central banker. The greater the emphasis the central banker places on losses from future inflation, the more inclined will he be to pursue low inflation policies today. Discussion Thus there are these two approaches to removing the inflationary bias of monetary policy. We have two chores left to do in this chapter. First, we want to get a sense of whether there is a lower boundary for the rate of inflation that can be set. Is there a floor below which the rate of inflation cannot fall? If yes, zero inflation is ruled out. The second question is—is there a mechanism by which the central banker can be compelled to take into account the true preferences of the people about inflation? We turn now to these important questions. The first question was addressed by Barro and Gordon (1983). The second idea was initiated by Walsh (1995) and followed up by several others. Barro and Gordon argue that it would never be possible to have a zero inflation


331

rate and, therefore, costs from inflation will always be positive and expectations of inflation will never be zero. The essential reason for this is already embedded in the dynamic inconsistency model. Even when a rules regime is in place, there is always the possibility that the central banker will find it profitable to cheat and opt for positive inflation in order to exploit the inflation-unemployment tradeoff. The public is aware of this temptation facing the central banker. What makes springing such surprises profitable for the central banker? The first factor we have already commented upon is that the inflationunemployment tradeoff can be exploited. In the context of developing countries, in particular, there are other motives such as seignorage and lowering the real value of the government debt. Barro and Gordon write a loss function from inflation such that zero inflation is socially optimal: (13.28)

This loss term has two constituents. The first term represents the loss due to current inflation. The second represents the gain to be had by being able to exploit the inflation-unemployment tradeoff, i.e., driving a wedge between actual and expected inflation. The impact of the first constituent is constant over time—that of the second varies with time reflecting standard natural rate type arguments. In equation (13.28) the optimal rate of inflation is, clearly, zero. Barro and Gordon, in fact, assume a distribution for bt such that bt has a constant mean and variance:

The crucial point now is to minimise the discounted loss over a horizon by choosing the rate of inflation. This objective function can be written as: (13.29)

Equation (13.29) recognises the fact that the choice of the inflation rate today will influence the expected inflation in the future and, hence, the value of the loss function. rt is the discount rate between t and We assume that rt is a random variable and write the edicount factor as qt. Hence qt follows a distribution which fact we specify as: (13.30)

In the symmetric case of the model no one knows bt or qt when they make judgements about macroeconomic variables in period t. Policy makers choose πt without this knowledge and people form expectations of inflation (πte) without such knowledge.

332


Discretionary policy Suppose that a discretionary policy is in place. The loss minimising rate of inflation is simply,

Given that people have rational expectations they will also except inflation to be this same value. Substituting in equation (13.28) the fact that we will have societal loss from inflation simply equal to Policy under a rule With policy following a rule for πt, the chosen rate of inflation is zero with the result that the loss Now suppose the people believe that the policy maker will indeed follow a rules type policy so that expected inflation is zero. If the policy maker then cheats and sets value of loss can be lower at

then the expected

Therefore one can define a

temptation for the central banker to cheat. This temptation is the difference between the social loss when the central banker does not cheat and when he cheats—and the people expect him to follow a rule. This temptation is clearly

Hence there are three types of equilibria: 1 2 3

A cheating equilibrium with people expecting zero inflation. We have discovered that in this case, social loss is negative. A rules equilibrium where the people expect zero inflation and the central banker fulfils this expectation. Actual and expected inflation are equal in this case and social loss is zero. Discretionary equilibrium when social loss is positive.

Rational individuals know of the existence of these equilibria and form expectations accordingly. Barro and Gordon consider a mechanism through which people ‘punish’ the central banker for giving in to temptation. If, in the previous period, a central banker who promised zero inflation delivered on that promise, people would expect zero inflation this period as well. However, if the central banker did not deliver on this promise, they would change their expectation of inflation to what they know to be his ‘optimal’


rate of inflation:

333

This is one kind of punishment mechanism with the

punishment having only a one-period memory and lasting for only one time period. We will have to discover whether this punishment is adequate. Given this fact the central banker will choose his inflation strategy. The central banker would be indifferent between cheating and not cheating if the expected present value of the loss from the punishment is equal to the expected gain from this. If the central banker cheats, the loss will be incurred only in the next time period. This loss would have to be discounted to express it in today’s dollars. The expected value of this loss is

The policy maker

will abide by the rule of setting inflation equal to zero during period t if the loss is at least as great as the temptation. But people know whether the policy maker will find it worthwhile to cheat. Hence if the cheating solution is preferred to the rules policy then it is irrational to expect zero inflation. It is clear then that equilibrium in this system must satisfy two criteria: (i) Rationality of expectations must obtain. This implies that each individual’s expectation of inflation, is the best possible forecast of the policy maker’s actual choice of the rate of inflation, given the way the policy maker behaves and given the way others form their expectations. (ii) The policy maker’s choice of the rate of inflation maximises his objective function given the way people form their expectations. We know that the temptation is whereas the penalty for this transgression is

and since

it

follows that the temptation is always greater than the punishment that can be inflicted. Hence it would be irrational to expect zero inflation. They would expect a rate of inflation which, if it prevails, would make the temptation to violate it equal to the penalty that can be imposed for this transgression. Suppose this condition is satisfied for the rate of inflation π. From the loss function, the temptation is

transgression would be

whereas the penalty for the

When the penalty is greater than or

equal to the temptation the central banker would not have the temptation to cheat and expected inflation could be sustained. This can be shown diagrammatically as in Figure 13.6. Clearly both temptation and punishment are zero when

Setting

temptation equal to punishment we can also solve for another value of the

334


Figure 13.6

rate of inflation where the two are equal. This is

Henceforth this

is called the optimal sustainable rate of inflation. It is sustainable since both temptation and penalty are equal to each other so that the central bank has no incentive to cheat and the people realise that the central banker has no incentive to cheat. It is the optimal rate of inflation since this is the lowest rate of inflation where such sustainability obtains. Remember that social this rate loss is monotonically rising in the rate of inflation. Since of inflation is less than

However, it is not zero. It will be in the best

interest of the central banker to announce this sustainable optimal rate of inflation. Thus the optimal rate of inflation is collapses to the myopic solution and as

As

this

we are back at the zero inflation

rate solution. However, the basic message from the Barro and Gordon story is that a zero rate of inflation is untenable and costs from inflation will never be zero.


335

WALSH CONTRACTS Walsh (1995) raised the question of excessive inflation within the context of the incentives faced by the central banker. He conceived of a transfer function that embodies a disincentive for the central banker to deviate from the ‘rules’ solution. Suppose the government wants to minimise the standard quadratic cost function. The supply equation for output is the same Lucas type relation: (13.31)

Actual inflation is given simply by the sum of monetary growth and velocity shocks as in: (13.32)

where m is the log of money so that ∆m is the rate of growth of the money stock and v represents shocks to the velocity of circulation of money. The private sector’s expectations are assumed to be determined prior to the central bank’s choice of a growth rate for the nominal money stock. Therefore, in setting ∆m, the central bank will take πe as given. The central bank is assumed to observe e but not v prior to setting ∆m. It is further assumed that e and v are uncorrelated. We write the central bank’s objective function as (13.33)

Equation (13.33) incorporates two considerations. First, gain in output gives positive utility (this is the first term). However, inflation detracts from welfare and this is the second term in expression (13.33). Maximising equation (13.33) with respect to ∆m and taking e and πe as given, we will get the optimal discretionary monetary policy rule as: (13.34)

Given this policy, actual inflation will equal However, rational economic agents will realise that the central bank will set monetary growth according to equation (13.34) so their expectation of inflation will not be zero. In particular, their expectation of inflation will be:

Hence, average inflation is fully anticipated so that inflation produces no gains in output, which is given by This is the discretionary equilibrium of the central banker. In the ‘rules’ setting the central banker is able to convince everyone that he or she is going to set Hence, in this case, and expected utility would equal

where UR and U d are, respectively, utility under rules and discretion

336


respectively. An alternative specification of the central bank’s objective function is:

(13.35)

where If ∆m is chosen after observing the supply shock e, but before observing the velocity shock, v, the first-order condition for ∆m, conditional on e and taking πe as given is whence we have (13.36)

Equation (13.36) is different from that obtained with the linear objective function (13.34) in two important respects. First, this expression involves the supply shock whereas (13.34) does not. Since the central bank wants to minimise the variance of output around its target level, policy should be made conditional on the supply shock. Thus a role for stabilisation policy is admitted. Second, the optimal money supply growth rule involves agents’ expectation of inflation. In line with our discussion above, rationality of expectations requires that economic agents take (13.36) into account to form expectations of inflation. This expected rate of inflation can be written as Solving for πe we get which when substituted into (13.36) using (13.34) enables us to derive an equilibrium rate of inflation. This expression can be used to form expectations of inflation: (13.37)

where the superscript ‘d’ stands for ‘discretionary’. The average rate of inflation is aλk. Since this is fully anticipated, this has no effect on output. Under these conditions social loss is:

(13.38)

The unconditional expectation of this loss is: (13.39)

where the σ2s are the respective variances.


337

Now suppose that the central bank has been able to pre-commit to a policy rule prior to the formation of private expectation. The monetary authorities would like to respond to the supply shock, e. In this case, the policy rule would not be a simple fixed monetary growth rate rule. To make things concrete suppose that the central banker is able to commit to a policy rule of the form: so that Substituting into the loss function we have: (13.40)

The central banker chooses b0 and b1 to minimise the unconditional expectation of this loss function. From this we get: (13.41)

With this as background we can now broach the subject of Walsh contracts. This was first discussed in a seminal paper by Walsh (1995). Walsh conceives of an independent central banker who shares the government’s preferences, V, but also receives a monetary transfer payment, t, from the government contingent upon the performance of the central bank. This payment could be either a direct payment to the central banker or the budget of the central bank. The central bank’s utility rises when the transfer goes up and falls when the loss to society from inflation and unemployment goes up. Thus we write the central bank’s utility as (13.42)

The socially optimal money supply rule is given by (13.41). If the government can monitor the value of e ex post, there are several contracts that could result in policy rules of the sort noted in equation (13.41). For instance there could be large penalties for departing from (13.41). However, this ex post verification of e could be hard to do. Keeping this in mind, Walsh thinks of the transfer function as a function of the realised rate of inflation. Thus This transfer function implements the optimal commitment policy rule ∆mc(e). Hence the realised rate of inflation is πc(e) if πc cb maximises for all e where E [.] denotes the expected value of the central bank’s objective function conditional on the realisation of e. The first-order condition for the central banker’s problem can be solved for the discretionary monetary policy ∆mcb(e): (13.43)

where

338


The last term in (13.43) shows that the optimal discretionary policy response to the supply shock is equal to the response under the optimal commitment policy ∆mc. It is important to realise the significance of this. This implies that the government wants to design a policy that eliminates the inflationary bias while leaving the central bank free to respond with discretion to movements in e. Taking expectations in (13.43) we have Substituting this for pe in (13.43) we obtain:

where E[.] represents the public’s expectation and Ecb[.] the central bank’s. If the central bank’s monetary policy must equal the optimal commitment policy then ∆mcb(e) should equal ∆mc(e) for all e. This will be possible when the first three terms on the right-hand side of equation (13.43) are equal to zero. This will be the case when

We can now integrate to

find out the value of the optimal transfer: where t0 is the constant of integration. This can be set at a level high enough to ensure participation by the central banker. On average, when inflation is zero, the return to the central banker is high enough to ensure participation. With this transfer function in place the central banker is induced to eliminate the inflationary bias while still ensuring optimal stabilisation policy in view of the realisation of e. To understand why the transfer function takes this simple form recall that the inflationary bias under discretionary policy was aλk. Realise that this is a constant. This is the temptation of the central banker. So, if the marginal cost of inflation to the central banker can be raised, then there will be no incentive to pursue discretionary policy. This is exactly what the linear transfer function accomplishes. HOW USEFUL IS INFLATION TARGETING FOR DEVELOPING COUNTRIES? Many developed countries following the cue of New Zealand started inflation targeting. Some developing countries have also given it serious consideration. These include South Africa and Korea. However, a key issue in deciding upon inflation targeting is whether shocks are more from the supply or the demand side. Typically inflation targeting has been used as a policy mechanism to address spurts in inflation arising from demand shocks.


339


2

3 4

In the seignorage model studied in this chapter real GDP was not growing. Consider an amendment to this model when the real GDP is growing at a per cent per annum. Work out the seignorage maximising rate of inflation for this case. How would your answer change if a was negative? In line with Cagan’s seminal (1956) paper on hyperinflation, assume that people adjust their expectations of inflation not at one go but by a constant fraction of the difference between the actual and the expected rate of inflation. Hence we write: with Find an expression for the change in expected inflation as a function of actual inflation. Trace the dynamics of seignorage, expected inflation and the real stock of cash balances. High rates of inflation are also more variable. Discuss the consequences of the variability of high inflation for consumers and producers. Many long-term labour contracts are not indexed for inflation or are at best only weakly indexed. For instance, in many countries a cost of living increase adjustment to earnings is provided for only once a year. Hence during periods of inflation real incomes often decline. Conversely during periods of disinflation real incomes often rise. Aggravating these tendencies is the fact that in most countries the rate of income taxation is based on nominal incomes and is not indexed for inflation. Explain why this might be the case using numerical examples. How would income tax rates need to be indexed to insulate taxpayers from higher real tax liabilities simply because of inflation?

Seignorage in highly indebted countries In an interesting paper McPherson (1999) argues that developing countries by virtue of the fact that their financial systems are weak and their tax bases narrow, rely more heavily on seignorage and inflation tax as sources of government revenue. This is usually triggered off through fast reserve money growth. However, they do not manage to collect much revenue in this manner because of the ensuing high inflation—including hyperinflation—and the consequent sharp fall in the value of real cash balances. He argues that governments and central banks see three benefits of seignorage. (1) Monopoly rents accrue when official financial instruments are held by the private sector and rates below market rates, e.g., reserve requirements of financial institutions. (2) Fixed coupon bonds provide a gain to the issuer because their real value falls when inflation rises up to the point of their maturity. The issuer gains an amount equal to the difference between its face value at maturity and that value deflated by the change in prices over the life of

340


the bond. (3) Second and higher-order benefits associated with the generation of seignorage are reflected in the gain accrued or costs avoided by the monetary authority and government through monetary restrictions, e.g., exchange control. With managed exchange rates rising local prices provide the monetary authority with rents. At the same time, however, there are costs to accumulating revenues through seignorage. Resource costs are associated with printing and issuing of money and maintaining the stock of fiat money. There is a capital loss on fixed value liabilities. When a country makes a commitment to repay a foreign liability it undertakes some risks. World interest rates may rise or the exchange rate may depreciate. Both will raise the cost of debt service—the former by increasing the cost in foreign exchange terms and the latter by increasing the cost in terms of domestic currency. Further, high rates of inflation may also lead to currency substitution. Asset holders in developing and transition economies, e.g., Bolivia, Zaire, Argentina, Peru, Russia, Ukraine and Serbia, where inflation has been high, have rapidly moved away from domestic to foreign currency holdings. This has, no doubt, been aided by the increasing globalisation of financial markets. Currency substitution generates seignorage for the country whose currency is being held—most often the US. High rates of inflation have their own costs in terms of distorting relative prices, eroding real values of fixed returns and those on fixed nominal incomes or incompletely indexed contracts. In addition, high inflation often has often been associated with bouts of capital flight. Furthermore high inflation has been associated with phases of dropping output. For example, in Zaire over the period 1991 to 1994 reserve money increased by a factor of 1.36 million whereas prices increased by a factor of 4.2 million. In the meantime, output declined. Similarly, in Nicaragua over the period 1988 to 1990 reserve money increased by a factor of 296,000 whereas prices increased by a factor of 364,000. Again there was a drop in output. Furthermore, excessive reserve money creation and the ensuing inflation led to exacerbation of problems arising out of factors such as financial repression, overvaluation of the exchange rate and the like.

NOTES 1 2 3

Cagan also introduced the concept of adaptive expectations in this paper. This is so because the nominal interest rate and not its logarithm appears in the money demand equation. One has to make a distinction between anticipated inflation and indexed inflation. The first deals with how economic agents form expectations about inflation, the second to how the economic system adjusts—ex post facto—to inflation. Whether


4 5 6

341

such inflation is anticipated or not is immaterial to this since we are talking of inflation ex post facto. According to Feldstein (1998) for the OECD countries a reduction in the steadystate rate of inflation from 4 per cent to 2 per cent would lead to a welfare gain of 1 per cent of GDP. See, for example, Barro (1995), De Gregori (1992a, b) and Sarel (1996). These follow from the existence of wage and price stickiness which results in drop in output, administered price rigidities, balance of payment disequilibrium arising from exchange rate misalignments of the past and export shortfalls accompanying output losses and costs associated with longer than necessary tight monetary policy to re-establish credibility and change inflationary expectations.

14


INTRODUCTION In Chapter 12 we have examined the problem of the twin deficits in developing countries and tried to come to grips with the problems that this poses for macroeconomic management. In many developing countries, however, the accumulated stock of external debt is posing an independent and even more pressing problem. Indeed the problem of high external indebtedness in some poor countries has taken serious dimensions with external debt to GDP ratios of these countries climbing to several times their GDPs and debt servicing charges amounting to well over 25 per cent of their export earnings. The indication that some developing countries were accumulating too much foreign debt too fast was present as far back as 1967. However, the current wave of debt-related problems for developing countries really began with the quadrupling of oil prices in 1973. Non-oil-producing developing countries found themselves in a spot as the price of oil as well as manufactured imports from industrialised countries skyrocketed. At the same time commercial banks were flush with funds deposited largely by industrialised countries and OPEC in the aftermath of such high inflation. Non-oilproducing developing countries reeling under the impact of such high prices borrowed heavily both from the banks as well as multilateral institutions such as the IMF in order to finance imports. Current account deficits and external debts soared. This debt accumulation was reported in the 1979 World Debt Tables of the World Bank, which noted ‘lagging debt payment’ on official loans to poor countries, although ‘debt or debt service forgiveness has eased the problems for some’. As time went by, the problem of the accumulation of debt of the poorest countries became more severe and erupted as a major crisis. It would not be an exaggeration to say that the growth prospects of some of the poorest countries of the world have been severely damaged by the existence of such large stocks of debt. In this chapter we will explore some issues connected with this debt crisis.

344


WHICH COUNTRIES BELONG TO THE HIPC GROUP? Global events in the 1970s and 1980s—particularly the oil price shocks, high interest rates and recession in developed countries and then weak commodity prices are usually referred to as the major contributors to debt explosion in the developing countries (IMF, 2000a). In the 1970s, on the demand side there was a pressing need in oil-importing developing countries for foreign exchange in order to finance balance of payments deficits and public projects following the sharp increase in the price of oil in the 1970s and the subsequent inflation in developed countries. Domestic factors such as high trade and budget deficits, low savings rates and poor project selection exacerbated the foreign exchange problem for developed countries. On the supply side the recessionary conditions in the developed countries made it lucrative for international banks to recycle their huge petro-dollar deposits in the developed countries. High levels of real interest rates in developing countries also aided this decision. This surge in external debt was accompanied by a change in the structure of the debt in response to the increased risk that accompanied the repayment difficulties that many countries faced. In the 1980s the debt crisis was propelled not only by the overborrowing of the 1970s but also by deterioration in the terms of trade of developing countries and high interest rates. Other contributing factors include lack of careful management of the currency composition of the debt, the absence of sustained adjustment policies, particularly when facing exogenous shocks, which gave rise to sizeable financing needs, the lending and refinancing policies of creditors, particularly lending on commercial terms with short repayment periods by many creditors in the late 1970s and early 1980s and non-concessional rescheduling terms for most of the 1980s, and political factors such as civil war and conflict. For a detailed discussion see Brooks et al. (1998). In recent years, the external debt situation for a number of low-income countries has become extremely difficult, prompting the IMF and the World Bank to design a framework in 1996 to provide special assistance to the heavily indebted poor countries (HIPC). There are 41 HIPCs, including 31 HIPC-LeDCs, which meet the three criteria to qualify for the enhanced initiative. These criteria are (i) a country is only eligible for highly concessional assistance (IDA); (ii) it has an IMF poverty reduction and growth facility supported-programme (PRGF) in place; and (iii) it has agreed to a rescheduling of debts on concessional terms with the Paris Club. According to the UNDP, the most impoverished and vulnerable countries of the world are grouped under the category of ‘least developed countries’ (LeDCs). Most, but not all, LeDCs are heavily indebted. Furthermore, there are some heavily indebted countries that do not belong to the LeDC category. What are LeDCs and what is an HIPC? A country is designated as a least developed country if it meets inclusion thresholds on the following three criteria (UNCTAD, 2000):


1 2

3

345

A low income: income to be below a GDP per capita of US$800. Weak human resources, measured by the Augmented Physical Quality of Life Index, which is based on indicators of life expectancy at birth, per capita calorie intake, combined primary and secondary school enrolment, and adult literacy. A low level of economic diversification, measured by the Economic Diversification Index, which is based on the share of manufacturing in GDP. The share of the labour force in industry, annual per capita commercial energy consumption and UNCTAD’s merchandise export concentration index.

The classification of HIPCs seems to be based on a rule of thumb rather than on clear-cut quantitative criteria. In 1996, when the category was introduced, the group of HIPCs consisted of 32 severely indebted low-income countries and nine other countries. To be classified as severely indebted in 1996 a country should have had 1 2

Present value of debt service to GDP to exceed 80 per cent, or Present value of debt to exports to exceed 220 per cent;

Two other common denominators of this group are that the countries only borrow on highly concessional terms from the World Bank (from the Bank’s International Development Association, IDA) and they have negotiated, or are prepared to negotiate, a concessional rescheduling with the Paris Club. Since 1996, the original group of 41 HIPCs has gone through some changes. Nigeria was soon ruled ineligible, as it also borrowed from the World Bank’s non-concessional window, the International Bank for Reconstruction and Development (IBRD). In 1999, Malawi was included and in the summer of 2000, Gambia was added as well. Equatorial Guinea was declassified as an HIPC in early 2000, as, with the onset of oil production, GDP levels rose above those required for IDA-only assistance. In Table 14.1 we list the countries that belong to the various poverty/ indebtedness categories. Thus even though LeDCs are, by definition, very poor and many are also heavily indebted, not all LeDCs are classified as heavily indebted poor countries (HIPCs). The World Bank and the IMF currently classify 41 countries as HIPCs. The group of 49 LeDCs shows considerable overlap with these HIPCs. More specifically, 31 out of 49 LeDCs are HIPC-LeDCs and ten HIPCs are non-LeDC but HIPCs. Eighteen LeDCs are not classified as HIPCs. Clearly, the classifications of HIPCs and LeDCs are rather arbitrary. Not only because several non-HIPC-LeDCs have unsustainable debts because all HIPCs are very poor and underdeveloped, even though they do not meet the criteria to fit in the LeDC category.

346


Table 14.1 Categorisation of countries by external debt and poverty status

WHY IS DEBT RELIEF NEEDED FOR THE POOREST COUNTRIES? One of the main reasons why development efforts have failed in LeDCs is that instead of investing in social and economic development, these countries are forced to use scarce government resources to finance external debt to foreign creditors. For example, in Burkina Faso, where one out of five children dies before the age of five, in 1998 the government spent as much on debt as on health. And in Niger, where 78 per cent of adult males and 93 per cent of adult women are illiterate, debt service amounted to 3.1 per cent of GNP in 1998, while spending on education was only 2.3 per cent of GNP. Hjertholm (1999) cites considerable empirical evidence suggesting that there is a strong and significant negative relationship between high debt burdens and poor economic performance, such as low growth, investment and human development. There are several channels through which these occur. But of these, two are particularly important: (i) cash flow effects arising from reduced public expenditures, and (ii) disincentive effects associated with a large debt overhang. Cashflow effects. There is evidence to suggest that public expenditures crowd in private investment especially when it comes to loosening the grip of structural bottlenecks and weak infrastructure. Weak infrastructure and low human capital act as significant brakes on development for many poor countries (Diaz-Alejandro, 1981; Taylor, 1983, 1993). If public expenditures are squeezed then these opportunities for growth are lost. A related cash flow problem associated with public debt service is import compression, which can occur for two reasons (Ndulu, 1991). First, since many developing countries have poor capital goods, production is low, they may find it difficult to substitute for imported capital goods. Thus a brake on imports will lead to a shortfall in investment. Second, as Hemphill (1974)


347

and Moran (1988) have argued, import compression can occur in cases where import volumes are determined by import capacity rather than relative prices. Clearly the magnitude of debt service matters for import capacity in such instances. Import compression can occur both at the balance of payments level and at the budgetary level (through the effects of public debt service on the import-content of government expenditures). Reductions in the import capacity of the government, as a result of debt service, can thus reduce government investment activity, whereby the complementarity effects are lost. Hjertholm (1997) presents evidence to suggest that such cash flow effects have indeed been at work in 23 indebted low-income sub-Sahran African countries. In addition to cash flow effects are disincentive effects associated with debt overhang. Since governments know that the beneficial effects of economic reforms will largely go to pay external debt, they have little incentive to pursue such reform. Affected by such disincentives are tax reforms as well as investment that cannot be buoyant in a climate of macroeconomic stability. A public debt overhang can affect macroeconomic stability by increasing the fiscal deficit, depreciating the exchange rate too fast, monetisation of deficit and reliance on seignorage and recourse to exceptional financing such as payments arrears and debt rescheduling, which tends to exacerbate uncertainty about the future debt servicing profile of the public sector. Hence tackling the debt crisis is extremely important for these countries.1 In Table 14.2 we give some indication of the burden of the external debt on these countries. Table 14.2 clearly indicates the weight of the burden of debt on the poorest countries. In 1998, the total debt stock of LeDCs (as defined above) amounted to US$154 billion. This is almost four times as high as the corresponding figure in 1980. For every single LeDC, the debt stock shows a steady and significant increase since 1980. The majority of the total debt stock is owed by the 31 HIPC-LeDCs with their share being 85 per cent or US$129 billion of the total LeDC debt. Some LeDCs, such as Mozambique, Angola and the Solomon Islands, owe a significant amount of debt to commercial banks and other private creditors. But for the majority of LeDCs, the main part of the total debt stock consists of official debt, i.e. bilateral debt owed to governments and multilateral debt owed to multilateral financial institutions. In 1998, 40 per cent of the LeDC debt stock was multilateral debt and 38 per cent bilateral debt. The remaining 22 per cent included private debt and related interest arrears as well as short-term official debt, plus interest arrears. Debt stock indicators have also worsened for most HIPC-LeDCs since 1980. As a result of rising debt levels, the present values of debt-exports ratio and of the debt-GNP ratio have increased since 1980 for most HIPCLeDCs, even though exports and GNP show some modest growth. However, for a number of HIPC-LeDCs, debt stock indicators improved from

Table 14.2 Characteristics of the debt profile of developing countries

Source: Computed from data sources from the World Bank and the IMF.

350


1980–98. Compared to the performance of other HIPC-LeDCs, these countries’ GNP and export income have improved significantly between 1980 and 1998. In other words, the improved debt indicators are to be attributed to improved economic performance rather than lower absolute levels of debt. In spite of this, however, these countries’ debts still remain far above the conservative World Bank and IMF sustainability threshold of a 150 per cent present value of debt-export ratio, as is the case for all the HIPC-LeDCs. Also, for 23 out of 28 HIPC-LeDCs for which data are available, the present value of debt-GNP ratio is above 80 per cent, the threshold used by the World Bank in the past to classify a country as ‘severely indebted’. The group of non-HIPC-LeDCs shows major differences in the level as well as the evolvement of debt indicators. Five out of 14 non-HIPC-LeDCs for which 1998 data are available, have a debt level that exceeds the 150 per cent present value of debt-export threshold. In one more country, Samoa, the present value of debt-GNP exceeds 80 per cent. Furthermore, for seven out of the 14 non-HIPC-LeDCs for which data are available, debt indicators are worse in 1998 than in the 1980 and 1990 levels, even though GNP and exports have increased compared to 1980. The debt indicators have improved for only two countries, Cambodia and former HIPC Equatorial Guinea. However, Cambodia’s present value of debt-export ratio is still far above the 150 per cent threshold. In addition to these debt indicators, the amount of arrears is indicative of the severity of the debt burden. A large number of LeDCs have huge arrears— in particular Congo, Liberia, Somalia and Sudan. In fact nine LeDCs have arrears of over 20 per cent of their total debt stock, and in the case of a further six LeDCs, these arrears amount to more than 15 per cent of the debt stock. There are three non-HIPC-LeDCs for which arrears are extremely high: Cambodia (43 per cent), Comoros Islands (22 per cent) and Equatorial Guinea (45 per cent). The amount of arrears clearly demonstrates that many LeDCs are simply incapable of paying their debt service obligations. For at least 18 LeDCs, including four non-HIPC-LeDCs, annual debt repayments decreased in 1998 compared to 1990, even though the debt stock increased. The repayments made by LeDCs during 1980–98 are given in Table 14.2. In 1998, these countries returned a total of US$4,740 million to foreign creditors. HIPCLeDCs account for 80 per cent of this total. The main receivers of LDC debt repayments are multilateral creditors. It should be pointed out, however, that in addition to the 31 HIPC-LeDCs, many LDCs also have unsustainable debts, even with the rather arbitrary and narrowly defined definition of the World Bank and IMF. Six non-HIPCLeDCs have debt levels that exceed the WB and IMF debt sustainability thresholds: Bangladesh, Cambodia, Comoros Islands, Haiti, Nepal and Samoa. Furthermore, Equatorial Guinea’s amount of arrears equally suggests that this country is unable to carry its debt burden. Tackling the debt crisis is


351

essential for these countries. If creditors fail to significantly reduce the debt level of LeDCs, social and economic development and poverty reduction remain a mirage. A question that needs to be raised is: how sustainable are debt levels of HIPC countries given the high levels of poverty that they have? Debt sustainability has been assessed only from the perspective of repayment capacity and the implications of such repayment for the domestic economy have rarely been considered. Some computations suggest that once adjustment is made for their high levels of poverty, most, if not all, HIPC countries should qualify for debt write-off. HOW HAVE DEBT REDUCTION STRATEGIES WORKED? Debt reduction efforts began with the 1977–79 UNCTAD meetings. These led to official creditors writing off $6 billion in debt to 45 poor countries. The measures by official creditors included ‘the elimination of interest payments, the rescheduling of debt service, local cost assistance, untied compensatory aid, and new grants to reimburse old debts’. The 1981 Africa report by the World Bank (usually known as the Berg Report) noted that Liberia, Sierra Leone, Sudan, Zaire and Zambia (all of which would become HIPCs) had already experienced ‘severe debt-servicing difficulties’ in the 1970s and ‘are likely to continue to do so’ and that ‘debt-service payments have risen to the point at which a number of countries face critical situations’. The Berg Report hinted of debt relief, namely ‘longer-term solutions for debt crises should be sought’ and ‘the present practice of [donors] separating aid and debt decisions may be counterproductive’. The 1984 World Bank Africa report, however, was more demanding. It argued ‘where monitorable programmes exist, multiyear debt relief and longer grace periods should be part of the package of financial support to the programme’. The wording got even stronger in the World Bank’s 1986 Africa report: low-income Africa’s financing needs will ‘have to be filled by additional bilateral aid and debt relief’. The World Bank’s 1991 Africa report continued escalating the rhetoric: ‘Africa cannot escape its present economic crisis without reducing its debt burden sizably.’ In addition, the June 1987 G-7 summit in Venice called for interest rate relief on debt of low-income countries. The World Bank noted ‘the past year has brought increasing recognition of the urgency of the debt problems of the low-income countries of Sub-Saharan Africa’. One year later, the June 1988 G-7 summit in Toronto agreed on a menu of options, including partial forgiveness, longer maturities and lower interest rates (these became known as the ‘Toronto terms’). Meanwhile, in order to help African countries service their official debt, the World Bank in December 1987 initiated a Special Programme of Assistance (SPA) to low-income Africa. The IMF complemented the SPA with the Enhanced Structural Adjustment Facility

352


(ESAF). Both programmes sought to provide ‘substantially increased, quickdisbursing, highly concessional aid’. The economics literature started noticing low-income African debt at about the same time. See Humphreys and Underwood (1989), Husain and Underwood (1991), Lancaster and Williamson (1986), Mistry (1988), Nafziger (1993), and Parfitt and Riley (1989). More recent analyses include Brooks et al. (1998). The 1990 Houston G-7 summit considered ‘more concessional reschedulings for the poorest debtor countries’. The UK and the Netherlands proposed ‘Trinidad terms’ that would increase the grant element of debt reduction to 67 per cent, from 20 per cent under the ‘Toronto terms’. The 1991 London G-7 summit agreed ‘on the need for additional debt relief measures…going well beyond the relief already granted under Toronto terms’. In November 1993, the Paris Club (the club of official lenders) applied Enhanced Toronto Terms that involved even more concessions. In December 1994, the Paris Club announced the ‘Naples Terms’ under which eligible countries would receive yet additional debt relief. Despite these efforts, many LeDCs have been forced to enter a seemingly endless cycle of debt restructurings, which, up to the year 2000, had not brought these countries to a sustainable debt level. The severity of today’s debt burden demonstrates how disappointingly inadequate official creditors’ response has been in the past to the debt crisis. Or, as UNCTAD (2000) puts it: ‘There has been a persistent tendency to underestimate what has been needed, which has in itself contributed to the build up of the debt.’ Attempts to significantly reduce the burden of bilateral debt through Paris Club debt negotiations have failed, mainly because of the exclusion of a large part of the debt, as the well-known case of Uganda demonstrates. In 1995 Uganda received a 67 per cent stock reduction, but mainly because previously rescheduled debt was excluded, its US$3.2 billion debt stock decreased by only 3.2 per cent. An important lacuna in these arrangements was that the multilateral debt problem was not even taken seriously. Even though the main part of the poorest countries’ debt repayments went to multilateral creditors, the World Bank and IMF denied for many years that there was a multilateral debt problem. It was only in 1996 that the two Bretton Woods institutions openly acknowledged this, when they introduced the HIPC Initiative. The HIPC Initiative was designed to allow the poor countries to ‘exit, once and for all, from the rescheduling process’ and to resume ‘normal relations with the international financial community, characterized by spontaneous financial flows and the full honoring of commitments’. The multilateral lenders for the first time would ‘take action to reduce the burden of their claims on a given country’, albeit conditional on good policies in the recipient countries. ‘Good policies’ often referred to an agreed upon poverty reduction programme. This has often been crystallised in the form of a Poverty Reduction Strategy Paper (PRSP), which the government in each such country


353

had to produce to the satisfaction of lending agencies—particularly multilateral lending agencies. The Paris Club at the same time agreed to go beyond Naples Terms and provide an 80 per cent debt reduction in net present value terms. By September 1999, debt relief packages had been agreed for seven poor countries, totalling more than $3.4 billion in debt relief in net present value terms. The seven countries were Bolivia, Burkina Faso, Côte d’Ivoire, Guyana, Mali, Mozambique and Uganda. According to the Bank’s web site, in addition ‘Ethiopia, Guinea-Bissau, Nicaragua, Mauritania and Tanzania have completed a preliminary review and could qualify for billions more in debt relief.’ There were renewed calls in 1999 for expansion of this programme. In common parlance the HIPC initiative in 1996 is known as HIPC I and the one in 1999 as HIPC II. Besides explicit debt relief, there also has been an implicit form of debt relief going on throughout the period—the substitution of concessional debt for non-concessional debt. It is remarkable that the net present value of future debt service for HIPCs rose throughout the period despite the large net transfers of resources from concessional lenders like the International Development Association of the World Bank and the concessional arms of bilateral and other multilateral agencies. The necessity to provide continuing waves of debt relief, one after another, may suggest that there is a problem with the way debt relief is structured. WHEN IS DEBT UNSUSTAINABLE? In Chapter 12 we have examined the notions of sustainability of internal and external debt. The definitions used there relied on the satisfaction of an intertemporal budget constraint. In the parlance of current international creditors of the poorest countries, however, the definition of debt sustainability used is whether or not a country can meet its current and future external debt servicing obligations in full, without recourse to further debt relief, rescheduling or accumulation of arrears, and without unduly compromising growth. If such conditions are not satisfied, it is felt the country in question would not be able to obtain a permanent ‘exit’ from the rescheduling process. This would mean that this country would never be able to re-establish normal financial ties with the rest of the world. From a developmental perspective, this definition is quite narrow— particularly since it does not deal with domestic debt sustainability. Furthermore, it does not address adequately the domestic resource cost of servicing the debt. In any event, the definition of sustainability used by international agencies encompasses the following: • • •

Net present value of debt-exports ratio of less than 150 per cent; Net present value of debt-government revenue of less than 250 per cent; Debt-service-exports ratio of less than 15–20 per cent.

354


Many authors have argued that these indicators do not truly reflect the burden of debt. First of all, the emphasis is on debt stock rather than debt service. Furthermore, many LeDCs have accumulated such large debt stocks that only a fraction of it is actually being repaid. It is further argued that from a current resources perspective it is debt servicing that counts and the focus should be on exports. But even though exports are an important source of foreign exchange, which is needed for debt repayments, the debtexport ratio can be misleading as rapid growth in exports does not always translate into more budgetary resources for the government. In addition, the volatility of currency and commodity markets also makes the debtexport ratio an unreliable benchmark to predict debt sustainability in the medium term. In view of this, some authors have proposed the use of an alternative fiscal criterion since it takes the amount of resources available to the government as a starting point. This point of view has been touched upon in Chapter 12. But there are difficulties with the application of such criteria as well. An important reason for this is that creditors view sustainability only from a debt stock perspective and not from a flow perspective. The present value of debt-export ratio is equally inappropriate. According to the World Bank and the IMF, the empirical threshold of the net present value of debt-export ratio should be 200 per cent. The institutions have decided to lower this to 150 per cent to ‘provide more of a cushion from exogenous shocks and to free up resources for poverty reduction’ (IMF, 2001a). However, the data used to derive the threshold of 200 per cent mainly come from middle-income countries, making no special allowance for the special situation of the LeDCs or HIPCs. Research suggests that the historical present value of debt-export threshold level for HIPCs is 140 per cent. In essence, then, the debt sustainability indicators being used by the creditors are inappropriate, arbitrary and conservative. SOME PROBLEMS WITH THE HIPC INITIATIVES It would appear obvious that private lending would withdraw because of the poor creditworthiness of HIPCs. Hence the process of debt relief has led to a substitution of official lending for private lending and foreign direct investment. Some economists have expressed the fear that official lending may not have followed the same standards of creditworthiness as private lending. In addition, there has been a redistribution of roles even among official lenders, with some agencies making net transfers (debt flows net of interest) to HIPCs and others receiving net transfers from HIPCs. However, these problems pale into insignificance when compared with the problems facing the HIPCs. As is evident from the figures presented above, even with the HIPC initiatives, the debt obligations of the poorest countries are still very severe. There is a popular call supported by some


355

economists such as Jeffrey Sachs for writing off the debt of the poorest countries. At the same time a number of economists have objected to these proposals. Their view is that many LeDC governments are corrupt and would use the funds released by debt relief for the personal use of those in power. This is a tricky issue since both debt relief and its appropriate use and follow-up are important. If debt writeoff is indeed contemplated, LeDC governments would need to accept far greater interference by multilateral agencies in their general administration and governance than they do at present. Some authors have argued that the granting of progressively more favourable terms for debt relief may also have perverse incentive effects, as countries borrow in anticipation of debt forgiveness and delay policy reforms waiting for the best deal. Burnside and Dollar (1997) and World Bank (1998) suggest that aid does not raise growth in countries with poor economic policies. The World Bank’s Africa report (World Bank, 1994b) suggested that many African countries failed to depart from poor economic policies during the process of receiving adjustment loans from the World Bank and International Monetary Fund. What is the evidence on the misuse of funds by HIPC-LeDC governments? It is impossible to indicate to what extent decreased debt service requirements contribute directly to additional spending on social and human development. First of all, it is difficult to estimate how many resources are actually freed due to debt relief, particularly since many HIPC-LeDCs in the past did not pay all of the outstanding debt service. In addition, two countries—Niger and Zambia—will pay more debt service in the years 2001–05 than they did in 1998–99, implying that no funds are freed at all. Second, increased social spending cannot be related directly to decreased debt service. HIPC-LeDCs have always borrowed money to pay part of their debt service. Consequently, it is impossible to increase spending on social and human development with the same amount as the debt service is lowered without new borrowing or grants. Increases in spending on social and human development cannot therefore be attributed solely to debt service relief, but must also be attributed to new grants and loans. Thus for a number of countries, increased spending on social services in 2001–02 is much higher than the amount of resources ‘freed’ by debt relief. Furthermore, for Zambia and Niger, whose debt service due after 2000 will increase compared to that paid in 1998, spending on social development will nonetheless increase after 2000. Hence, the statement that ‘corrupt’ HIPC-LeDC governments will necessarily misuse funds made available through debt relief does not find overwhelming empirical evidence. It follows from this that cancellation of HIPC-LeDC debt may not be as bad an idea as may appear at first glance. What is the numerical significance of the implied burden on multilateral agencies of following such a strategy? There is evidence to suggest that for the World Bank, cancellation of HIPC

356


debt would be possible with the prudent use of a small proportion of the Bank’s reserves and a continuing commitment from its net income. Greater bilateral donor support and bigger contributions from the World Bank to the HIPC trust fund will be needed to cancel further regional multilateral development bank loans such as those from the African Development Bank. In the case of conflict-ridden LeDC, it is clear that the World Bank and IMF should make substantial efforts to support them. These countries have huge arrears with no prospects of getting paid in the foreseeable future. It certainly makes sense to have these arrears cancelled now instead of their piling up and being amortised and then written off as accumulated debt in a later period. That, certainly, would be a farcical exercise. To be sure, the World Bank and the IMF have produced several papers exploring different options to help these countries. Since these proposals depend on bilateral donor contributions, and these have hardly been forthcoming, they have not materialised. It is evident to the serious observer that for preventing LeDCs from falling into the same debt trap in the future as they are now, much deeper debt relief would be needed. In addition, current borrowing and lending practices should be changed. LeDCs remain dependent on external support, and it is essential that this support is on proper terms and that it is properly used. Irresponsible borrowing could be prevented by (i) making sure the internal disbursement of external finance is carried out in an accountable and transparent manner; (ii) the use of funding is consistent with an appropriate poverty reduction strategy; (iii) a monitoring system, involving civil society and parliament, is put in place to allow for the monitoring of internal disbursement at the national, regional and local level. In addition to responsible use of resources, responsible borrowing and lending requires that the type of finance is adjusted to its spending purpose. For example, finance for non-income-generating purposes, such as health, should preferably be on a grant basis, and finance for directly income-generating projects, such as the building of factories, can be on a concessional loan basis. Some other problems with the HIPC-LeDC initiative are as follows, (i) Several LeDCs with significant debt burdens are not included in the initiative. For instance, Angola is considered to have a sustainable debt burden, even though all its 1998 debt indicators are above the threshold level and Angola’s arrears almost quadrupled from US$700 million (or 8 per cent of total debt) in 1990 to US$2,704 million (22 per cent of total debt) in 1998. Furthermore, there are seven LeDCs that do not meet the criteria to be designated an HIPC but which do have unsustainable debts according to the rather conservative World Bank and IMF criteria. These countries are not even considered for the initiative, even though they are heavily indebted and extremely poor, e.g., in Bangladesh, 30 per cent of the population lives below US$1 per day. (ii) Debt service levels will start to rise again. For eleven out of the thirteen HIPC-LeDCs for which post-2005 debt service data are available,


357

debt service will start to increase after 2005 and for nine HIPC-LeDCs future levels are predicted to be far above present debt service levels. Thus, it appears that in the intermediate run most HIPC-LeDCs will be back where they are now. (iii) As was pointed out by the US General Accounting Office, debt reduction only frees resources if HIPCs continue to borrow at the same rate as in the past (GAO, 2000). In effect, increased spending on poverty reduction is now being financed by new debt that will come due in the future. For example, the World Bank is funding HIV/AIDS programmes with loans rather than grants, despite the fact that these loans will not generate new productive capacity with which to pay them off.2 The World Bank and the IMF forecast that debt levels will be sustainable by 2018. If this were to happen then this must be the result of better economic growth. These projections appear very optimistic, if not unrealistic, and it appears unlikely that these countries will live up to these expectations. For example, for eligible HIPCs that have reached the point where they could meet the criterion of debt sustainability, the average annual real GDP growth rate is projected to increase from the 3.1 per cent achieved in 1990–99 to 5.6 per cent in 2000–10 and this high growth rate is expected to be sustained for fifteen years. The implied projected export figures are much higher than their ten-year historical averages: the average export growth in 1990–99 was 4.2 per cent, and the average projected growth for 2000–10 is 8.9 per cent, which implies export performance more than doubling. There is simply no doubting the fact that the figures are overly optimistic. Actually, there is reason to think that export earnings will come under increasing pressure in future years. LeDCs and HIPCs export earnings are concentrated in a few primary commodities and most of these countries depend on just two or three exports. This makes their economies very vulnerable to fluctuations in commodity prices, and in recent years commodity prices have fallen sharply. Furthermore, no account is taken of the likelihood that, as many countries increase export production in the absence of a simultaneous upswing in world demand, commodity prices will fall, exacerbating the shortfalls in export earnings of LeDCs. In addition, LeDCs face increasingly difficult conditions because of (i) fluctuations in the price of important products such as petroleum, (ii) exchange rate devaluations and consequent impact on domestic prices of essential imports, and (iii) non-economic shocks through weather and the occurrence of HIV/AIDS epidemics on a large scale. In addition, LeDCs carry the burden of unfulfilled promises made by Paris Club creditors which they made at the G-7 in Cologne in 1999 when they announced that in order to meet the sustainability ratios, they would cancel pre- and postcutoff non-aid debt. Reduction of aid debt was purported to be in addition to this, and would therefore reduce the ratio below the thresholds. However, for Uganda, the first country to reach the completion point, Paris Club creditors failed to come to an agreement on this. The debt relief that was finally agreed upon was left to the creditors themselves to decide. It is likely

358


that the final reduction may not be sufficient to redress the problems of these countries—primarily because domestic debt is not included in the World Bank and IMF’s debt sustainability analyses, even though some countries have significant domestic debt burdens. In many LeDCs, domestic debt repayments amount to 20 per cent of the government budget. On the one side, the need to develop and implement a poverty reduction strategy (PRS) for one year postpones debt relief, while on the other side the rush to achieve the completion point diminishes the quality of the PRS. The current linkage is thus damaging to both processes and therefore other ways should be explored to link debt reduction to poverty alleviation. The next section describes alternative ways to ensure that debt reduction is really linked to social and human development. CONCLUSIONS The problem of high debt in developing countries is difficult to manage. Some arguments for debt reduction have focused on development of an appropriate strategy of poverty reduction. Others, for example, Jha and Schatan (2001), have argued that there may be a role in linkage to environmental issues. However, it appears that the scale of the HIPC indebtedness problem is so large that drastic reduction of debt should be serious and deep and bring the HIPC to a sustainable debt growth path using reasonable assumptions. It remains to be seen whether the current international arrangements would be able to rise to this challenge. DISCUSSION QUESTIONS 1

2

An important reason why many developing countries face large external debts is because of stagnating prices for their exports. For instance, the price index of metals—a leading export of many LDCs—has both fluctuated and had a downward trend. Using the price index of metals for the period 1992–2001 are given in the following table:

Source: IMF. Examine the links between episodes of accumulation of debt and commodity prices for one LDC—say copper for Zambia. A number of organisations analyse the LDC debt situation. For instance, see the proceedings of a conference in ESCAP on their website http:// www.unescap.org/drpad/projects/fin_dev/issuenote6.htm. In light of the arguments in this paper comment on the possibility that trade constitutes an engine of economic growth for the poorest countries.


359

The institutional response to LeDC indebtedness—the World Bank The debt crisis of the 1980s prompted the International Bank for Reconstruction and Development (IBRD) to make market-based adjustment loans. Most IBRD development activities fall under the category of investment lending for projects or programmes. The Bank also provides ‘adjustment lending’ designed to support fundamental changes in economic and financial policies of member countries. This type of lending emerged with the onset of the debt crisis of the 1980s. The IBRD typically approved ‘Structural Adjustment Loans’ and ‘Sector Adjustment Loans’ aimed at liberalising domestic and foreign trade and privatising public enterprises. These were market-based measures policy makers felt were necessary to reform inefficient, state-dominated economies—which developed prior to the debt crisis thanks in part to World Bank support. More recently, the IBRD approved billions of dollars in adjustment loans to Indonesia, Korea and Thailand to help those countries restructure their financial and corporate sectors in the wake of the Asian financial crisis. Now the Bank is trying to improve its responsiveness while continuing to stress poverty alleviation. In the 1990s the World Bank came under criticism from many quarters. Critics claim it is a top-down, unresponsive institution that is out of touch with grassroots development realities in member countries. In response, the Bank has formed an Inspection Panel to monitor its compliance with its own policies. It has also embarked upon a reform programme called the Strategic Compact. The programme’s objectives include improving the effectiveness of its lending and non-lending services; improving its responsiveness to client needs; diversifying its products and services; and reducing overheads while decentralising its activities. In the meantime, the Bank has opened up another front in its antipoverty war via the HIPC, inaugurated in autumn 1996. Developed along with the IMF in response to worldwide activism, the Initiative’s purpose is to enable poor (mainly African) countries pursuing marketbased policies to rid themselves of their unsustainable debt burdens. The HIPC Trust Fund provides debt relief by prepaying a portion of debt owed to multilateral institutions, or purchasing and cancelling a portion of the debt, or paying debt service as it comes due. Uganda became the first beneficiary with a debt-service relief package in April 1998 of $650 million. Only a few other countries have qualified thus far. In response to criticisms that the Initiative has failed to provide adequate debt relief, the IMF Executive Board in August 1999 agreed in principle to IMF/World Bank-proposed modifications intended to bring speedier, broader and deeper debt relief. Soon, other World Bank institutions got involved in debt relief. These include the IFC, the MIGA and the ICSID. The International Finance Corporation (IFC), formed

360


in 1956, promotes private sector investment in poor countries that would otherwise not easily attract private investment. It does this by providing long-term market-priced loans and equity financing for private sector projects. The IFC’s participation in a project acts as a seal of approval, encouraging other private investors to become involved. Established in 1988, the Multilateral Guarantee Agency (MIGA) is the most recent addition to the World Bank Group. MIGA is an investment insurance agency that encourages foreign direct investment in developing nations. It does this by providing guarantees against political (non-commercial) risks. In this way, investors are more willing to invest in countries that may not be politically stable. The International Center for Settlement of Investment Disputes (ICSID) was established by treaty in 1966 in order to provide a forum for arbitration or mediation of disputes between foreign investors and their host countries. Its purpose is to promote increased flows of international investment by providing a forum outside the host state for settlement of investment disputes. Parties cannot be forced to use ICSID conciliation and arbitration services. But once they have consented to arbitration under the ICSID Convention, neither can withdraw unilaterally. The work of the World Bank Group is intended to facilitate foreign private investment in one form or another. Such investment is frequently critical to the economies of host countries. However, many people and organisations believe the benefits of investment fail to reach society as a whole. In fact, many believe foreign investment frequently increases the gap between the poor and the rich.

NOTES 1

2

In a recent paper Dijkstra and Hermes (2001) argue that large debt overhang may impede economic growth in LeDCs. This hypothesis states that with high levels of debt, the government has no incentive to carry out macroeconomic and structural reforms since the beneficial effects of such reforms will have to be passed on as debt service payments to foreigners. Thus the economy gets caught in a low productivity-high debt trap with each feeding on the other. Dijkstra and Hermes (2001) show further that the problem of the uncertainty of debt service payments—in addition to the level of the external debt as such—may compromise economic growth in HIPC. This is because the large variations in debt service payments from year to year may make it difficult for the governments in the HIPC to know exactly how many resources are available for debt servicing and how many for economic reforms. They show that this uncertainty seriously compromises the ability of these governments to carry out serious economic reforms. Thus large amounts of debt seriously compromise poor countries’ ability to pursue economic policies to reduce poverty and encourage economic growth thus condemning to HIPC to seemingly unending poverty. For some of these reasons, it soon became apparent that the original initiative failed to reduce debt service burdens. Creditors have recognised some of the original initiative’s flaws and introduced a number of changes, culminating in


361

the ‘enhanced initiative’, or HIPC II Initiative. This initiative was agreed by the G-7 in June 1999. This initiative is divided in two stages. During the first stage, the country establishes a three-year track record of good performance on an IMF-and World Bank-supported structural adjustment programme, of which one year immediately precedes the decision point. The previous track record is taken into account. Just as in the original initiative, Paris Club creditors provide debt service rescheduling according to the Naples Terms. The major important difference compared to the first initiative is that countries must prepare a PRSP during this first stage. Thus, the objective of poverty reduction has been made an integrated part of the process. The actual debt relief comes in the second stage of the initiative. A drawback of this approach besides the exclusion of many heavily indebted countries and inadequate amounts offered for debt reduction, is the complexity of conditionality. The introduction of the PRSP as a condition for the initiative has brought poverty reduction to the centre of policy conditionality. However, this not only involves a change of emphasis, but also an extension of policy conditions. In addition to traditional macroeconomic and structural reforms, the country must also implement a number of agreed social development policies. At the same time as conditionalities pile up, the PRSPs have not succeeded in better aligning macroeconomic issues and poverty issues than in the past and thus macroeconomic frameworks have not significantly changed. It is important to ensure that the poverty focus is included in the design of macroeconomic adjustment measures. Even though the change of emphasis must be welcomed, it cannot be denied that the piling-up of different sets of conditionalities is delaying the road to the completion point, at which point countries receive unconditional and irrevocable relief. According to the World Bank and IMF, in theory it should take two years to produce a PRS. However, Uganda has been working on a PRS for five years, and even after this the Bank and the Fund wanted Uganda to provide some more details on the cost of poverty reduction programmes and to increase the link between expenditures on poverty reduction and poverty indicators.

15


INTRODUCTION In this chapter we take up for discussion three issues related to exchange rates in developing countries. The first is the concept and measurement of real exchange rates as well as exchange rate misalignment and its impact upon economic growth. The second topic taken up is factors that are important in deciding upon the exchange rate regime for developing countries. We then discuss the problem of dilemmas of policy in the face of sustained capital inflows. In Chapter 17 we discuss the problem of policy design to prevent exchange rate crises. Real exchange rates Real exchange rates are a variable of key interest and importance in all countries—developed and developing. We have already encountered them in Chapter 7, for example, where we modelled the trade balance as being influenced by the real exchange rate. However, it might be the case that the real exchange rate may not be very flexible—either because the domestic price level or the nominal exchange rate is slow to adjust. The foreign price level is taken as a parameter. If this were indeed the case, then, it would be meaningful to develop the notion of an equilibrium real exchange rate. Real exchange rate misalignment takes place when the actual real exchange rate deviates from this equilibrium value. An exchange rate is labelled as being ‘undervalued’ when it is more depreciated than this equilibrium rate and ‘overvalued’ when it is higher than this real rate. The equilibrium real exchange rate, itself, is an unobserved construct and refers to the value of the real exchange rate that would have prevailed had there been simultaneous external and internal balance in the economy. A departure of the actual exchange rate from this magnitude is, then, referred to as ‘misalignment’. The literature on modelling of equilibrium real exchange

364


rates as well as misalignments of real exchange rates is quite extensive and includes Bayoumi, Clark and Symansky (1994); Borenzstein (1990); Cottani, Cavallo and Kahn (1990); Devrajan, Lewis and Robinson (1993); Elbadawi and Soto (1995); Ghura and Grennes (1993); Montiel (1999); Paul (2001); Razin and Collins (1997); and Williamson (1995). Works relating to developing countries usually cite the pioneering work and papers of Edwards (1989a, b; 1995a, b) as their starting point. Concurrently with these developments there was progress in empirically quantifying the notion of a real equilibrium exchange rate and misalignments therefrom. However, the corresponding work on developing countries is rather scarce. Notable exceptions include Patel and Srivastava (1998); and Paul (2001). The simplest and most widely used notion of an ‘equilibrium’ real exchange rate is that of purchasing power parity. The simplest form of purchasing power parity holds when the nominal exchange rate moves to correct for all movements in the relative price level of the home and foreign countries. Thus the simplest form of this hypothesis is to estimate the equation: (15.1)

where e is the nominal exchange rate (domestic currency per unit of the foreign currency), and p, p* are the (logs) of the home and foreign price levels. If α=0 and β=1, the strict form of PPP is said to hold. In this case the nominal exchange rate movements exactly mirror relative price level movements. A weaker form of the PPP hypothesis is that β=1. In this case nominal exchange rate movements mirror relative price level movements up to a constant, α. Williamson (1995) suggests that in sharp contrast to the PPP hypothesis, there is empirical evidence to believe that the real exchange rate is nonstationary. It is certainly not a constant. In line with the emphasis on PPP, there exists some literature emphasising misalignments as departures from PPP. This has the problem of assuming that real exchange rates are at some constant level that prevailed in some base period when macroeconomic balance was thought to have obtained. Further, this approach fails to take into account the movements in a country’s real exchange rate over time in accordance with changes in the macroeconomic fundamentals. As a consequence of this, most studies have visualised the equilibrium real exchange rate as one that is consistent with the simultaneous attainment of internal and external balance. Internal balance requires clearance of the market for non-tradable goods not just in the current period; it also requires that people expect this equilibrium to hold in future time periods. External balance requires that the intertemporal budget constraint should be satisfied— in other words that the discounted sum of the country’s current account balance should be zero—thus implying that the country is not a net borrower or lender forever. Clearly, the value of the equilibrium real exchange rate implied by this definition is unobservable.


365

The empirical literature using this notion for developed countries is quite extensive. It has generally been focused towards multilateral exchange rate models as well as multi-currency macroeconometric models of Bayoumi et al. (1994) and others. Stein (1995) used the concept of natural equilibrium exchange rates to provide estimates of the US dollar equilibrium exchange rates. For developing countries such an extensive database is typically unavailable. This prompted Elbadawi (1995) to use a much smaller number of fundamentals to estimate equilibrium real exchange rates. Edwards (1989, 1995) defines equilibrium real exchange rate as being determined by the fundamentals of the economy. Both real as well as nominal factors are modelled to determine the equilibrium real exchange rate. He begins by defining the real exchange rate (er) as: (15.1a)

where PN is the price of non-tradables and is the price of tradables with being the foreign price of tradables and t the rate of tariff. Thus a fall in er is associated with a fall in the relative price of non-tradables and implies a real exchange rate depreciation. Since we are interested in the allocation of resources between tradables and non-tradables in the domestic economy, the domestic relative price of tradables to non-tradables should be used in the definition of the real exchange rate. A limitation of this approach is that it assumes that all tradable goods are subject to the same tax rate although different goods may be subject to different rates of taxes. One way out of this is to define sector-specific or good-specific indices of real exchange rates after taking into account good-specific tariff rates. However, this does not have the advantage of an economy-wide measure. Edwards, therefore, prefers to define an economy-wide index of real exchange rates exclusive of these taxes, i.e., (15.2)

In the case of developing countries non-availability of data on the prices of tradable and non-tradable goods makes it difficult to find an empirical counterpart to equation (15.1a). Hence proxies are used. The consumer price index, it is argued, has a greater share of non-tradables and is hence suggested as a proxy for the prices of tradables. Since the share of tradables is higher in the wholesale price index (WPI), the foreign WPI is often suggested as a proxy for the price of tradables. The ‘fundamentals’ typically included are: (i) terms of trade; (ii) government consumption; (iii) level of tariffs; (iv) a measure of technical progress; (v) capital flows; and (vi) sundry other variables such as the rate of investment. Typically these variables as well as the

366


computed real exchange rate are all non-stationary. If there is a single cointegrating relation between them the predicted value for the real exchange rate from this cointegrating relation is taken to be the equilibrium real exchange rate. Departures from these values—essentially the error correction mechanism—are modelled as ‘misalignments’. In the context of developed countries where financial flows are well integrated, the real exchange rate is influenced by asset market conditions.1 To see the influence of financial markets consider the following example. Suppose an investor in Thailand is deciding to invest the equivalent of one Thai good for n years in either n-year Thai bonds or n-year US bonds. Suppose that she decides to hold Thai bonds. Let rnt be the average yearly real interest rate she can expect to get if she holds the n-year bond for n years. Thus she can expect to get goods in n years. Now suppose that she decides to hold n-year US bonds instead. Let er be the real exchange rate. Let be the n-year US interest rate. Let the expected real exchange rate n years from now be Then for every Thai good she invests in n-year US bonds, she can expect to get Thai goods in n years. If we assume that expected returns have to be the same for arbitrage opportunities to cease and equilibrium to be attained, then the following condition must hold:

A good approximation is

To firm up matters let us take a numerical example. Suppose that n=10 and the 10-year Thai interest rate is 10 per cent and further that the annual rate of inflation is expected to be 6 per cent on average for the next 10 years. Hence the Thai domestic rate of interest, rnt, is equal to Suppose that the 10-year US interest rate is 6 per cent with average expected inflation of 3 per cent, so that the 10-year foreign real interest rate is equal to Hence she can expect to get 3 per cent a year in terms of foreign goods. When will she be indifferent between holding Thai and US bonds? Only when she expects the relative price of US goods to increase by 10 per cent over the next 10 years. In that case, the two expected returns in terms of Thai goods are equal. Or


367

Hence we can write the real exchange rate as follows:

If the investor holds Thai bonds she can expect to receive 4 per cent a year in terms of domestic goods. Modelling the equilibrium real exchange rate and its dynamics Any country has many trading partners and the computation of a real exchange rate has to recognise this fact. The standard method of computing the real exchange rate is as follows: 1

Suppose the country in question has n trading partners. The nominal exchange rate with respect to the ith country is ei and the fraction of this country’s total world trade2 with the ith country is ßi. An index of nominal effective exchange rate (NEER) is computed using the formula (15.3)

2

A relative price index is then computed as: (15.4)

3

where CPI is the consumer price index for the country in question and is the wholesale price index of this country’s ith trading partner. Finally a real effective exchange rate (REER) is computed as: (15.5)

This REER is used in analyses of the real exchange rate. It is clear from this that the REER will depend upon the base year chosen for the computation of prices indices as well as the trade weights. There is no easy remedy for this. In most cases, the real exchange rate calculations described above are converted into indices by taking a base year. From this it would follow that the characteristics of the base year become a key issue in the question of determining how to interpret the developments in the real exchange rates. To be sure, real exchange rate movements based on different years can be interpreted quite differently. Ideally we would like to choose the base year such that both internal and external balances obtain in that year.

368


Another important point to be taken into account is the number of countries according to which the effective exchange rate would be calculated. The most popular method is to choose the countries according to the weights of their currencies in the foreign trade volume of the country in question. The IMF, for example, determines country weights using the foreign trade data of the commodities in the agriculture and the manufacturing sectors. These weights are used in the calculation of the nominal exchange rate of a given country. Then, in order to render comparison possible between many countries, real effective exchange rates are calculated after correcting the nominal rate with the consumer price indices of each country. Typically base years of real effective exchange rate indices are modified every ten years. In assessing real and nominal effective exchange rates, it should be remembered that monetary authorities have only limited control over real exchange rate movements. Developments like foreign capital movements, changes in the terms of trade or relative productivity growth differentials due to technological change may alter the value of the real exchange rate. A consequence of the above is that real depreciation of the exchange rate may cause unexpected adverse effects on the trade balance. This is known as the J-curve phenomenon. We encountered this in a different context in Chapter 6. An important reason for the J-curve effect is that both export and import connections are confirmed in advance. Therefore, in the short run, the real depreciation of the exchange rate does not cause the expected reduction in import volume. On the other hand, the value of the existing import volume in terms of domestic currency increases with real depreciation. Hence, an adverse influence on the foreign trade balance is observed. However, since the quantity effect dominates the price effect in the long run, the expected influence of the real depreciation of the exchange rate on foreign trade is observed eventually. On the other hand, an important point particularly in the context of developing countries is that when the share of intermediate goods and investment goods in total imports is high, the real depreciation of the exchange rate might not lead to a substantial reduction in imports as expected by theory. In Figure 15.1, I report the computed behaviour of the 36-country NEER and REER for India. The choice of countries is made on the basis of their importance as trading partners of India. In Figure 15.2, five-country real effective exchange rates and equilibrium real exchange rates for India are reported. The five countries chosen are the most important trading partners of India. The general tendency is for both the equilibrium real exchange rate as well as the real effective exchange rate to fall over time, although there were short-lived rises in the early 1960s and early 1980s. The experience of many developing countries has been broadly similar. However, there are substantial differences in detail. The five-country REER calculated as above is related to key variables in a cointegrating framework. Those values consistent with magnitudes predicted

Figure 15.1 Nominal and real effective exchange (1985 = 100) Source: Handbook of Statistics 2000, Reserve Bank of India

Figure 15.2 Five-country real equilibrium exchange rates and real exchange rates for India Source: Handbook of Statistics 2000, Reserve Bank of India


371

Figure 15.3 Misalignments of the real exchange rate in India Source: Handbook of Statistics 2000, Reserve Bank of India

from this cointegrating relation are referred to as real equilibrium exchange rates whereas departures from it are called ‘misalignments’. These are shown in Figure 15.3. For most of the period under consideration, the Indian real exchange rate seems to have been overvalued. This would explain the relatively inadequate export performance of the Indian economy during much of the pre-reform period. With market-oriented reforms since 1991 put in place, this overvaluation appears to have ended. Why are misalignments important? If the real effective exchange rate were equal to the equilibrium real exchange rate then, one would presume, that the value of the real exchange rate is consistent with the fundamentals of the economy. However, this is true only in the long run. In the short run one would expect that real exchange rate movements outside of equilibrium would take place. Such movements are almost invariably associated with movements in net exports: a depreciation (appreciation) of the real exchange rate leads to an increase (decrease) in net exports. However, this effect appears with a lag of a year or more in the case of developing countries. Thus, as noted above, we do have a J-curve phenomenon operating in developing countries. Actually this seems to be the case in OECD countries as well. An important point to understand about real exchange rate misalignments is that they incorporate movements in the nominal exchange rate as well as in relative price movements. Hence the choice of the exchange rate regime is critical to real exchange rate misalignments. If the country is on a nominal

372


exchange rate peg, real exchange rate adjustments will be reflected in changes in domestic prices since the cushion of nominal exchange rate adjustment does not apply. The differential impacts of real exchange rate misalignment across developed and developing countries was investigated in a contribution by Razin and Collins (1997). They discover that net terms of trade and capital flows are important determinants of equilibrium exchange rates in developing countries—but not necessarily in developed countries. Greater net trade surpluses in developing countries are associated with more appreciated RERs. Terms of trade improvements are associated with more appreciated RERs. Further, terms of trade have a stronger effect in developing countries as opposed to developed countries. Output shocks are important determinants of real exchange rates in developing countries but monetary shocks are more important in developed countries. Razin and Collins include exchange rate misalignments as a determinant of growth in cross-country growth regressions. They discover that volatility of the real exchange rate has been negatively associated with economic growth. Thus macroeconomic instability, for which real exchange rate volatility is a proxy, negatively impacts upon economic growth. Further, there is some evidence to suggest that exchange misalignment negatively impacts upon economic growth. Also, the effect of misalignment upon economic growth is asymmetric between over- and undervaluations. When the real exchange rate is overvalued there is a negative impact upon economic growth. This is particularly true for developing countries and for large overvaluations. Undervaluations, however, do not appear to have any significant impact upon economic growth. EXCHANGE RATE REGIMES IN DEVELOPING COUNTRIES Developing countries face an external economic environment undergoing considerable change. Of particular relevance to developing countries in their choice of exchange rate regimes the following considerations are increased capital mobility and exchange rate risk issues. As Table 17.3 indicates, gross capital flow, to select developing countries have increased considerably since the early 1980s. This is particularly true of portfolio investment and less so for FDI. Several factors have contributed to this increased flow—the most important of which has been the increased capital market liberalisation of several emerging market economies. Net private flows to developing countries after hovering around the 0.5 per cent of GDP in the 1970s and 1980s rose to about 3 per cent of their GDPs around the mid-1990s. Soon after the East Asian crisis of 1997 the figure went down again and in 1998 stood at 1.5 per cent of GDP. Such large capital flows—particularly if they are short-term FII flows—have the potential of having reversals. Such reversals have often been cited as reasons for currency crises.


373

Exchange rate risks Residents, and even governments of developing countries, find it difficult to borrow abroad in ‘hard’ currency and non-residents are unwilling to take net long positions in the currencies of developing countries. In net terms, the foreign currency liabilities of residents of developing countries typically exceed their assets. This implies that they are exposed to exchange rate risks through their balance sheets as well as through trade. Contributing to these exchange rate risks is the question of choice of the appropriate exchange rate regime— a question to which we now turn. Choice of exchange rate regimes The choice of the appropriate exchange rate regime and the use of capital controls by developing countries are perennial problems to which there are no simple and final answers. It would appear that a solution to these issues is highly dependent upon the circumstances in which a country finds itself. Thus, in the literature, everything from a completely floating exchange rate to dollarisation (whereby a country takes on the US$ as its official currency and abandons its own currency) have been suggested. To understand the context-driven nature of these solutions consider a small open economy that is undergoing high inflation. If policy makers in this country are desperate for monetary and price stability, a currency board or even dollarisation is a good option. Abandoning the national currency and choosing the US dollar as legal currency is the most extreme form of fixed exchange rate.3 This would lead to a sharp reduction in the inflation rate and in equilibrium the rate of price inflation in this country would equal that in the US. The advantages of this arrangement are (i) quick attainment of price stability and (ii) adverse expectations about future financial performance of this economy are cooled. The disadvantages of this arrangement are as follows: (i) the country gives up monetary policy as an instrument, (ii) The government of the country loses all seignorage revenue that it would have got had it printed its own currency (Chapter 13). All seignorage revenue on account of the use of currency in this country would now flow to the US. (iii) The loss of national pride in using the currency of another country for domestic use.4 (iv) Last but not the least is the problem of not being prepared for dollarisation because of inadequate domestic preparation. With inappropriate external and internal borrowing policy and adopting the dollar as the national currency even though the bulk of its trade originated in Europe, Argentina is currently encountering severe problems. Adhering to the US currency reduced its flexibility to respond to European price movements. The extant literature distinguishes between the following nine exchange rate regimes. They are arranged below in order of increasing flexibility.

374


Fixed exchange rate niche 1

2

3

Monetary union. This occurs when a country adopts a foreign currency as legal tender. Dollarisation is a special case of this. Alternatively, two or more countries may adopt a single currency. A little more flexible than the monetary union is a currency board. This has three defining characteristics. (i) The exchange rate is fixed by law. (ii) The monetary base is increased one-for-one with foreign exchange reserves. (iii) As soon as a balance of payments deficit occurs monetary policy is tightened in response. Thus expenditures are adjusted automatically. The exchange rate is fixed and the central bank of the country is committed to buying or selling however much foreign currency is necessary at the given exchange rate. The central bank also expresses its commitment to maintaining the fixed exchange rate peg.

Intermediate niche 4

5

6 7

The variable peg. The exchange rate is fixed but, unlike regime 3, there is no open-ended commitment to protect the peg. Thus there is no indication that devaluations or revaluations would forever be resisted. The crawling peg. Under this arrangement, there is a pre-announced policy that the currency would be depreciated against a major currency or a basket of currencies by a fixed percentage each time period. Basket peg is a regime in which the peg appears with respect to a basket of currencies rather than to one currency. Target zone. The central bank is only committed to maintaining a margin of fluctuations around some central rate rather than to a single peg.

Flexible niche 8

9

Managed float is a situation in which there is no specified target for the exchange rate regime. But the central bank does intervene from time to time in the currency markets. The full or free float is a situation in which the central bank does not have any target for the exchange rate nor does it intervene in the foreign exchange market.

How should a country, particularly a developing country, choose between these exchange rate regimes? Frankel (1999) argues that there is no particular exchange rate regime suitable for any country at all times. This is an obvious enough proposition but needs to be stated as such. Which regime is optimal for any country depends upon its circumstances and this choice would be


375

influenced by this country’s experience, its current and anticipated trade orientation, the structure of financial flows into this country and so on. Thus countries in South East Asia or Russia, which pegged their exchange rates and went through currency crises, would endorse any move towards flexibility of exchange rates. Certainly there are clear-cut cases when a full float is cptimal. The conditions for a full float to be optimal include (i) the economy in question must be large; (ii) there must be greater economic integration between the constituents of the economy than between these constituents and the rest of the world; (iii) there should be no obvious reason for this economy to subvert its monetary policy to another country. Two oft-cited cases of currencies that satisfy these conditions include the dollar and the euro. Conditions for a successful full float are satisfied in both cases. Both Europe and the US are large economies. There is greater economic integration within Europe and within the US than between individual countries in Western Europe (or individual states in the US) and the rest of the world. Furthermore, there is no reason for these economies to subordinate their monetary policies to that of any other country. This does not mean, of course, that the US or Europe should never intervene in foreign exchange markets but only that this would be rare and certainly not a part of their core monetary policies. Certainly hardly any developing country would qualify to be treated as a sure case for a full float of its currency. On the other hand, countries such as Hong Kong, which opted for currency boards and were able to avoid the exchange rate crisis to a considerable extent, would clearly support a move towards fixed exchange rates. The classic conditions for an exchange rate to be fixed are if the country constitutes an optimum currency area (OCA).5 Conditions for an OCA include small size of the country and openness to international trade. In addition, labour mobility should be high within the country and, since the country’s monetary policy would be tied to maintaining the fixed exchange rate, the country should have credible fiscal policy to manage downturns in the business cycles. Furthermore, there should be high correlation between the business cycle of the country in question and the country to whose currency the country’s exchange value is pegged. Such countries are likely to see considerable benefits from fixed exchange rates—monetary union or currency boards—and are less likely to need monetary independence in the first place. Examples are Panama (tied to the dollar) and Luxembourg (tied to the euro). Some such countries (e.g. Argentina since 1991) urgently need to bring in monetary and price stability even though they do not satisfy the conditions for an OCA. Countries that are pegging their currencies should have a strong, wellsupervised and regulated financial system. Otherwise, the country might simply convert currency-crisis vulnerability into banking-crisis vulnerability. Furthermore, such countries should also have adequate reserves in order to smoothen the impact of destabilising speculation in their currency markets.

376


Figure 15.4 Frenkel’s impossible trinity

It would thus appear that many of the preconditions for the success of fixed exchange rate regimes are not true of most developing countries. They would have neither the foreign exchange reserves nor the sophisticated fiscal policy to ensure the success of the fixed exchange rate regime. Yet another proposition states that policy choices in the intermediate niche regime are not possible and that countries would have to move in the direction of a full float or a fixed exchange rate regime. The rationale for this appears to be the existence of what Frankel (1999) has called ‘the impossible trinity’. This is depicted in Figure 15.4. Each side of the triangle is a desirable policy objective whereas each vertex is a policy tool. Frankel argues that each of the three policy objectives in Figure 15.4 is important. But these are mutually inconsistent. If a country chooses full capital controls, for example, it can achieve monetary independence and exchange rate stability but will have to give up financial integration as it would have to impose extreme capital controls and shield its capital markets from responding to external events. As the degree of capital mobility increases the country is pushed towards monetary union or a pure float. With a monetary union it obviously enjoys exchange rate stability (complete stability with respect to the members of the union and reduced instability with respect to the rest of the world) but it has to give up monetary independence and pursue monetary policy consistent with such union. If it chooses a pure float, it will sacrifice exchange rate stability but get monetary independence and enjoy integration of its financial markets with the rest of the world. Our discussion so far does not imply that the intermediate niche is not viable, although in recent times, most intermediate niche type policies have


377

been tried and failed and some of these failures have been spectacular. Thus the Mexican peso, for instance, had a spectacular collapse. Particularly since the exchange rate collapse of 1997–99 it might be the case that going to the fixed exchange rate niche or the floating rate niche might be construed, by international investors, as the only viable options available for developing countries. However, it is likely to be the case that the search for a single regime that will eliminate currency speculation is doomed to fail unless strict capital controls can be put into place. This, of course, carries the cost that it will deprive the developing countries imposing these restrictions with much needed capital flows. It also might be the case that capital flow restrictions might be difficult to put into place since these restrictions may become porous over time. This, then, is a no-win situation but it cannot be said that the intermediate niche is entirely unviable for developing countries. The experience of India, for instance, is an important indicator of the possibility of such success. Another proposition being discussed is that the world will ultimately crystallise into three major currency blocks. The first is the dollar zone of the Americas, the second the euro of Western Europe and the third the yen block of the Far East. All countries of the world will, perforce, become members of one block or the other. The three currency areas have monetary masses more or less corresponding to their respective GDPs, the ratio of about $9.5 trillion, $7.0 trillion and $5.0 trillion respectively, together making up perhaps 60 per cent of world GDP.6 However, the yen block is not as developed as the other two. A final view—popular in Europe—is that the cross rates between the dollar, the euro and the yen should be stabilised. Other countries in the world should then be free to adopt the currency regime of their choice. With stability in the cross rates between these major currencies, persistence of instability in other currencies would not be possible. Excessive volatility between these three major currencies is responsible for repeated currency crises, this view argues. There is some evidence in this point of view. For twenty-five years after 1948 one dollar was worth 360 yen; in 1985, before the Plaza Accord, it was around 240 yen; ten years later it had fallen to 79 yen. And then three years later, in June 1998, it had soared to 148 yen, bringing on the Asian crisis; and then suddenly it came down to 105 yen. Volatility of this type destabilises financial markets, disrupts trade and creates extremely difficult conditions not only for Asia and developing economies but also for all countries of the world. CAPITAL CONTROLS As has been noted in Chapter 5 with perfect capital mobility it is not possible to have an independent monetary policy with fixed exchange rates. This country must accept that its interest rate will be tied to world interest rates. However, a country that does not have full capital mobility can have some

378


leeway in setting its interest rates and having an independent monetary policy at the same time. The case of the two Asian giants India and China is of this nature. They have both opted for capital controls and retained an ability to influence both their interest rates as well as having independent monetary policies. Another interesting and successful application of capital controls is to be found in the case of Malaysia. Kaplan and Rodrik (2001) note that Malaysia recovered swiftly from its currency crisis after it imposed capital controls in September 1998. Korea and Thailand also recovered smartly but the crisis was deepening in Malaysia in September 1998 while it was weakening in Korea and Thailand. Kaplan and Rodrik (2001) make the important find that compared to the IMF programmes initiated by Thailand and Korea, the capital controls used by Malaysia were associated with lower output, employment and real wage losses, and faster economic recovery and stock market upturns. This seems to point out the possibility that capital controls are useful in dealing with crises. They may even have a role as a useful backdrop for the conduct of independent monetary policy. Furthermore, capital inflow restrictions may be useful in the presence of sudden surges in capital inflows. When a country is faced with a sudden upsurge in such inflows it must be able to distinguish whether it is becoming the next tiger or is merely the subject of the next speculative bubble. Only if the capital inflows have been sustained for a while can the country be sure that it is not just experiencing a speculative bubble. However, as noted above, capital controls have a cost. The efficacy of restricting inflows is likely to be greater than that of restricting outflows. It is easier to scare away investors rather than to keep them in. A history of capital controls may stem the flow of needed development finance capital into these countries. This probably explains the comparatively poor record of attracting FDI in India as opposed to some Latin American countries and points to the importance of targeting the composition of capital inflows. The higher the reliance on FDI the lower is the probability of crisis. On the other hand, external borrowing particularly of a short-term nature is likely to be risky. Thus one form of capital controls would not try to affect the magnitude of capital inflows but to alter the composition of such inflows. Such policy usually takes the form of placing some penalty on short-term borrowing—in terms of foreign currency—from abroad. Chile and Colombia seem to have been able to alter the composition of capital inflows in this manner without affecting very much the total magnitude of the inflows. In the Chilean case, for example, this took the form of requiring a non-interestbearing deposit with the central bank once such short-term borrowings are made. Banks can be penalised from borrowing abroad by placing a higher reserve requirement on their borrowings. Such policy initiatives are necessarily of the type that is considered prudent in international banking circles, including by the IMF.


379

In the case of capital controls, too, then there is no simple answer that suits all countries for all times. These controls may well have a use during crises. However, their application has a cost to the country. As countries develop, it would be less and less useful for them to consider capital controls as a long-term solution to their capital account problems. In the intermediate run, countries are better advised to undertake the banking and financial sector reforms that would achieve their smooth integration into the world economy. MONETARY POLICY DURING CAPITAL INFLOWS An inflow of foreign capital has important implications for macroeconomic equilibrium. Such an inflow raises domestic expenditure, which would then lead to an increase in the demand for non-tradable goods. This would lead to an appreciation of the real exchange rate. The demand for tradables (imports) goes up as well and this would tend to widen the trade deficit. The net effect on the domestic economy depends upon the exchange rate regime in place. If the central bank does not intervene and the exchange rate is floating, the nominal exchange rate will go up. If the exchange rate is pegged the accumulation of foreign assets will lead to an expansion of the monetary base and hence of the money supply. This would then have further effects on aggregate demand and inflation. To understand this point we go through an example—the recent experience of India. The path of adjustment of the Indian nominal and real effective exchange rate over a 36-country export weighted average is given in Figure 15.1. REER is constructed as a weighted average of NEER adjusted by the ratio of domestic WPI to foreign CPI. Both NEER and REER depreciated after 1985. At the time of the regime switch in 1993 the NEER depreciated— however the REER fluctuated around a constant value. In August 1995 and August 1997 there were sharp REER appreciations. It appears that the authorities have been reluctant to permit a nominal appreciation in the face of sustained capital inflows. The fact that the monetary authorities have been reluctant to permit a nominal appreciation implies that they have been willing to absorb the capital inflows and pass these on. Either these would be accommodated as higher monetary base and result in higher money supply and inflation or be sterilised and result in a change in the composition of the central bank’s balance sheet— or a mix of the two. The precise form of the mix chosen is a policy option. Figure 15.5 plots the balance on the current and capital accounts as well as the overall balance and foreign exchange reserves as percentages of the GDP and provides interesting evidence of the development of the Indian external sector. Confining ourselves to the period since the reforms the end of the 1980s and the beginning of the 1990s were associated with a deterioration of the current account as well as the overall balance. There

Figure 15.5 Current, capital, overall balance and foreign exchange reserves as percentage of GDP Source: Handbook of Statistics 2000, Reserve Bank of India


381

was a dip in foreign exchange reserves. The subsequent revival in the current account saw an enhancement of reserves. What is most striking about the behaviour of the reserves is their sharp rise since 1992–93—quite unrelated to current, capital or overall balance movements. In fact the capital account surplus and overall balance actually fell while the reserves were rising sharply. Except for a mild downturn in 1994–95 this upward trend continued. By all accounts the foreign exchange reserves as a percentage of GDP are now hovering at or above the 10 per cent of GDP mark.7 Although India has had a market-determined exchange rate since 1993, the flexibility permitted by the RBI has been limited. This feature has been common to many developing countries and has been called ‘fear of floating’ by Calvo and Reinhart (2000). But such a policy course has had its adverse effects. This accumulation of foreign exchange reserves in the face of restricted movements in the exchange rate has led to the money supply becoming endogenous. This is documented by Rath (2001) in the context of financial inflows but applies more directly to the case of accumulation of foreign exchange reserves. Only part of the inflows has been sterilised.8 This is one important impact of the accumulation of reserves. A second important implication follows from the link between fiscal and monetary policy in the presence of capital inflows. Sterilisation is often a controversial issue and there are arguments both for and against it (Calvo, 1991; Spiegel, 1995). Sterilisation involves an exchange of foreign currency assets for domestic assets; so that the rate of interest must be kept higher. However, this interest differential may cause further capital inflows and be destabilising in extreme cases. Open market operations accompanying sterilisations would also exert pressure on short-term interest rates. Such pressures on interest rates aggravate the difficulty of servicing the public debt. The increased cost of servicing the public debt imposed by interest rate differentials needed to attract capital flows has been termed quasi-fiscal cost and has been estimated to be in the range of 0.25 to 0.5 per cent of GDP for some Latin American countries. Although no firm estimates exist for India such costs could well be large in the context of the overwhelming importance of interest payments on the public debt. As Edwards, Gregorio and Valdes (2000) argue, the best response from fiscal policy to a situation of increased capital inflows is to enforce fiscal austerity. In the Indian case not only is fiscal austerity hard to implement but also the perceived need for accumulating foreign exchange reserves helps keep interest rates high. Why is it necessary to keep acquiring more foreign exchange reserves and, in the process, keep interest rates high? Surely the reason has to lie in the need to control the impact of sudden reversal of capital flows. As is well known this is more likely the greater the fragility of the banking system9 and the greater the public debt burden. The high level of NPAs with Indian banks is, thus, a cause as well as a consequence of high interest rates since, as argued above, the high level of NPAs in itself makes it difficult to move to a

382


lower interest rate regime. The high level of government debt and the thinness of the government bond market also make it difficult to lower interest rates. These factors also increase the perceived probability of a sudden capital outflow and necessitate a large reserve of foreign exchange, which in turn requires higher interest rates. Only financial sector reforms encompassing public debt restructuring, reduction of NPAs and reorientation of the sterilisation of foreign inflows policy can help break this circle. Many developing countries are caught in the predicament described above for the Indian case.

Malaysian capital controls In a Fortune Magazine article published in the midst of the Asian financial crisis (1998), world-renowned economist Paul Krugman suggested that one solution to the crisis was the implementation of temporary capital controls. Krugman had by that time propagated the view that the entire Asian economic model was flawed, arguing that the tremendous economic growth in the region prior to the crisis was fuelled by ‘excessive capital inputs’. Thus Krugman was suggesting that Asian economies had accumulated so much savings and attracted so much foreign capital, that their economic growth was fuelled more by the sheer volume of investment in their economies, rather than by the productivity of those individual investments. Capital controls, Krugman argued, would (i) help reinforce the idea that capital inflows were not the solution to Asia’s woes, and (ii) provide a buffer against the instability short-term capital inflows can sometimes cause. Almost immediately thereafter, Malaysian President Mohammed Mahathir, a vocal critic of the global financial system and its players (in particular, international speculator George Soros), announced a complex and fairly comprehensive system of capital controls. The system was designed to tax foreign capital exiting the country at increasingly higher rates depending on the brevity of its stay within the Malaysian financial system. Krugman, who was admittedly stunned that Mahathir would follow his advice, then wrote an open letter to the Prime Minister, Fortune Magazine explaining in greater detail what he ‘really meant’ by advocating capital controls. He urged Mahathir to view capital controls as a short-term, stop-gap measure and not as a substitute for liberalising reforms. Most US agencies did not have even the guarded enthusiasm for capital controls that Krugman expressed. A Washington Post article (1998) claimed that US Treasury Department officials privately voiced hope that the controls would damage the Malaysian economy further, so as to dissuade other countries in the region and


383

around the world from adopting Mahathir’s approach, which was a direct challenge to the IMF approach. Mahathir refused to implement the austerity programme of high interest rates and budget cuts, and instead opted for a stimulus package. Fearing that investors would flee Malaysia and its ‘unorthodox’ economic plan, Mahathir essentially prohibited foreign money from leaving the country for one year. He fixed the value of the ringgit to the dollar and effectively barred trading in the currency. Mahathir was accused of making Malaysia an ‘international pariah’, and observers predicted economic doom as a consequence. Those predictions did not materialise. Malaysia’s growth rate was respectable in 1999—higher than both Indonesia and Thailand, each of which implemented IMF-sponsored plans. Michel Camdessus, Managing Director of the IMF, was convinced to say in a television interview in 1999, ‘I praise the way in which Malaysia has been able to adopt a soft system of controls.’ The reference to a ‘soft system’ is in regard to changes made to the regime that improved investor confidence in Malaysia. The prohibition on the withdrawal of certain foreign capital was converted into a simple, albeit hefty, tax. Nevertheless, despite this easing of the capital controls, the Malaysian policy stands as an example of an alternative to the liberalisation approach favoured by the IMF. In 2000 as well, the Malaysian growth rate was higher than Indonesia’s or Thailand’s according to World Bank data. Over the period 2000–04 though, Malaysian growth is likely to slow down. This makes reforms necessary in the medium term and emphasises the fact that the efficacy of capital controls is confined to the short term.

DISCUSSION QUESTIONS 1 2 3

4

Why might a development economist argue that a small short-run surplus on the capital account is good? Why might the same economist be wary of large surpluses on the capital account? Suppose there is a local war as a result of which government purchases increase. What will happen to the current account and the real exchange rate? Why? Would your answer change if this were a world war? Why? Why do developing countries opt for overvalued exchange rates? Discuss the implications of such overvaluation on domestic producers and consumers. Does this indicate why such overvaluation may not be sustainable over time?

384


NOTES 1 2 3 4

5 6

7

8

9

The same could be true of developing countries as well if they had capital account liberalisation. It is standard to calculate ßi as (exports and imports from country i)/(total value of exports + imports). If the country keeps its own currency there is the possibility of a devaluation and financial markets are constantly monitoring for that. With dollarisation this possibility is ruled out. This is to be contrasted with the use of a global currency as advocated by former Federal Reserve Board Chairman, Paul Volcker and the Nobel laureate, Robert Mundell. These authors have emphasized the importance of stabilising not just price levels or inflation rates but also exchange rates. A global currency would certainly help achieve this. However, as Mundell (2000) notes, each time the idea of a world currency comes up, it does not find favour with the dominant power which is content to see its own currency being elevated to the status of reserve currency. It was Britain in the nineteenth and America in the twentieth century that rejected the idea of world currency. This is simply because with only its currency being used, the dominant power does not have to share seignorage revenues. Dollarisation, then, plays right into this sentiment. The classic exposition of the OCA is Mundell (1962). Mundell (2000) argues that this could ultimately lead to the formation of a world currency. A monetary union of any two of the areas would make it the dominant currency area and thus make it very attractive for the third area to join because the alternatives would be worse. The drive to acquire additional foreign exchange reserves even at rather unfavourable terms saw the issuance of the Resurgent India Bonds in the wake of sanctions placed on India after the nuclear tests as well as the India Millennium Deposit Bonds issued in 2000. On December 29, 2001 foreign exchange reserves had gone above $48 billion according to newspaper reports. There is some evidence that the fraction of capital inflows absorbed as reserves by the RBI has increased from about a third at the beginning of the 1990s to about a half towards the end. The greater the desired stability in the exchange rate the greater would be the required extent of sterilisation. For a theoretical analysis of why high NPAs may raise the chances of capital flow reversal see Detragiache (1999).

Part IV

Issues in Developing Countries’ Macroeconomics

16

Financial liberalisation and economic growth in developing countries

INTRODUCTION Surely one of the most important tasks before developing countries is to achieve higher rates of economic growth. Consequently an important question that we must confront is why countries grow at different rates. In Chapter 10 we have discussed some of these reasons that range from differences in rates of technological progress across countries to factor accumulation and differences in research and development (R&D) profiles. Other reasons cited in the literature include differences in macroeconomic stability and in legal and other institutional structures. Surely, one can go on adding to this list but that would lead to a rather vacuous statement that differences in economic growth depend on a whole range of factors. A focused response would be more meaningful. In recent years, there has been considerable emphasis on the degree of financial development as a determinant of the rate of economic growth. The literature on this is quite large. Levine (1997) is a useful survey of this literature and Khan and Senhadji (2000) is a recent restatement and empirical test of this proposition. This literature begins by pointing out that only in the perfect world of the Arrow-Debreu framework where there are complete contingent markets is financial intermediation not necessary. In all other situations financial intermediation is necessary to facilitate transactions. Financial systems play an important role in the process of economic development. They mobilise savings and allocate investment funds. They facilitate the trading, hedging as well as the pooling of risk. In addition, they help monitor managers and exert corporate control. Further, financial development helps facilitate the exchange of goods and services. To take just one example of this—just think of how much credit cards and web-based payments systems have helped in conducting transactions. Hence an important function of financial development is to ameliorate transactions and information cost across space as well as across time.

388


What role does financial development play in the growth process of developing countries? The literature distinguishes between four different possibilities. 1

2

3

4

One branch of the literature argues that economic growth is largely the result of factors—such as savings and investment and technological progress—emanating from the real sector of the economy. Hence financial development and economic growth are not causally related. This would be the argument of schools of thought that ignore the institutional background of the growth process in developing countries. A second view argues that financial development follows economic development. Economic growth, it is argued, induces changes in financial institutions and practices. Financial development is, thus, demand driven. A third point of view argues that financial development is a determinant of economic growth. This school of thought argues that financial development is one factor—but an important factor—influencing economic growth. One version of this hypothesis argues that financial development is a precondition for economic growth whereas a second version argues that sophisticated financial systems help invigorate the climate for rapid economic growth provided there are no other impediments to economic development. A final point of view ascribed to authors such as Diamond and Dybvig (1983), Keynes (1936), Krugman (1998) and Singh (1997) states that financial development is an obstacle to economic growth. This is because of the inherent instability of the financial system. Thus this school would argue that there is a role for government intervention in the working of financial markets. This is in sharp contrast to the work of McKinnon (1973) who argues that state intervention in financial markets leads to their repression and, therefore, stunts economic growth.

Which of these four views is right is an empirical issue to be tested with data. As will be discussed, Graaf (1999) provides quite convincing evidence that financial development aids economic growth. WHY FINANCIAL DEVELOPMENT MAY HELP ECONOMIC GROWTH Formal growth models have highlighted two aspects of the impact of financial development on economic growth. In the models of Lucas (1988), Rebelo (1991) and Romer (1986), for instance, the financial system affects capital accumulation either by altering the savings rate or by reallocating savings among different capital-producing technologies. In the models of Aghion and Howitt (1992), Grossman and Helpman (1991) and Romer (1990) the financial sector performs functions, which affect steady state growth by altering the rate of technical progress.


389

In addition, financial development helps ameliorate risks. The literature typically distinguishes between two types of risks: liquidity and idiosyncratic. Liquidity is the rate at which economic agents can convert assets into purchasing power at agreed prices, i.e. into real purchasing power. Real estate is less liquid than bonds or equity. Furthermore, the liquidity of bonds or equity may vary across markets. Shares of AT&T traded in the New York Stock Exchange would be more liquid than equity traded in the Egyptian or Nigerian Stock Exchanges. Liquidity risks emanate from the fact that there may be uncertainty associated with being able to convert these assets into purchasing power. There may be transactions costs as well as informational asymmetries. As a response to both these problems capital markets that make liquidity available may come up. Within the context of developing countries the emergence of such capital markets is particularly important. Many potentially high-return capital or infrastructural projects in developing countries have long gestation lags and yield returns only after a considerable period of time. However, savers may not be willing to relinquish control over their savings for long periods of time. Hence financial institutions that allow savers to invest their savings into bonds and equities, which afford easy liquidity to them while, at the same time, permitting long-term investment would serve a critical role. In fact, Hicks (1969) argued that many of the inventions that we associate with the industrial revolution in England were available much before they were adopted. This adoption, however, had to await the emergence of appropriate financial institutions that were willing to make long-term investments. This effect, however, has to be distinguished from the effects that increased liquidity may have on savings. There is little evidence to suggest that greater liquidity has a beneficial effect on savings. Greater liquidity would have a substitution effect and an income effect that would work in opposite directions making the effect on savings ambiguous. In addition, the development of financial institutions would lower risks and make the precautionary motive for savings weaker. However, financial institutions could facilitate risk diversification. Holding a diversified portfolio of innovative projects reduces risk. This would naturally encourage innovation and investment in projects that would, by themselves, appear to be too risky. Thus, as King and Levine (1993a) argue, financial development that permits risk diversification can accelerate technological change and economic growth. Another important role performed by financial institutions in economic development is in acquiring information about, say, technology or projects. In the absence of appropriate financial institutions each prospective investor would have to expend resources to assess these costs. Financial institutions play a useful role in collecting and disseminating such information. Surely, the fees of such institutions would be lower than the aggregated costs for such prospective investors acting on their own. Since many firms would solicit capital, financial intermediaries and markets that are better at selecting the most promising firms and managers will induce a more efficient allocation

390


of capital and faster growth. Greenwood and Jovanovic (1990) have emphasised this point. In a similar vein, as stock markets develop, participants may have greater incentives to find out more about firms whose stocks are being traded (Grossman and Stiglitz, 1980; Holmstrom and Tirole, 1993). As a stock market gets larger and more liquid it can stimulate the acquisition of information. The acquisition of this information would improve resource allocation and therefore aid in accelerating economic growth. Also, if stock prices better reflect the managerial efficiency within firms, tying managerial compensation to such stock prices would help in better monitoring the firm (Diamond and Verrecchia, 1982). In addition to reducing the costs of acquiring information, financial markets and intermediaries may help reduce the acquisition and information costs of monitoring managers of firms and the costs of exercising corporate control. Firm owners would like to create arrangements within the firms that make it imperative for managers to operate the firm efficiently. Similarly financial agencies to which the firm is attached—say banks and other credit groups—would like to see the firm operated under the most efficient conditions and would, therefore, exert pressures on firm shareholders to better monitor the performance of the managers. Thus financial development helps in the optimal management of the managerial staff of firms. Financial development plays an important role in the mobilisation— pooling—of capital from savers. With the development of financial markets, individual savers get a range of instruments in which to invest. These can provide various degrees of liquidity—varying lock-in periods, involve different degrees of risk and so on. With the proliferation of such instruments the chances that savers would be able to allocate their savings in the portfolio of their choice improve. This, then, leads to an improvement in savings. In addition to individual savers, resources for investment can be raised from other sources including from abroad. Such mobilisation may involve multiple bilateral contracts between agents requiring productive capital and agents with surplus resources. Effective mobilisation of savings also facilitates the adoption of improved technology. To quote McKinnon (1973): The farmer could provide his own savings to increase slightly the commercial fertilizer that he is now using, and the return on this marginal new investment could be calculated. The important point, however, is the virtual impossibility of a poor farmer’s financing from his current savings the whole of the balanced investment needed to adopt the new technology. Access to external financing resources is likely to be necessary over the one or two years when the change takes place. Without this access, the constraint of self finance sharply biases investment strategy toward marginal variations within the traditional technology. McKinnon (1973:13)


391

In addition to mobilising savings, financial institutions help in facilitating exchange thereby leading to increased specialisation, higher economic growth and lower transactions costs. Within the context of a developing country, the first step in this is, of course, the greater monetisation of the economy. As other financial instruments develop these transactions, costs continue to fall. Markets that promote exchange and reduce transactions costs lead to productivity gains. Some of these productivity gains may spill over into further gains in financial development. Thus a virtuous cycle may be set in motion. THE EMPIRICAL EVIDENCE A number of studies have examined the empirical link between financial development and economic growth. Typically, they find that cross-country differences in financial development explain a significant portion of the crosscountry differences in growth rates. A typical such equation would look like:

where yi is the growth rate of country i, FD is the index of financial development, X is a vector of control variables and εi is the error term. Various indicators of financial development have been used—for instance, the ratio of various monetary aggregates such as M1 M2 or M3 to GDP. This, of course, has two problems associated with it. First, non-monetary developments may be differentially important in various countries. In particular, in richer countries where the benefits of monetisation have long since run their course, financial development may largely take non-monetary, non-banking characteristics. (Other measures of financial development that have been used include liquid liabilities of the banks and non-banking financial institutions as a share of GDP, the ratio of bank credit to bank and central bank credit, the ratio of private credit to domestic credit and finally the private credit as a ratio of GDP.) Second, as has been argued above, there may exist feedback effects from economic growth to financial development. Hence this equation should ideally be estimated using instrumental variable type techniques. In a recent paper Khan and Senhadji (2000) conduct a test of this proposition. They use a panel data framework for the period 1960–99 for 159 countries. But data for a number of developing countries have a shorter span. Hence it is an unbalanced panel. They use the following four indicators of financial development: (i) domestic credit to the private sector as a share of GDP; (ii) (i) + stock market capitalisation as a percentage of GDP; (iii) (ii) + private and public sector bond market capitalisation as a percentage of GDP; (iv) stock market capitalisation. They also find that the simultaneity between financial development and economic growth a matter of some importance for some indicators. They, too, find strong causal links from

392


financial development to economic growth. But their results lead them to conclude that differences in financial development help explain differences in growth rates but not differences in growth dynamics. Another important effort to understand the link between financial development and economic growth is by Graff (1999). He draws on a large panel data set of 93 countries—predominantly developing—from 1970–90. He visualises the testing of the relation between financial development and economic growth as a three-step procedure: In the first step a new proxy for financial development that captures the share of resources a society devotes to run its financial system at any point in time is constructed. This is in contrast to other standard indicators of financial development such as the M 2/GDP ratio. The motivation for this new index comes from the neoinstitutionalist view that financial development would reduce costs of economic transactions. (See North (1990), Williamson (1985).) The fraction of resources devoted to the financial sector can be taken as a proxy for the real value society places on these transactions costs. Possible proxies include (i) the share of the workforce in the financial system, (ii) the share of the financial system in the GDP, (iii) the number of bank branches per capita. A principal components type of analysis is undertaken to develop an aggregate index from these three indicators. Graaf (1999) then uses this index of financial development in a cross-country growth regression type framework to understand its contribution to the growth of output per capita. From our discussion in Chapter 10 the growth of output per capita in a human capital type framework can be written as: where i, t are index country and time respectively, A is exogenous growth of productivity, and Y, K, L, H represent GDP, physical capital, labour and human capital and with Letting gx represent the growth rate of any variable X, we have (upon removing country and time subscripts): This equation has been estimated in a number of studies to assess the contributions of the three factors to economic growth. However, as Benhabib and Spiegel (1994) argue, the level of human capital may affect the growth of A since a labour force with higher human capital would be able to absorb higher exogenous technical progress. Graaf writes the estimable equation (after omitting time subscripts) as:


393

where g(T) represents pure technical progress and FD is the index of financial development. FD also interacts with Y/L and H/L. There are two intercept terms—one that varies over time and the other, which is country specific. Thus ‘fixed effects’ are allowed for. It is likely that financial development and GDP growth could be mutually endogenous—with one affecting the other. However, the results of a causality test indicate that this is not the case. Financial development leads output growth and not the other way round. However, finance matters more in countries with higher adult literacy. Thus for countries with very poor development of human capital, financial development may be less important. Hence, there seems to be considerable agreement in the literature about the beneficial role of financial development. Graaf (1999) does point out that benefits of financial development do not remain significant for all levels of income. When it comes to developed countries the benefits of further financial development taper off and reach a plateau. However, there seems to be little doubt that financial development is useful for developing countries. THE CONSEQUENCES OF ECONOMIC REPRESSION When financial development is inadequate, economists have sometimes called it a facet of economic repression. To be sure, economic repression is taken to imply more than inadequate financial development. It includes controls on ceilings on interest rates and exchange rates, among other measures constraining economic activity. Economists have long been aware of the widespread prevalence of economic repression in developing countries. Two influential economists, Jagdish Bhagwati (1989) and Anne Krueger (1978), published separate monographs detailing the maze of physical and quantitative controls prevalent in many developing countries and examining the economic consequences of such controls. With the reduction in the appeal of quantitative control techniques in light of the disastrous experience with centralisation in most of the former socialist countries, it has become very fashionable to advocate the removal of these quantitative controls and to see the consequences of economic liberalisation as the shortest route to economic prosperity in developing countries. Be that as it may, it has to be understood that although economic repression has had significant economic costs, there was a rationale for them. An influential economic planning model developed independently by the Soviet economist Feldman and the Indian economist Mahalanobis posed the development problem in the following manner. A typical less developed country (LDC) faces adverse trading conditions in international markets because it exports mainly primary goods—the demand for which is inelastic. Thus the home economy faces a serious foreign exchange constraint. Coupled with this constraint is the poor availability of saving in the home

394


economy—primarily because of low incomes. The LDC is, hence, faced with the problem of allocating scarce foreign exchange and domestic savings in order to maximise consumption over the long run. Should these scarce resources be used to produce consumer goods or producer goods? If the future was severely discounted, the answer would be straightforward: the economy should produce consumer goods. As the discount rate becomes smaller the attractiveness of producing producer goods increases. Emphasis should then be placed on transferring surplus from agriculture to industry. Since the private sector would be unable to gather the resources for manufacturing producer goods and its ability to undertake the risks associated with initiating a programme of industrialisation almost from scratch would be severely limited, much of the production of the manufactured goods would have to take place in the public sector. The logic of the model states that in LDCs a programme of industrialisation would involve considerable investment across a broad range of sectors. The transport sector cannot be established unless the steel sector is simultaneously established and so on. If matters were left to the private sector, the steel producer, even if willing to establish a steel plant, would not have any way of communicating this decision to the producer of transport services. This has been called ‘secondary uncertainty’ as distinct from ‘primary uncertainty’ associated with random acts of nature, by authors such as Dobb (1937). The paucity of resources makes it necessary for the public sector to invest in these industries. The presence of secondary uncertainty makes it necessary for all these production decisions to be planned and coordinated. The public sector must be at the ‘commanding heights’ of the economy. The fact that markets for these goods do not exist in the home economy implies that price signals are of little use and quantitative control would have to be exercised. The basic philosophy that industrialisation was most important led to an implicit neglect of agriculture. The need to transfer resources from agriculture to industry was emphasised. The LDC experienced with development is ample proof of the failure of the Feldman-Mahalanobis model. The countries that were more ‘outward looking’ and learnt the lessons of international comparative advantage are precisely those that have developed fast. Countries that initially followed the logic of the Feldman-Mahalanobis model are now racing each other to dismantle the legacy of centralised planning as well, discussed in Bhagwati (1987), Chenery and Srinivasan (1989) and Krueger (1978). A cynic might argue that these quantitative restrictions were used as policy instruments in order to placate powerful political and economic interests. Industrial labour stands to gain because food prices are controlled. Industrialists gain because they get cheap inputs and enjoy the benefits of protected markets. Bureaucrats stand to gain because they get to make key economic decisions and may profit from such decisions. Above all, a whole generation of politicians has profited from making key decisions. This


395

bunching of interest groups presents a formidable opposition to genuine economic reforms in many LDCs. Moreover, during a process of economic reformers there are likely to be high costs of adjustments. There is to be considerable unemployment, for instance. Economic reforms, then, have significant political, social and economic costs and it may not be feasible for the developing country in question to liberalise very quickly. These costs constitute the principal reason why genuine reforms are often made only under the pressure of a severe economics crisis. There are a number of examples of this, for example, India, which started genuine economic reforms in July 1991. At that time India was about to default on foreign loans, its domestic debt had reached very high proportions, inflation was persistently high after being in single digits for decades, industry was stagnating and exports were sluggish. FACETS OF ECONOMIC REPRESSION In developing countries economic repression takes several forms. On the trade front, in order to placate powerful urban groups, governments restrict the export of foodstuffs and raw materials. Urban labour needs cheap food and urban industry needs cheap agricultural inputs. The domestic terms of trade are often turned against agriculture. More broadly, foreign trade is repressed by quantitative restrictions and/ or high tariffs. The prices of these goods, then, are usually determined by domestic demand and supply considerations and have little to do with global prices. The domestic financial system is usually insulated by exchange controls of various kinds. Obtaining foreign exchange for purposes of foreign trade is very difficult. Given the wide variety of controls on the foreign trade sector, multiple exchange rates exist. The ‘official’ exchange rate typically undervalues the true scarcity of foreign exchange and there is a thriving black market for foreign exchange. Moreover, in many countries, investment is subject to controls. Government licences have to be obtained for investing in certain sectors. In several other ‘priority’ sectors only public sector investment is permitted. To be sure, the prevalence of this licence quota raj encourages corruption among the government bureaucracy, which must make key economic decisions without reference to any market values. Official interventions in the allocation of credit may be as pervasive, detailed and bewildering as the proliferation of quantitative restrictions on foreign trade. In most developing countries stock markets and markets for bonds are rather underdeveloped. Thus the monetary system has to perform the role of intermediating between savers and investors. Most of private savings take the form of bank deposits and highly liquid post office savings, or claims against commercial banks. Control over a broad monetary aggregate such as

396


M3 is, therefore, important not just for adjusting the money supply but also for affecting savings. The backwardness of financial markets means that avenues for government borrowing are rather limited. Governments typically indulge in forced sales of government bonds to the banking system. This is done through an elaborate system of reserve requirements. It is not uncommon for the government to unload onto the central bank. The central bank uses some of these as a base for monetary expansion and, quite often, commercial banks are required to hold part of their reserves on terms of government bonds. These reserve requirements are, in turn, often quite high, the fact that the central bank uses government bonds as a base for monetary expansion means that the government debt tends to get monetised. The ensuing inflation interacts with the reserve requirements to harness even more revenue for the government (the inflation tax). The process of revenue extraction by the government is much too complicated to be captured in a single diagram but one can imagine this process taking the form depicted in Figure 16.1. Primary savers—the public—save in the form of currency and cash, which are liabilities of the central bank. They are also liabilities of the commercial banks. Part of these are effectively transferred to the central bank through reserve requirements and purchase of government debt, leaving only a fraction for free lending. The public also saves with savings banks and post offices. These savings are channelled in a similar manner. The central bank has to transfer resources to the government (ministry of finance) through purchase of government bonds. It also has to lend directly to certain ‘priority’ sectors labelled I, II, etc. in Figure 16.1. These priority borrowers may be industrial development banks,

Figure 16.1


397

agricultural development banks, public sector undertakings, private sector undertakings in the ‘core’ sector and the like. COSTS OF ECONOMIC REPRESSION One of the principal problems with economic repression is the inflationary bias that sets into the economy as a consequence of the repression. This inflationary bias is the result of both demand-pull and cost-push factors. The fact that market prices are highly distorted pushes up costs of production. Moreover, the government adds very significantly to demand-pull pressures on inflation. On the one hand, the government has large and growing expenditures, particularly on an unproductive and unprofitable public sector. Their losses have to be absorbed and the government has to spend vast sums on subsidies of various types—principally those on food, fertilisers and industrial inputs. At the same time, government revenues are rather limited. Since incomes are low, income taxes are unimportant. A significant portion of commodity taxes is avoided because of irrational tax structures and corrupt tax officials. Government deficits are then financed by the government issuing bonds that it forces on to the banking system by making it mandatory for them to buy government bonds (private investors would not purchase these bonds readily). The banking system then uses these bonds as reserves for monetary expansion. Hence, the government deficits get monetised and inflationary tendencies are exacerbated. It has sometimes been argued that perfectly anticipated inflation has relatively small costs because markets have a chance to adjust completely and the values of real variables such as the real interest rate or the real wage rate remain unaltered. In most developing countries, however, inflation is the result, as argued above, of the government spending beyond tax collections and monetising the deficit by forcing its bonds onto a captive banking system. Hence, inflation is largely unanticipated. Further, key economic variables such as the interest rate and the exchange rate are not free to adjust so that inflation, even if fully anticipated, involves significant economic costs. Government policy is inflationary: sometimes the attempt is to get additional revenues in order to provide additional resources for investment, and at other times it is a direct consequence of the system of controls, e.g. a food subsidy or a fertiliser subsidy. Once inflation begins, governments have a tendency to exacerbate its ill effects by imposing additional controls. One of the most important of such controls is the pegging or slow adjustment of the nominal exchange rate. Other significant distortions are the fixing of prices of key intermediate and infrastructural sectors such as coal, steel and transport in order to ‘control inflation’. As a consequence relative prices get distorted further. These are all very important markets and controlling them imposes significant costs on the economy. Even when there is an attempt to liberalise the economy

398


there is still an attempt to ‘control inflation’ by controlling administered prices rather than freeing these prices. It is difficult to direct controls and it is not possible to eliminate them without taking into account the underlying macroeconomic structure. The same political and economic forces that lead to inflationary tendencies lead to balance of payments difficulties as well. The nominal exchange rate is adjusted much too slowly. This leads to an excess demand for foreign goods. The government either rations this excess demand or allows it to be satisfied by running down reserves or borrowing from abroad. The piling up of debt permits continued expansionary fiscal policies and excessive government expenditure. Once the accumulation of the debt becomes a crisis its payment imposes additional strains on the home economy as budgetary, and trade surpluses have to be accumulated in order to pay off the debt. The urgency of paying off the debt has the undesirable consequence of focusing attention on the external sector to the neglect of the excessive distortions in the home economy. Indeed the ill effects of high internal debt, the blurring of the distinction between fiscal and monetary policies and the endogenising of the money supply to reflect budgetary policies tend to get understated. External debt tends to get emphasised for the simple reason that it is more visible. Liberalisation of the external sector would require that the domestic price of import-competing home goods fall relative to the domestic price of exportables. This can be achieved by sharp devaluations combined with a freeing of imports. This devaluation will itself lead to a large jump in the price level and some contractionary effect on domestic employment and output. This could well turn out to be counterproductive as we have learnt earlier in Chapter 12. Some argue that, because of such difficulties, it is more appropriate to opt for a floating exchange rate. However, freeing the external sector with other financial markets repressed may not be the best policy measure. Moreover, the simultaneous freeing up of all financial markets—the so-called cold shower effect—is too extreme a policy measure and its short-term social costs in terms of the unemployment of workers in failing previously protected industries may be too much to bear for countries with very little social insurance and safety nets for the unemployed and the poor. However, more liberalisation programmes have faltered on their inability to control inflation than any other single factor. In practice liberalisation can be combined, albeit imperfectly, with an anti-inflation programme. Three significant components of such a policy initiative would be: (i) a crawling peg rather than sharp devaluations which dampen inflationary tendencies and expectations before freeing up the exchange rate completely, (ii) safeguards to ensure that the government does not resort to price controls to ‘control inflation’ and does


399

not resort to excessive spending, and (iii) simultaneous efforts being made to loosen controls in other financial and key sectors of the economy. CONSEQUENCES FOR ECONOMIC GROWTH It is now fairly well established that countries that have sustained high real rates of interest and more stable price levels have had more robust financial growth. In Table 16.1 we present evidence of this from various issues of The International Financial Yearbook published by the International Monetary Fund (IMF). Ratios of M3 (currency, savings deposits and shorter-term deposits in banks and other quasi-financial institutions such as post offices) to gross national product (GNP) are typically higher in the high growth Asian economies than the highly inflationary Latin American economies. As a matter of fact even the slower growing Asian economies tend to be more financially developed than Latin American countries. However, all slowly growing economies tend to have low financial development—low levels of M3/GNP ratio. A number of authors, including Gelb (1989), show that high real interest rates tend to be associated with high growth rates of real gross domestic product (GNP) as well. The evidence is definitely suggestive, if not compelling, evidence of the deleterious effects of financial repression on economic growth. Gelb discovered that in a cross-section of LDCs varying investment efficiency was strongly and positively correlated with the average real deposit rate. It might be argued that apart from the investment efficiency effect, changes in the deposit rates will have effects on the volume of savings and investment. It might be argued that higher deposit rates gained by increasing other real interest rates might lead to a drop in the volume of investment. Gelb quantified the strengths of the two effects and discovered that any negative effect on the volume of investment was more than outweighed by the efficiency of investment effect. Table 16.1 Domestic credit/GDP

Source: IMF, International Financial Statistics, various issues.

400


PROBLEMS OF LIBERALISING INDIVIDUAL MARKETS The general theory of second best tells us that when there are several distortions in an economy it may or may not be second-best optimal to remove the distortions in individual markets. We consider here, somewhat tentatively, the welfare gains and losses from liberalising individual markets/sectors in developing countries. Consider, first, the most important sector in most developing countries: agriculture. Farmers may safely be assumed to base their supply decisions on the relative prices for their produce they face. Usually producer prices for agricultural goods are fixed in developing countries in order to provide food at relatively low prices to urban workers. Freeing up producer prices will, invariably, increase agricultural output, ceteris paribus. In the very short run urban workers may have to pay higher prices for food. But the combination of higher agricultural supplies and a higher demand for industrial goods in agriculture as rural incomes increase will mean that their welfare will also increase. Hence it must be welfare improving to free agricultural prices even when the home currency is overvalued. This argument does not carry over very easily to the case of a country with comparative advantage in cash crops. If resources were allocated efficiently these countries would import food. Freeing up producer prices would then lead to higher demand for food and, if food is bought under licence in a regime of fixed nominal exchange rate, higher food prices. This could lead to a drop in welfare. Freeing up financial markets and reducing financial repression can improve the growth rate of the economy and welfare. As interest rates rise and, in particular, when the real interest rate becomes positive, savers are willing to save more. Overvaluation of the home currency and ceilings on domestic interest rates work in the same direction—they tend to reduce the incentive to save and increase the capital intensity of projects. Hence, irrespective of whether the home currency is overvalued, it will be welfare improving to reduce or eliminate financial repression. We can make similar comments about the labour market. In most developing countries urban wages are well above market-clearing levels, mainly because of unions which exercise considerable political clout. Freeing up the labour market would improve welfare. Particularly, if there are exchange rate overvaluation and/or interest rate controls, a reduction in the real wage would lead to the adoption of more labour-intensive techniques of production and higher employment. Similar remarks can be made about changing the value of the exchange rate to more realistic levels even if there are other distortions in the economy. Apart from bringing domestic prices more in line with true international prices, such a measure will reduce the import substitution bias of domestic


401

industrial strategy. Foreign industrial goods become relatively attractive and the world demand for exports improves. It may safely be concluded that liberalising capital outflows and inflows may not always be welfare improving. If domestic interest rates are controlled, liberalising capital outflows will mean a flight of capital from the home country as savers shift savings to foreign markets where rates of return are higher. Even if domestic financial markets are completely liberalised there might be considerable exchange rate risk in holding one’s savings in the form of home currency. It would, therefore, appear that except for capital flows and the agricultural sectors of countries with comparative advantage in cash crops, liberalisation of individual markets/sectors is likely to be welfare improving. THE SEQUENCING OF ECONOMIC REFORMS It has often been argued that simultaneous and large-scale reform on a wide front is the best way to proceed. New signals will be set up that will not involve the resource misallocation in response to altered signals before the transition is complete. This conclusion gets further strengthened by empirical evidence that suggests that uncertainty about the directions of future reforms during a process of slow change delays response to altered signals. There are some who argue that the complete and total dismantling of all restrictions may not be a good idea, i.e., the costs of large-scale reforms are very high in LDCs. With large-scale reforms many inefficient industries may have to shut down with consequent unemployment. Typically there will be a significant time lag before new more efficient industries can come up. Moreover, these new industries may have considerably different skill requirements as compared to the industries that had closed down. Workers may have to suffer protracted unemployment. With very little social programme to support these workers, the short-run welfare costs of quick and widespread reforms may be very high. A government that has time on its hands can afford to go slow. Typically this will involve reasonably good economic performance at home and a comfortable foreign exchange position. Unfortunately, most developing countries do not have such luxury. When the domestic economic performance is satisfactory and the foreign exchange position comfortable, governments hardly ever think of deep economic reform. If they did, the social costs of such reform would be much lower and the path of transition to a more efficient economy much smoother. Usually reform is initiated in the face of deep economic crisis—a sluggish domestic economy combined with the possibility of not being able to meet external payments obligations. The more severe the situation the greater the possibility that reforms will have to be widespread and quick. The consequent costs to the poor are then high. It is indeed deplorable that many LDC governments have, in the past, delayed reforms when they were easy to carry out and been

402


forced to effect much deeper adjustments when the social costs of adjustment have gone up considerably. Suppose now that large-scale and complete reforms cannot take place all at once. Is it possible to provide some direction for governments on the sequencing of reforms? This is the issue of optimal sequencing of economic reforms that we consider now. To be sure, the nature of economic crisis that precedes serious economic reforms is different for different countries. In most developing countries this involves the government controlling its expenditures and increasing its tax revenues. Control of government expenditure is extremely important for controlling inflation. As we have remarked earlier, the ability to control inflation is perhaps the most important measure of the success of a programme of economic reforms. Government budgetary deficits in developing countries are highly inflationary because governments tend to dump their debt on captive central and commercial banks so that the debt gets monetised. This, in turn, exacerbates inflationary tendencies. Control of government expenditure involves reducing or eliminating subsidies. Typically, these are subsidies for food, fertilisers and for inputs used in the manufacturing sector. Reducing these subsidies has the dual advantage of reducing the government deficit and reducing price distortions. In addition, the government may contemplate selling off unproductive and unprofitable state undertakings. At the same time as the government attempts to reduce its expenditure it must also attempt to increase its tax revenues. In most developing countries the most important forms of taxes are indirect taxes on domestically produced goods and foreign produced imports. Because personal incomes are so low, direct taxes are relatively unimportant. Governments must try to broaden the tax base by bringing a larger number of goods, services and incomes under the tax net. Further, governments must attempt to reduce tax evasion by lowering marginal tax rates and streamlining the tax collection mechanism. In the 1990s the credibility of most governments, particularly LDC governments, was low. Hence strict fiscal discipline at home was essential to send the right signals to domestic and foreign investors. Governments must demonstrate their ability to finance their expenditures in a non-inflationary manner, i.e. without printing additional money. In addition governments must reduce accumulated foreign and domestic debt. The next step in the process of economic reform is the freeing up of domestic credit markets so that depositors receive, and borrowers pay, substantial real interest rates. But unrestricted borrowing and lending in free capital markets can take place only when the price level is relatively stable and government finances are under control. Without price level stability, real and nominal interests are going to be unstable. This will increase the risk associated with and adversely affect the volume of borrowing and lending.


403

This implies that until such time as the price level has stabilised, bank credit must be strictly controlled by the authorities. Once domestic financial markets have been liberalised the government is well advised to decontrol foreign transactions. The prevailing wisdom is that the current account is best liberalised before the capital account. The capital account should not be liberalised too soon because this might lead to substantial capital flight from the domestic economy. A liberalised current account means that home exports become more competitive in foreign markets. Before such liberalisation of the trade account is undertaken domestic prices must be rationalised and decontrolled, government finances must be in order and the system of multiple exchange rates with several different exchange rates (depending on tariff and other restrictions) must be abandoned. In their separate works on trade liberalisation Bhagwati (1978) and Krueger (1978) defined liberalisation as consisting of the replacement of quantitative restrictions which are not visible with tariffs that are. The next step in this arrangement is the reduction of the mean and variance of the tariffs. As remarked earlier, the rationalisation of foreign trade policy does not imply that full convertibility of the home currency be extended to the capital account. Before allowing enterprises and individuals to borrow and lend freely from abroad, domestic financial markets must be fully liberalised and exchange rate risk associated with the home currency reduced so that liberalisation of the capital account does not involve large-scale capital flight from the home economy. This depends crucially on the restoration of fiscal discipline on the part of the government and the attainment of price stability. THE INTERNATIONAL CONTEXT OF ECONOMIC REFORMS An implicit assumption behind the analysis of economic reforms contained in this chapter is that the international economy is responsive to the reforms being undertaken by the developing countries. In concrete terms this assumption means several things. For instance, it could mean that when the developing country gets its finances in order and starts producing efficiently it is able to sell more abroad and earn additional foreign exchange. It could also mean that the international community is willing to transfer funds on easy terms to the concerned LDC in order to reduce the pain of deep-rooted structural reforms. Clearly both these conditions are likely to be satisfied only when the economies—particularly those of the developed countries— are growing rapidly. It is ironic that when the world economy was growing most rapidly—in the decades of the 1950s and 1960s—most developing countries were inclined towards economic repression. The Feldman-Mahalanobis model was at the

404


peak of its popularity because it provided an economic rationale for the suspicions of the newly independent countries towards trade with the western world, important members of which were their former colonisers. The newly industrialised countries (NICs) (Hong Kong, Singapore, Taiwan, South Korea and the like) were among the very few exceptions to this rule. It is not surprising that these NICs attained very high rates of economic growth whereas much of the rest of the developing world lagged behind. It should further be remembered that the rapid expansion of the world economy during this period was itself, at least partly, a result of far reaching tariff and other trade barrier reductions in most developed countries in Europe, the USA and Japan. The NICs went in for deep economic reforms, indeed integrated themselves into an expanding world economy. Their products found ready markets in the post-war economic boom. The costs of economic reforms in these small economies were low. These economies prospered. A second major wave of economic reforms in developing countries was initiated during the 1980s—a consequence primarily of IMF pressures in the wake of LDCs seeking loans in the aftermath of the oil price shocks of 1973 and 1979. The early 1980s were a period of economic expansion for the world economy but the oil shocks had created a tentative atmosphere for continued trade liberalisation. Protectionist pressures were strong in most developed countries as they tried to minimise the harmful effects of the oil price shocks on their own economies. The last few years of the 1980s were characterised by deep world recession. Protectionist pressures were again strong. In this atmosphere there can be no presumption that developing countries will readily find markets for their products in the developed countries’ markets. Moreover, the atmosphere for easy credit availability to tide over the pains of structural adjustment has certainly not improved. How bad was this protectionist environment? A World Bank study in 1987 showed that about 21 per cent of developed countries’ imports from developing countries in 1986 were subject to formal non-trade barriers (NTBs): about 17 per cent in the USA, to 22 per cent in Japan and 23 per cent in the European Common Market. Eighty per cent of clothing, 61 per cent of textile yarn and fabrics, 55 per cent of steel products, 27 per cent of footwear and 13 per cent of chemicals imported by developed countries from developing countries in 1986 were subject to NTBs. NTBs in developed countries as a whole increased by about 16 per cent between 1981 and 1986. The protectionist movement had been most conspicuous in the USA. But other countries including Japan were quite protectionist in more subtle ways. One of the principal ways in which NTBs are applied to developed country imports from developing countries is punitive restrictions following a determination that the LDC has been ‘dumping’ its products in the markets


405

of the developed country. In the USA there is a special law to deal with this— Article 301. Article IV of the General Agreement on Tariffs and Trade (GATT) defines dumping as an act in which a product is sold in another country at less than the ‘normal’ value of the product which is defined as (i) the home market price and/or (ii) a ‘constructed value’ based on an estimated cost of production plus ‘reasonable sales expenses and normal profits’. This last value is a very abused concept. In estimating ‘constructed value’ developed countries often add an amount for administrative expenses for its export sales because such expenses are paid by the importer, the importing country may add all the expenses associated with advertising in the domestic market. The US antidumping law even sets a minimum level of selling costs at 10 per cent and a statutory minimum profit margin at 8 per cent, effectively raising the normal value to meet the dumping charges. The European Community (EC) take the selling costs actually incurred and the profit margin actually realised by the firm in question. In recent proceedings the latter varied between 1 and 33 per cent. Moreover, it is easy in developed countries to initiate an anti-dumping or countervailing duty charge. Since processing procedures are inordinately long, and since very harsh ‘provisional’ measures based on information supplied by the petitioners are applied, the anti-dumping and countervailing duty cases have produced monstrous ‘procedural protectionism’. Unfortunately, these arrangements have allowed vested interest groups in developed countries to undermine the efforts of exporters in developing countries. Anti-dumping provisions have been widely used in the EC as well. As a matter of fact this has been the most important instrument of EC protectionist policy. Beseler and Williams (1986) report that between 1980 and 1985 there were 254 investigations and punitive action was taken in 209 cases. The prevalence of NTBs and other protectionist measures in developed countries introduces a measure of uncertainty in trade planning in developed countries. To be sure, the WTO recognises that developing countries may have to impose tariffs to protect ‘infant’ industries. But it is required that these tariffs be removed soon after the industry gets going. For various reasons (see Corden, 1977, for example) developed countries may find it necessary to subsidise some industries. Some of these industries may, for example, create external economies for other industries. Typically, in such situations, it is advisable to subsidise the costs of this industry since its expansion has beneficial effects on the rest of the economy that are not completely captured by market prices. In practical situations there might be disagreement between developed and developing countries as to which industries deserve protection and this may lead to the imposition of NTBs against LDC exports. Nam (1986), for instance, notes that the burden of proof as to whether factors other than subsidies may have been a major cause of injury, or whether initial findings on the amount of subsidies were appropriately estimated,

406


normally falls on the exporters. Such a proof requires not only a large amount of information, but also expensive legal costs which may be too burdensome for many developing countries to bear (Nam, 1986:19–20). As a result of the Uruguay Round and national tariff reforms, average tariffs of many countries have been reduced to relatively low levels. However, it would be erroneous to assume that tariffs have ceased to be a major problem for developing countries trying to make inroads into the markets of developed countries. Even after the full implementation of all Uruguay Round concessions a substantial number of high tariffs on imports from developing countries will remain in place in developed countries. Both frequency and tariff levels are a matter of concern. After adjusting for all presently applied tariff suspensions as well as general GSP concessions applied in favour of developing countries, it turns out that developed countries would still maintain high and quite variable rates. As Table 16.2 shows, for important products their tariff peaks reach 350 to 900 per cent, although the majority of their tariffs range from 12 to 30 per cent. However, one-fifth of the peak tariffs of the US, about 30 per cent of those in the European Union and Japan and about one-seventh of those in Canada exceed 30 per cent. In contrast, developing countries apply rates above 12 per cent ad valorem more frequently than developed countries but have fewer extremely high rates. The degree of success that can be achieved by the reforms of developing countries, then, depends partly on the outcome of the tussle between protectionist and liberal forces in developed countries. Past experience suggests that if multilateral trade institutions function effectively it would be possible to contain the protectionist forces. At the transnational level, then, an important element of the strategy of developing countries should be to work toward the success of the WTO and other multilateral institutions such as UNCTAD. This is their best bet for helping the forces favouring free trade in developed countries. Several basic principles should guide LDC efforts at multilateral organisations and bilateral negotiations. First and foremost, policy makers in developing countries must take these negotiations very seriously. A history of import-substitution type strategy had led many developing countries to adopt a rather lackadaisical attitude toward such negotiations. It is amply clear that reciprocity is important. If developing countries seek reductions in tariffs and other barriers in developed countries they should be prepared to grant similar concessions on their own part. Moreover, it should be realised by all concerned that the establishment of a truly international liberal trading order is in the best interests of all trading countries. The current tendency towards forming regional free trade blocks has to be eschewed.

Table 16.2 Distribution of tariff peaks by product groups


Table 16.3 Post-UR tariffs on exports from developing countries

Source: UNCTAD TD/B/COM.1/14/Rev.1 (January 28, 2000)



413

Attitudes toward the WTO have to be improved. At the present time the future of the current round of negotiations for tariff reductions known as the Uruguay Round is uncertain. There is considerable ill-will and suspicion among the major developed countries over agricultural policy and intellectual property rights. Table 16.3 indicates the high level of tariffs on developing country exports. The fact that most of these countries are currently going through a prolonged recession is not helping matters either. Developing and developed countries should realise that the scope of the WTO should be expanded to cover all aspects of trade. The present recession is fanning protectionist pressures in most countries. But at the same time some of the most productive areas of international trade cannot easily be controlled through national tariffs and NTBs. These include items like information systems transfers, which can be trade via satellite. Such trade and the progressive liberalisation of many previously highly repressed economies (developing countries and former socialist economies) provide strength to the forces that want international trade to be free. It is hard to predict how matters will ultimately be resolved. Suffice it to say here that a more liberal world trading arrangement will be to the benefit of all. The success of the current phase of reforms in developing countries depends critically on the liberalisation of trade in developed countries. To the extent that they may influence developed countries’ policies, multilateral institutions such as the IMF and the World Bank should attempt to make developed countries’ markets more open for developing countries’ exports. Such efforts would supplement and enhance the effectiveness of the multilateral institutions’ insistence that developing countries open up their markets and reduce economic repression. DISCUSSION QUESTIONS 1

Consider a two-country world economy—a rich ‘North’ and a poor ‘South’. Both economies are Keynesian. Demand for the output of the North is written as:

where yn is Northern output, cn is the propensity to consume, An is autonomous expenditure in the North, Xn are Northern exports and mn is the propensity to import. The parallel equation for the South is written as:

where the subscript ‘s’ refers to the ‘South’. The regions’ interdependence arises because Northern exports are Southern imports and Southern exports are Northern imports. Thus:

414

2


Solve for the equilibrium levels of income in the North and South and show how these depend on respective propensities to consume and import. Which region will have the higher income? Draw a diagram with yn and ys on the axes. Graph the yn(yn, ys) and ys(yn, ys). Under what conditions will Now suppose that the two economies are asymmetric. In particular, in line with the Columbia economist Ronald Findlay suppose that the Northern economy is of the Solow type whereas the Southern economy is of the Lewis type. The rate of growth of the North is the natural rate of growth discussed in Chapter 10, i.e., n + g. Growth in the South, however, is constrained by the rate of capital accumulation. Further, North exports capital goods to the South. Let P be the terms of trade, i.e., the rate at which the North’s capital good exchanges for the South’s primary product (taking Ps and Pn as price per unit of Southern and Northern goods, Hence growth of Southern capital is given by If North and South were to grow at the same rate we must have This equality of growth rates is, however, consistent with rising inequalities in per capita income across North and South. Why?

An alternative view of economic growth For the structuralist school state-led development was the key to rapid economic growth. Although its influence has declined considerably, the structuralist school of development economics has had a lasting impact on debates regarding development, especially with respect to Latin America and other regions with similar problems. The structuralist school argues that for poor countries economic development may only be achieved through an internal expansion of the home economy. The structuralists focus on the mechanism by which ‘underdeveloped’ economies transform their domestic economies from a traditional subsistence agricultural base into a modern economy, defined as one in which most of the population is urban and the bulk of the country’s output is in the form of manufactured products or services. Hence the most important issue to be addressed is how to expand the modern economy while contracting the indigenous traditional economy of the country or region. The object of development is the structural transformation of underdeveloped economies so as to permit a process of self-sustained economic growth. This can only be achieved by eliminating the underdeveloped country’s reliance on foreign demand for its primary exports (raw materials) as the backbone fuelling economic growth. Economic growth must be fuelled through an expansion of the internal industrial sector. The structuralist school


415

emerged in Latin America in the 1940s. In the latter part of the nineteenth century and the beginning of the twentieth century Latin American countries were exporters of raw materials. Classical economics held that the region had a comparative advantage in raw materials, meaning they could produce raw materials more efficiently than other regions. As such, they should have concentrated on expanding such exports. By the 1940s Latin American economists began to attack this notion. They argued that export-led growth of raw materials was no longer a feasible path to economic development. This was because the price of primary exports was declining while the price of manufactured products was increasing. In addition, the supply of manufactured goods was decreasing due to World War II. All of these factors created large disruptions in the economies of Latin American countries. Since prices for primary goods in the international market have a tendency to not rise as fast as prices of manufactured goods, developing countries were unable to earn enough to pay for all of the imports they needed—particularly high-priced manufactured products. Many Latin American economists believed that the situation would not improve following the conclusion of the war. They cited two reasons in support of this conclusion. First, advances in technology, which lowered the production costs of manufactured goods, were not resulting in lower-priced imports of such goods. Structuralists argued that the fruits of those advances were being retained by the industrialised nations in the form of increased profits for the manufacturer and higher wages for the workers. Second, structuralist economists warned that, given the United States’ role as the world’s new industrial leader, the demand for raw materials was going to diminish because the United States was rich in natural resources. Consequently, the United States would be willing to buy raw materials abroad only if it was cheaper than extracting them at home. But this scenario was unlikely because the post-war global economy differed substantially from the situation during the previous century. At that time, Great Britain, an island nation, was the pre-eminent economic power, but it had few natural resources. In order to feed its industries, Britain had to import raw materials. This gave providers of raw materials more bargaining power, power that developing countries would no longer have in the post-war world. Structuralists also argued that focusing on overall economic growth numbers was a necessary but not sufficient step in pursuing economic development. In an underdeveloped economy the technological levels of one or more sectors of the economy fall below the technological level of the most advanced sector, especially if technology exists that will enable those sectors to be more productive. To the structuralists, development had to include the expansion of new technology and methods of production in order to eliminate the gap between the most advanced sectors of the economy and those that lagged.

416


For the structuralists a situation in which the bulk of the economy’s output was derived from the primary product sectors was unacceptable. In addressing the cause of underdevelopment, structuralist economists focused on the evolution of economic relationships between developed countries and the rest of the world. Developing countries were brought into the international economy to serve two purposes: (i) to supply cheap raw material and (ii) to purchase finished manufactured goods from industrialised economies. This gave rise to ‘enclave’ economies in developing countries that expanded the primary product export sector at the expense of the industrial sector. The structural relationships in the international economies led to a dual economic structure in developing countries, where a modern economy (the export sector) coexisted with a backward and undeveloped one. The modern sector was maintained not through internal innovations and advancement but by purchasing new technology from the developed countries. As long as dualism persisted, autonomous economic development would be impossible; that is, growth would be dependent on the industrial countries. Structuralists argued that economic growth had to stem from internal demands. From this followed the conclusion that the structural transformation of the economy could only be achieved through government intervention. Hence, structural changes needed to bring about economic development could only be achieved by state intervention. For example, government-imposed tariffs on imports were designed to stimulate the internal market by protecting new industries within the country. A tariff was viewed as a way to even the playing field between a manufacturer in an industrialised country and one in a developing nation. The former tended to have better access to capital and technology as well as a more productive workforce. These factors enabled manufacturers in industrialised countries to produce a given product faster and cheaper than ‘infant industries’ in developing countries. Another important component of the structuralist approach was state-owned enterprises. The structuralists believed that, given the underdeveloped capital markets in developing countries, only the state could generate and manage the sizeable investments needed to industrialise. Protective fiscal and monetary policies were also advocated. In sum, all of these policies, known collectively as ‘importsubstitution’, were geared at encouraging the country to industrialise. Thus, the structuralists accepted the notion that development was to be achieved through capitalism. But they were not convinced that the market alone could achieve the type of thriving capitalism that industrialised countries were enjoying. Governments of developing countries had to actively promote industrialisation through government regulation of the economy. However, structuralist policies enjoyed only limited success. Countries that adopted the import-substitution model


417

of development began to notice in the 1960s that government-led initiatives to industrialise could not effectively create the most important phase of industrialisation relating to heavy machinery and plant installation. Moreover, the heavy involvement of the state in the market created inefficiencies that eventually caused major internal and external economic problems. The drive to industrialise led, ironically, to increased dualism in developing countries as the gap between the rich and the poor widened. Thus although the analysis of the structuralist school that stagnating terms of trade for developing countries exporting primary goods is a problem is right, the school underestimated the costs of state-led industrialisation.

17

The international financial architecture and the developing countries

INTRODUCTION The international financial architecture (IFA) is a crucial aspect of the world economy. As the name itself suggests the IFA is the base on which the superstructure of the international payments system rests. The basic argument for considering the IFA seriously is that it is essentially an international public good. Having an orderly system of international payments and transfers creates a conducive atmosphere for the growth of trade and commerce, which ultimately is of immense benefit to the developing countries. When the IFA is ill-formed or does not function smoothly there can be severe repercussions as was most recently discovered during the East and South-East Asian crisis of 1997 and thereafter. It is in the interest of everyone to have a wellfunctioning IFA in place. However, the design of this IFA requires cooperation among these players and the distribution of the benefits from such cooperation becomes a contentious issue. The IFA is a catch-all term involving many substantive issues. At the very minimum it covers the following range of topics: 1 2 3 4 5

6 7

exchange rate policy coordination of exchange rates policies toward the capital account, i.e. whether there are one sided or both sided capital controls1 the system of international payments including the functioning of institutions such as the BIS the role of international institutions particularly the World Bank, the WTO and the IMF, which may act as an international lender of last resort (ILLR) the framework for dealing with financial crises, trade collapse and other emergent situations structuring policy responses toward the long-term development needs of the developing countries

420

8 9


the development of standards for reporting of financial data in a uniform manner by all countries developing appropriate regulatory structures for international financial institutions

The IFA as it exists today is inherited from the Bretton Woods arrangements made just prior to the end of the Second World War. Basically three types of international institutions were created: the IMF to oversee the system of international payments, the World Bank to oversee the development requirements of the developing countries and the General Agreement on Tariffs and Trade (GATT) to coordinate international trade policies.2 As the world economy grew and, particularly, since the last few years, a need has been felt for a review of these international arrangements as the land-scape for international economic relations has gone through considerable change. To mention just three of these—international financial policy in the 1950s and 1960s occurred against a background of (largely) fixed exchange rates. Public capital flows to developing countries were then much more important than private capital flows and the problem of public bads—such as global warming—were unheard of. In view of these and other qualitative changes some economists have called for the development of a new financial architecture. See, for example, Eichengreen (1999), Goldstein (2000) and United Nations (1999). Another factor underscoring the importance of IFA is the completion of the Uruguay Round of Trade Talks and the expansion and consolidation of the WTO. Despite appearances to the contrary, there does seem to be an impression that the future holds even lower barriers to trade and investment flows. Another reason for the increasing importance of IFA is the need to support the credit and financial requirements of transition economies as well as the heavily indebted developing countries (HIDC). Given the declining role of official development assistance to many of these countries, the IFA needs to evolve market-based solutions to these problems. Furthermore, there is the whole issue of choice of exchange rate regime by developing countries in the face of possible sharp reversals of capital flows. A further important question is the appropriate policy response of these countries and that of international financial organisations such as BIS and IFC and, in particular of an international lender of last resort (ILLR) such as the IMF. This chapter cannot hope to cover all of these issues in any great depth. Hence we must choose to focus. We consider three aspects here. First, we consider the problems associated with the explosion in financial flows to developing countries. Second, we briefly evaluate the role of three major international institutions: the World Bank, the IMF and WTO. Third, we consider some safeguards that could be developed to ward off currency and stock market collapse of the sort that were seen in the late 1990s.

International financial architecture and the developing countries

421

Several factors have combined to make the IFA a matter of singular importance in recent times. First, is the sharp drop in aid flows, particularly to the poorest countries as detailed in Table 17.1. Further, in many developing countries there has been a sharp increase in the net capital flows although most of these flows occur between developed countries. As can be seen from Table 17.2, portfolio flows were much larger than FDI flows. Countries that are aid dependent also have the least capital flows going to them. There have been ups and downs in the volume of total and components of development finance over the years. For instance, while there was an upsurge in private flows in the early 1990s, the volume of official flows has shown a downward trend since the beginning of that decade. In addition, there has been a change in relative shares of foreign direct investment (FDI) and non-FDI components of private finance, just as there has been changing composition of official flows. Multilateral flows are expanding at the expense of bilateral flows, project aid takes precedence over project finance and aid is getting increasingly tied. CAPITAL INFLOWS AND RESOURCE TRANSFERS TO DEVELOPING COUNTRIES Developing countries typically have low rates of savings and investment. A higher savings rate would raise their economic growth. In this sense, the primary benefits of capital inflows are the opportunities these provide to augment national savings, thus increasing economic growth and living standards. Foreign savings, especially if they are embodied in new machines, often act as conduits for the transfer of modern technology to developing countries. This is typically the case with foreign direct investment (FDI). FDI often also brings with it management skills. However, there is the risk that capital inflows may be used to augment current consumption. This will, of course, reduce savings. Obstfeld (1998) has summarised the literature on this topic. An important question to investigate, then, is the extent to which capital flows augment savings and investment. Bosworth and Collins (1999) consider this important issue. They use the following specification:

where I/Y and S/Y represent investment and saving as a percentage of GDP; i indexes countries and t time. Xit denotes explanatory variables such as (i) total capital inflow as a percentage of GDP,3 (ii) the rate of growth of real GDP with one or two year lags, (iii) change in the terms of trade. Countryspecific effects are modelled through country-specific intercept terms α0i and ß0i. Saving is measured as the sum of investment and the current account

Table 17.1 Aid flows to developing countries: select characteristics

Source: 2001 World Development Indicators, The World Bank.

Source: The OECD’s DAC Journal (2000) Development Co-operation, 2000 Report, 2(1), pp. 180–1.

Table 17.2b The total net flow of long-term financial resources from DAC countries to developing countries and multilateral organisations by type of flow (current US$ billion)

Source: The OECD’s DAC Journal (2000) Development Co-operation, 2000 Report, 2(1), pp. 178–9.

Table 17.2a Total net resource (current US$ billion)


425

balance. Hence the impact of a change in X on the real resource transfer (the current account deficit) is simply They recognise that capital flows may be endogenous: just as investment and saving depend on capital flows so, too, can capital flows depend on saving and investment. This two-way endogeneity may lead to biased coefficients in regressions. To uncover the effect of capital inflows on saving and investment, therefore, instrumental variables techniques are used. Bosworth and Collins find external factors—particularly US interest rates, deviations of real US GDP from trend and a measure of total gross capital flows to developing countries to be important determinants. The results of the estimation are very interesting. First, the type of capital inflow has an important bearing on investment. FDI has the strongest (and most positive) relation with savings and investment. Portfolio inflows have the smallest and least significant relationship and loans are in between. Thus FDI flows are a more important method to augment domestic savings and investment than other types of external flows. Further, the three types of financial inflows also appear to have sharply different implications for the current account and, therefore, for savings. In particular, the coefficient of FDI in the savings equation is particularly large. Furthermore, all the observed negative correlation between total capital inflows and the current account is associated with loans, which raise investment, but lower savings. Portfolio inflows have little impact on investment, savings or the current account. There is a widespread belief that international capital flows can help augment domestic savings for purposes of investment. Bosworth and Collins show that FDI is the most important factor influencing investment. Other types of financial flows—particularly portfolio flows—have typically involved countries incurring current account deficits, which they can supplement with a capital account surplus (capital inflows). However, such a strategy is fraught with difficulties if pursued without adequate caution as the recent East Asian experience has shown. Thus, there is a need for international management of these changes. With such large capital flows, the emphasis has shifted from the current account imbalances of developing countries to their capital account imbalances. At the same time, it is important to recognise that most developing countries have had rather limited success in attracting FDI inflows although portfolio flows have been substantial. China has been the most successful followed by some Latin American countries. Even large, rapidly growing LDCs such as India have had rather limited success in attracting FDI inflows. In Table 17.3 we report the FDI experiences of select Asian countries. As can be seen, many large countries such as India finance only a small portion of their investment through FDI. Portfolio investments are higher and more volatile. The 2001 World Investment Report brought out by UNCTAD predicted that although global FDI was to rise to one trillion dollars

426


Table 17.3 FDI inflows into select Asian countries

Source: World Investment Report 2001.

in 2002, its distribution would be highly uneven. In 1998, for instance, FDI flows into developed countries were $636 billion while it was only $208 billion for developing countries. Of this amount more than $43 billion went to China alone (Table 17.3). For the entire continent of Africa FDI inflows in 1998 were only $8.08 billion—less than one-fifth of China’s. Some pickup is expected in the transition countries of Asia, whereas Latin America was expected to increase its share. This will now be affected by continuing grim prospects of Argentina. The prospects for Africa are hardly improved.


427

THE ROLE OF INTERNATIONAL FINANCIAL INSTITUTIONS The IMF as an ILLR So far as developing countries are concerned, the IMF serves a very important purpose. It is the one organisation that they can go to when they face an international payments crisis. In Chapter 11 we have discussed the problem of structural adjustment in which the IMF asks the countries that borrow from it to engage. It is important to understand that the external payments problems that were discussed in Chapter 11 originated in the current account. The external payments problem that arose during the Mexican peso crisis or the East Asian crisis originated in the capital account. For example, in 1997 the Thai economy had a healthy fiscal situation so cutting down public expenditures was not really warranted. However, it has taken a little while for the IMF to adjust to this change in the origin of crises. We will evaluate the Fund’s approach to the handling of these crises later in this chapter. At this time we want to discuss the structure of the IMF and the design and evolution of its programmes. Particular attention will be paid to its role as an ILLR. Since a number of excellent treatises exist on this issue, our discussion will necessarily be brief. We will consider, in turn, the decision-making structure within the IMF and then the effects of the IMF programmes themselves. The decision-making structure of the IMF Since the IMF is an international organisation, its decisions have to be a properly weighted aggregate of the decisions of various member countries. The weights that any member country’s decisions will carry depend upon three factors: (i) the country’s quota position, (ii) the quota formula which determines a country’s votes, and (iii) the voting structure within the IMF’s board of directors and the majority rules for aggregating directors’ votes. Quotas The role and importance of quotas within the IMF structure are ultimately related to the raison d’être of the IMF as the only transnational guardian of the world system of payments. The objectives of the IMF as laid down in Article I of its Articles of Agreement are to (i) promote international monetary cooperation, (ii) facilitate the ‘expansion and balanced growth’ of international trade, and contribute to the promotion and maintenance of high levels of employment and real income, (iii) promote exchange stability, (iv) assist in the establishment of a multilateral system of payments in respect of ‘current transactions’, (v) make the general resources of the Fund temporarily available to them to provide them with an opportunity to correct maladjustments in their balance of payments, and (vi) shorten the duration

428


and lessen the degree of balance of payments disequilibrium. If the IMF is to successfully pursue these objectives, it must have resources and mechanisms to affect both the level of aggregate world liquidity as well as member states’ contributions and access to it. While the mandate of the IMF is well short of that of a central bank of the world economy, it nevertheless serves as an international lender of last resort (ILLR). IMF quotas play a critical role in facilitating this. On the supply side, the aggregate quota of all members directly affects the liquidity position of the Fund. On the demand side, the distribution of quotas among member states helps determine limits to their borrowing from IMF resources. Therefore, it is important to distinguish between the desired level of aggregate quotas and its distribution among members. Presently, quotas are targeted to multiple objectives. These are (i) determining the contributions of members to the Fund’s resources, (ii) setting the members’ maximum permissible access to Fund’s resources under its various schemes, (iii) deciding the shares in any special drawing rights (SDRs) allocation among members and (iv) providing a basis for members’ voting power in the Fund and for the constitution of the IMF’s Executive Board. Each member is allotted 250 basic votes plus one additional vote for each SDR 100,000 of the member’s quota. This relative voting power also determines representation in the IMF Executive Board.4 Furthermore, quotas by providing the resources to the Fund have enabled it to act as a lender to meet temporary balance of payment deficits and, by determining the access limits of members, facilitated their access to funds to close financing gaps. This has facilitated maintenance of a multilateral payment system and exchange rate stability that has contributed to the expansion of world trade.5 In a sense, quotas play a role even more important than that of equity in a corporate enterprise. Yet, as is evident, there is one instrument assigned to at least four different objectives leading to the classic problem of having more targets than instruments.6 Under the present dispensation, each member country is assigned a quota expressed in special drawing rights (SDRs).7 All transactions and operations of the Fund are carried out using SDRs and ‘usable currencies’.8 The Fund obtains ‘usable’ as well as ‘unusable currencies’ as part of the quota subscription it receives from the members. These are converted into SDRs for purposes of accounting. A member’s subscription to the Fund equals its quota—one-quarter of which is paid in reserve assets specified by the Fund (SDRs and usable currencies), and three-quarters in member’s own currency. Quota subscriptions form the foundation of the general resource account (GRA) of the IMF.9 These are called ‘ordinary resources’ and consist of: (i) a reserve tranche, not exceeding 25 per cent of the quota and paid in reserve assets specified by the Fund (SDRs or usable currencies), (ii) a balance of quota subscription, paid by member countries in domestic currency (generally in the form of non-interest-bearing promissory notes denominated in domestic currency), and (iii) undistributed net income originating from the use of


429

resources in the GRA. Borrowed resources of the Fund can take the form of (i) general arrangements to borrow (GAB) and (ii) new arrangements to borrow (NAB).10 In addition, the IMF may borrow. Quota subscriptions provide the financial base of the GRA funds. Quotas define the maximum amount a member country can borrow under various schemes (including standby and extended arrangements) as these are fixed in terms of percentages (common to all countries) of their quotas. Hence, quotas have a bearing on the possible demand of financial resources from the Fund and determine members’ access to Fund resources. As quotas also influence the supply of financial resources to it they help determine the liquidity position of the Fund.11 Quota structures have been determined primarily by measures of the size of the economy (measured in terms of US dollars), measures of openness and variability in export earnings. IMF quota structures are reviewed every five years. So far eleven quota reviews have been carried out. The IMF appointed a committee under the chairmanship of Harvard professor Richard Cooper to go into the rationale for quotas. This committee submitted its report in April 2000. This report presented a thorough evaluation of the quota structures but made few recommendations for reform. Jha and Saggar (2000) have evaluated the quota structures of the IMF and found these to be without much rationale. They have then advanced a series of proposals for quota reform that reward a country’s size, ability to stabilise the economy and extent of involvement in the international economy. Voting structures The voting structure within the IMF revolves around its executive directors (EDs). The US, Japan, Germany, France and the UK get to appoint their own EDs directly. Elected members represent other countries. The split up of votes among the EDs on May 21, 2001 is shown in Table 17.4. The asymmetry in the voting structures is apparent from this table. Whereas the US, UK, France, Germany and Japan get their individual EDs, and two small countries—Austria and Belgium—share one ED, many developing countries have to share ED. It may well be the case that these EDs may not be able to portray the requirements and policies of each of these countries in the best possible way, particularly if there are conflicts within groups. This asymmetry in the voting structure of the IMF gets reflected in the design of IMF programmes. We have commented on some aspects of this in Chapter 11. We now provide some brief comments on possible reform of the IMF. A RUDIMENTARY AGENDA FOR THE REFORM OF THE IMF In Chapter 11 we have discussed in some detail the effects of IMF programmes and conditionalities. As would be expected, a number of studies have taken

430


Table 17.4 IMF executive directors and voting power, May 2001



431

432




433


1

Voting power varies on certain matters pertaining to the General Department with use of the Fund’s resources in that Department. 2 Percentages of total votes 2,166,739 in the General Department and the Special Drawing Rights Department. 3 This total does not include the votes of the Islamic State of Afghanistan, Somalia and the Federal Republic of Yugoslavia, which did not participate in the 2000 Regular Election of Executive Directors. The total votes of these members is 7,073–0.33 per cent of those in the General Department and Special Drawing Rights Department. 4 This total does not include the votes of the Democratic Republic of the Congo which were suspended effective June 2, 1994 pursuant to Article XXVI, Section 2(b) of the Articles of Agreement. 5 This figure may differ from the sum of the percentages shown for individual Directors because of rounding. Source: IMF (2001).

434


up the question of reforming the IMF programmes. We now take up some of these issues. A reform of the IMF programmes such that they are more responsive to the needs of developing countries should begin with a restructuring of the power structure. As has been argued above, the quota structure within the IMF is highly arbitrary. Furthermore, it does not truly reflect the influence of various countries on the world economy since it measures GDP at current nominal exchange rates rather than PPP rates and does not reward good macroeconomic performance. Some suggestions for the revision of IMF quota structures along these lines have been offered in Jha and Saggar (2000). Second, as also has been noted above, voting structures within the IMF are unduly restrictive. Given the 85 per cent majority rule for most decisions, the US has an effective veto more for blocking decisions rather than pushing its own agenda. A majority voting structure for most decisions should be completed. These two reforms would increase the stake and the power of the developing countries. It is likely then that loans made by the IMF would more accurately reflect developing country preferences. Moreover, with greater say in the operation of the IMF, developing countries would have less reason to complain about an important aspect of the IFA. A reform of the actual policy package of the IMF is a more difficult job. The situations in which the IMF is asked to intervene are usually not of its own making. For instance, if an HIPC asks for an IMF loan, the overhang of the debt might make it necessary to prescribe a rather restrictive credit policy if matters were left simply to the IMF. In fact at least one other policy option is suggested. Independent attempts should be made to reduce the debt directly. If such efforts are not made and the IMF is asked to tackle an emergent situation on its own, it would get unfairly blamed for ‘excesses’. Thus some of the complaints against the IMF are not entirely valid and it is taking the blame for faults with the entire IFA. The IMF faces some inherent constraints in the design of its programmes, which leave it open to criticism. One such example is the interest rate that the IMF charges on loans advanced. Several economists have argued that these interest rates are too high and place an undue burden on developing countries. However, there is a deeper problem here. If the interest rates were too low then there would be a problem of moral hazard. Countries would realise that since IMF bailouts would be available at low rates of interest there would be no need to pursue prudent macroeconomic policies. Imprudent macroeconomic policies would become an attractive policy option. In fact almost all proponents of the IMF’s role as ILLR have emphasised that it must charge interest rates higher than the market rates. The argument is that if the country could borrow at the market rate it would go ahead and do so. It has approached the IMF because under the conditions it finds itself in, the country finds market loans hard to get and approaches the IMF as a last


435

resort. The IMF must ensure that the country uses the funds prudently and has all incentives to pursue sound macroeconomic policies. In order to pursue this objective, the IMF would have to charge higher than market interest rates. In support of this policy stance there is the further argument that loans for longer-term purposes are made at concessional rates anyway. It is possible, however, to streamline the conditionalities imposed by the IMF. There is room for the view that the credit restraints imposed by the IMF’s policy package are excessive. These restrictions would appear to be correct only within the purview of the IMF’s excessively monetarist models. The IMF model should be expanded to include the possibility that credit conditions affect not just the demand for output but its supply as well. THE GATT AND WTO The GATT and later the WTO were formed with the express intention of lowering barriers to international trade. Subsequently and particularly since the Uruguay Round (UR) attention has also been directed toward facilitating the global flow of international capital and protection of intellectual property rights through patents and other methods. The Trade and Intellectual Property Rights (TRIPS) agreement is part of this. Efforts are on to widen and deepen the content of such agreements. However, it is common knowledge that even after the full implementation of all UR concessions a substantial number of high tariffs on imports from ADCs will remain (Table 16.2) with peaks reaching 350 to 900 per cent although the majority range from 12 to 30 per cent. One-fifth of the peak tariffs of the US, about 30 per cent of those in the EU and Japan will exceed 30 per cent. In contrast, developing countries apply rates above 12 per cent ad valorem more frequently than DC but have fewer extremely high rates.12 Another area of contention is non-tariff barriers (NTBs). NTBs are pervasive in almost all countries and take many forms from import quotas, licensing of import/export, anti-dumping and countervailing duties, sanctions and voluntary export restraints to preference procurement of domestic goods, customs valuation and clearance procedures, copyrights and intellectual property rights. Particularly in the US and EU, NTBs are large (Table 17.5, especially items 3a to 3d). Thus there is some justification for the view that over the past 40 years international trade policies have been driven by the needs of the industrialised world and its principal corporations. The participation of the developing countries has been sought on the basis of the argument that freer trade by fostering world growth would encourage the development of poor countries. This assumption has now been shown to be faulty. In particular, the WTO regime seems to have underestimated the costs of transition to freer trade for developing countries. Current data reveal that for developing countries as a

436


Table 17.5 Frequency ratio of NTBs by commodity groups, 1993 (Product categories subject to NTBs expressed as a percentage of total number of product categories in corresponding group)

Source: Deardorff and Stern (1998).

whole, terms of trade fell by more than 5 per cent during the 1990s. Further, commodity prices fell by over 16 and 34 per cent respectively for oil and non-oil commodities since 1996. Many manufactures exported by developing countries are now beginning to behave like primary commodities as a growing number of countries simultaneously attempt to raise their exports in the relatively stagnant and protected markets of developed countries. Although LDC exports grew sharply during 2000 some of this could be attributed to regional trade agreements as well as to the software technology boom of the 1990s. However, in the first half of 2001 global trade grew only by 2 per cent compared to 12 per cent the year before. In addition, income inequalities have grown since the UR. The income gap between the OECD and developing nations was about 5.5:1 in 1993, whereas in 1999 it had increased to a little over 7:1. The strong growth in trade in services and knowledge-based goods in the post-UR phase has further marginalised a number of developing countries. The levels of human development are also unequal. Ironically those left behind in terms of HD indicators are those deeply integrated into the world trading system. Sub-Saharan Africa has an export to GDP ratio of 29 per cent, but being mainly dependent on primary products, the gains from trade has not accrued to them. Polarisation between countries has been accompanied by increased inequality within the countries themselves. Increased trade liberalisation has increased the wage inequality between skilled and unskilled workers both within countries and in a north-south sense.


437

Within developing countries this has been accompanied by declining industrial employment of unskilled workers. Capital has gained in comparison to labour and profit shares have risen everywhere.13 As a result of such trends there is much scepticism about the usefulness of the WTO in many developing countries. Concurrently there have been doubts expressed in many developing countries—particularly among labour and environmental groups. Some of these fears found expression in the hostilities that broke out at the so-called Millennium Round of the WTO in Seattle, Washington, USA in 1999. However, there can be no doubt that the greatest disappointment has been for the developing countries. The usefulness of the free trade regime is not being questioned but the route to be taken is in dispute. Future trade rounds, if they are to be successful, must create a regime that offers the third world a real chance to bootstrap itself out of poverty. This would require, among other things, that developing countries must have clear-cut access to agricultural markets as well as markets for labour-intensive manufactures and associated items like leather, textiles, processed items and the like in Western Europe and North America, protected as these are at the moment through a number of implicit and not so implicit subsidies. In addition, efforts must be made to reconcile patent agreements, for example, with the genuine needs of public health in the developing countries. A recent example of the need for this reconciliation was the prolonged legal battle to make available generic forms of drugs to help treat HIV-AIDS patients in South Africa and other HIV-AIDS-ravaged countries. Pursuing a merely legalistic agenda, as the West has often done in the past, will simply not do. The developing countries also need to persist with their programme of tariff reduction but action at the level of the developed countries seems logically prior, at least at this stage. THE WORLD BANK The World Bank has been at the forefront of project and programme lending to the developing countries. This aid has taken several forms—from financial to medical to capacity building and technical support. The Bank has provided crucial aid to many developing countries. However, the Bank’s lending has not kept up with the drop in bilateral aid from OECD governments. As has already been discussed, private capital flows have not been attracted to the poorest countries whereas some of these countries are facing crushing problems of debt overhang. Hence, there is a pressing need for enhancing the extent of financial and other flows to the World Bank. This must depend upon increases in the flow of funds to the Bank. In recent times, multilateral debt (i.e. debt owed by developing countries to the World Bank and International Monetary Fund (IMF)) has grown phenomenally. It appears that the most severely indebted countries of the developing world are trapped on a debt treadmill forced to take new loans

438


to pay old ones or risk default and potential economic collapse. Sub-Saharan African countries, for example, paid more in debt between 1990 and 1993 than they spent on health care and education. Still they watched their debt load double. As southern governments reel under financial pressure to generate foreign currency to pay the rising debt, poverty and environmental devastation increase. This situation is becoming increasingly unsustainable, economically, socially and environmentally. The scale of the problem is quite high now. The total developing country multilateral debt rose from US$61.6 billion in 1980 to US$313 billion in 1994. A significant amount of new IMF and World Bank funds are used to repay outstanding interest on debts owed to commercial banks. For example, from 1983 to 1989, US$32.7 billion in loans from multilateral sources went to service commercial bank debt, representing 17 per cent of total debt service over the period. The various rescheduling plans of the 1980s, have resulted in a situation in which the IMF and the International Bank for Reconstruction and Development (IBRD—the non-concessional wing of the World Bank) are receiving far more in debt servicing from the poorest countries than they are lending. The multilateral debt problem affects the poorest countries with high debts most severely. Debt servicing obligations to multilateral institutions rose from less than US$8.5 billion in 1982 to almost US$40 billion in 1994, an almost five-fold increase. In most cases, debt service to multilateral banks takes precedence over all other debt service requirements. The World Bank’s International Development Association (IDA), which is the arm of the Bank set up to provide concessional loans and grants to the poorest countries, now provides a significant amount of these funds for ‘structural adjustment support’ programmes. The World Bank also provides debt relief through a fund known as the International Development Association’s Debt Reduction Facility and the so-called ‘Fifth Dimension’ Fund. The IMF created an Enhanced Structural Adjustment Facility (ESAF) which channels aid dollars from donor countries to the poorest countries on more concessional terms. However, there is growing concern that most foreign aid to indebted lowincome countries simply goes straight back into the coffers of the World Bank and IMF, because this aid is being used to service outstanding World Bank/IMF debt. AVOIDING CURRENCY CRISES Currency crises can be classified into two types (Dornbusch, 2001).14 Early type currency crises typically began as a real appreciation of the home currency because of domestic overspending at fixed exchange rates. This would then eat into exports and lower the rate of growth. Devaluation would follow. If there is no monetary accommodation following the devaluation, the economy


439

would stabilise. However, typically money supply is passive so that inflation rises necessitating further devaluation. This is then a devaluation-inflation vicious cycle. Dornbusch cites Egypt as an example of this sort of currency crisis. The more recent types of currency crises are more complex in nature. The essential questions involved are the balance sheets of a significant part of the economy and the valuation of the exchange rate. But once it becomes entangled into one question the other factor immediately becomes relevant as a result of capital flight. For example, if the currency is not overvalued but the financial sector has considerable bad debts, capital may start leaving the country, as it would not feel safe with the weakness in the financial system. This results in a sharp rise in the demand for foreign exchange that puts pressure on the currency. The central bank may try to arrest this fall by selling off foreign exchange thus making a big hole in its foreign exchange reserves. Ultimately, this process becomes unsustainable and the currency collapses. This collapse increases debt with drastic implications for already weak balance sheets. So the process continues. Currency crises are extremely expensive for the economy undergoing the crisis. If there is a history of currency crises in a country every subsequent currency crisis becomes even more expensive. Thus once a country has had a crisis it becomes more vulnerable to future crises and once a new crisis occurs finds it harder and more costly to deal with it. The costs of the currency crises arise from three sources (Dornbusch, 2001). The first is a substantial loss in output and growth potential. Second, the currency crises lead to immense public finance problems. In the face of a collapsing economy, tax revenue sources dry up and since the currency crisis raises the domestic resource cost of servicing liabilities, debts start to pile up. This is fuelled by the rise in the interest cost charged on the debt. The burden of servicing this debt persists beyond the crisis and, unless attended to, may contain the seeds of the next crisis. This would typically be the case when the ongoing crisis occurs in a situation where the government does not have the ability to meet the higher debt service burden by taxation or reduction in spending. Third, the pile-up of the debt leads to payments to creditors, and the downturn of economic activity leads to a drop in wages and higher unemployment. This redistribution of income away from labour to capital may be socially resented. Dornbusch (2001) estimates the numerical values of the costs of currency crises to be as high as 20 or even 30 per cent and more of GDP for a bank bailout. In addition, there is easily a 10 or 15 per cent increase in debt from high interest rates applied to a large debt and from recession-induced tax losses. There is also always a large loss of reserves, which are sacrificed in the run-up to a crisis as the central bank tries to defend the currency. Furthermore, a currency crisis lowers the credit rating of the country and makes it harder and more expensive for it to borrow from abroad.

440


Thus periods of currency crisis are associated with abysmal economic growth, a sharp rise in the discount rate as people’s decisions get biased in favour of the short term and deteriorating social and economic climate. Such effects can persist for a long while even after the crisis is over. Thus Dornbusch (2001) estimates that during the debt induced in Latin America during the 1980s GDP growth per capita actually fell by 0.7 per cent. Even by the end of the 1990s the recovery was far from complete with GDP per capita growing at only 1.7 per cent. How would one go about arresting currency crises? This is a rather complex question with which the literature is still trying to come to terms. Surely, it would be important to ensure that the currency is not overvalued for long periods of time. In addition, one has to keep a careful eye on the strength of the financial system and nip potential slippages in the bud. However, even these efforts may not be enough because of the phenomenon of contagion. Even if the currency and the financial system are in reasonable health, there may be problems because the country’s finances are closely aligned with another country whose finances are weak and which may, therefore, face a crisis. Dealing with some of these complex issues at a national level is a problem. Some countries like Malaysia opted for strict capital controls. There is some evidence, as discussed in Chapter 15, that these may have worked. However, this is a rather drastic solution and may not be an unambiguous blessing. What is there to guarantee that behind the veil of capital controls, an overvaluation of the currency is not being perpetuated or that genuine weaknesses in the banking system are not being camouflaged? Controls are rarely efficient or temporary. An alternative to capital controls is following the IMF medicine. At first glance this medicine sounds counterproductive since raising interest rates in a situation of weak balance sheets would make these weaker. Fiscal and monetary austerity at a time of economic slowdown would make the slowdown even worse. In dealing with some of these complex issues, potential reforms of the IFA are important. Failing rapid accomplishment of this, hard choices may have to be made about the exchange rate regime. However, if the interest rates remain unchanged and the central bank keeps issuing money at these interest rates, the potential for capital flight will exist. Within a slightly longer-term framework, the country may have to make a hard choice about its exchange rate regime. It might be compelled to opt for a currency board or dollarisation. However, as explained above, such arrangements may not be ideal for all countries. Thus there are no easy choices for policy at the national level. At the international level, work needs to be done at reorganising the IFA in order to give early warning signals about the possibility of crisis and facilitate and support corrective measures.


441

A POSTSCRIPT ON STATISTICAL STANDARDS Our analysis in this chapter has pointed over and over again the importance of early warning signals in preventing a crisis. However, statistical early warning systems are hard to establish. First, there is widespread variation in world standards of national and financial accounts collection and reporting. The IMF, for instance, has been at the forefront of efforts to institute uniformity in standards of data dissemination across the world. It has instituted Global Data Dissemination Standards (GDDS) and Special Data Dissemination Standards (SDDS). A principal difference between the two is the frequency of data collection and reporting. The SDDS, for example, requires GDP data to be collected and reported quarterly whereas the GDDS requires annual collection and reporting of GDP.15 Uniformity in standards of collection and reporting have been stressed by international financial bodies like the IMF and the Bank of International Settlements (BIS) got involved in the task. Later even the World Bank (WB) has become an important player in this effort. These different agencies have approached this problem from various angles depending upon their immediate areas of concern. The BIS, for instance, lays considerable emphasis on updating its 1991 International Accounting Standards (IAS30).16 The IMF, on the other hand, has viewed this problem within the context17 of developing and implementing its GDDS and SDDS in countries around the world. The WB is interested because, among other reasons, the adoption of statistical standards and practice for banking and finance has implications for conducting and assessing the success of some of the WB’s own programmes.18 There are essentially three steps in the debate on international standards. First, is the question of what constitutes acceptable standards and uniform norms for collecting and reporting financial and related statistics. Second, there is the question of what will determine the pace with which countries around the world adopt these standards. Third, is the important question of the use to which the information so collected would be put. Would this information be made public? Or would only select bodies like the IMF, governments and banking circles be privy to this? What would be the time lag between the collection of this information and its disclosure to the public?19 Countries around the world have reacted cautiously to the proposals. Whereas some developed economies like Australia were early and enthusiastic supporters there have been voices in dissent. To cite just two examples, the Governor of the Reserve Bank of New Zealand in his address to the SEANZA Forum of Banking Supervisors’ Conference in Auckland in November 1999 emphasised the view that ‘(supervisory arrangements)…must be tailored to the individual circumstances of each country’ (BIS Review 124/1999, page 2, available on the BIS website www.bis.org). In a similar vein, the Deputy

442


Governor of the Reserve Bank of India in a speech to the IMF-WB Conference on International Standards and Codes held in March 2001 emphasised the necessity of allowing for differences in institutional development and legislative practices around the world. In particular, he wanted the countries’ adoption of these standards to be voluntary and at their own pace (BIS Review 20/2001, page 1, available on the BIS website). The first point to note is that in view of the original purpose of developing these codes of good practice (of providing early warning), these codes should be seen as an international public good since they would, potentially, benefit the entire international financial community. However, the clients of this public good are not individual economic agents but sovereign governments. Two points need to be remembered here. First, in the absence of a world government, the provision of this public good will necessarily be voluntary. To that extent the fears of some of the countries may be exaggerated. The only viable leverage that international bodies have on the participating countries is to categorise them as performers or non-performers in the area of adopting the international standards, which may then affect these countries’ abilities to borrow. Second, as in a typical problem of voluntary provision of public goods, there is a fear that, from the global point of view, less than the optimal amount of this good will be provided. Hence, there is a genuine tussle between the needs of the global community to have in place uniform international standards that would help in improving the global allocation of capital and the difficulties of various countries— particularly developing countries with weak statistical and/or banking bases to come up to the requirements of these standards. Resolving this tussle is a pressing matter for global governance. The ultimate aim of these standards would be best served if the procedures used to resolve this tussle inspire confidence among the major, if not all, players in global capital markets. It would be myopic to use established voting structures like those in the IMF’s Board of Directors to decide on this important issue. As has been argued above, the IMF’s voting structures are excessively biased in favour of the US. An additional issue of concern that has been raised in recent times is developing prudential norms for banking. Once these norms were put in place they would serve as guideposts for judging whether banking practices were sound. For instance, banks would be required to keep certain minimum capital reserves to support their lending profiles, the so-called capital adequacy norms. These have been classified as the Basel Capital Adequacy Accords. Several authors have, however, raised objections to some of the stipulations as discriminating against developing countries. See, for example, GriffithJones and Spratt (2001). The Basel Capital Adequacy Accord was an important achievement in the area of regulation of international banks. Agreement was reached between the member countries of the Basel Committee that a minimum capital requirement of 8 per cent would be required of internationally active banks.


443

Since that time the Accord has been adopted by more than 100 countries, many of which are developing countries. The Basel norms make a distinction between OECD/non-OECD countries in risk-weightage. Thus, ceteris paribus, a bank with a higher non-OECD component in its portfolio would be judged to be facing higher risk. GriffithJones and Spratt (2001) argue that from a developing country perspective, the OECD/non-OECD distinction in risk-weights is crude, unfair and provides a distorting incentive for developing countries to seek OECD membership. Furthermore, there was a lower risk weight to loans for emerging markets and this may have led to excessive short-term lending to them, thus contributing to the East Asian crisis. The new Basel Accords are to come into effect in 2004. Griffith-Jones and Spratt (2001) argue that these new accords do not entirely eliminate the bias against developing countries. The most important reason for this seems to be that the new accord would unduly exaggerate the risks in developing countries and thus impede the flow of capital to them. APPENDIX ON CAPITAL FLIGHT The rapid integration of capital markets throughout the world has facilitated the ease with which residents of developing countries can transfer their financial assets abroad. Both the dismantling of capital controls as well as the increased efficiency of financial markets have contributed to this. Some people have argued that this represents an efficient portfolio diversification and should, therefore, be welcome. However, recent experience has pointed to the large volume of capital flight from developing countries. Later in this chapter we will discuss the perceived deleterious effects of such capital flight. However, defining capital flight, particularly the type that hurts developing countries, is a non-trivial task. Recent attempts at such definition include Ajayi (1997), Dooley and Kletzer (1994) and Schineller (1993). It would be inappropriate to call all financial transfers abroad by domestic residents as capital flight. The literature typically distinguishes between four different types of private sector acquisition of assets overseas. (1) Direct investment overseas by large domestic companies may, for example, represent their ‘newly industrialising’ status. An example of this is flows from Mexico, Malaysia, Korea and Taiwan financing equity purchases in productive units in the US and the EU. Moreover, companies may be using financial flows abroad in order to access developed country markets. Thus Venezuelan oil firms may be seeking to establish distribution outlets in the US. Furthermore, domestic firms in some countries have been engaged in regional investment (Brazil in Mercosur, India in South Asia) and domestic banks in developing countries often lend regionally. Certainly such transfers are not a cause for worry.

444


(2) When domestic assets are particularly risky, a diversification of portfolios toward foreign assets may be justified. This becomes particularly useful when domestic financial markets in developing countries lack depth or when the need is felt to protect savings from business cycles or exogenous shocks at home. While some of these motives may be justifiable, it can well be argued that ‘illegitimate’ transfers abroad can often be disguised as belonging to this category. (3) An important reason for transfer of financial assets overseas is for the purposes of evading domestic taxation and other charges. Governments in developing countries often find it well-nigh impossible to tax capital stored overseas. Large wealth holders in developing countries are attracted towards overseas assets. Furthermore, the proceeds of patently illegal activities such as the sale of narcotic drugs, commissions on arms deals, etc., are routinely kept in banks and other financial institutions overseas. Transfers of this sort are patently illegal and a cause for worry. These involve not only breaking the law in the developing country but also depriving the developing country of much needed capital. (4) A shift from domestic to foreign assets can occur for other reasons as well. For instance, property rights may not be well defined in the developing country and the functioning of the judicial system may be arbitrary and slow. Furthermore, some wealth holders might dread nationalisation without adequate compensation. Also there may be high inflation in the home economy leading to erosion in the real value of assets whose prices are insufficiently indexed to inflation. Fearing such eventualities, wealth holders may transfer funds overseas. This would again constitute inappropriate transfers. The coexistence of ‘legitimate overseas asset acquisition’ with capital flight may often lead to the latter being disguised as the former. One way of achieving this is to under-invoice exports (thereby understating their value in foreign exchange) and over-invoice imports (thereby exaggerating the foreign currency value of imports). The difference between the two can then be kept as foreign currency deposits abroad so that the accounts presented by exporters and importers to domestic authorities are essentially false. This ambiguity in what genuinely constitutes capital flight makes their measurement difficult. Ajayi (1997) identifies five different methods of measuring capital flight. The simplest measure is that of ‘hot money’ which involves taking the net short-term capital outflow and adding the ‘errors and omissions’ element of the balance of payments. This is the narrowest measure of capital fight. In its International Financial Statistics the IMF publishes data on capital flight measured as a change in cross-border bank deposits of non-banks by country of the depositor. Ajayi considers this an underestimate of capital flight. The World Bank uses a broader measure. It compares the sources of finance, i.e., the change in external debt and net foreign direct investment with the uses of finance, i.e., the current account deficits and the changes in official reserves. Hence this method includes the


445

assets of both the banking and the non-banking sectors. The best-known method of measuring capital flight is due to Dooley (1986, 1988). This method captures that part of an increase in external claims that yields income not reported to domestic authorities. Thus this measure focuses on ‘abnormal’ flows and not declared flows that may actually be the result of legal portfolio diversification. The Dooley and residual methods have a high correlation coefficient. Most statistical methodologies rely on (a) the ‘errors and omissions’ item in the balance of payments. This measures the difference between the recorded (and therefore authorised) transactions and observed change in reserves; and (b) estimates of mis-invoicing based on the difference between official figures and those of major trading partners. Thus the Economist Intelligence Unit (EIU) measures capital flight as capital flight=— current account—change in reserves—net FDI inflows—change in the external debt stock. It should also be noted that not all capital flight is lost to the country on a permanent basis. In the case of countries like Russia and China, for example, there have been reversals of capital flights. The intuition behind this is that the domestic investor can get the benefit of both the tax exemption and the protected status of a foreign investor. In this context, some authors have advanced a ‘revolving door’ view on capital flight. This refers to the contemporaneous causal relation between external debt and capital flight. For instance, a Morgan Guaranty Trust Company study published in 1986 estimated that for the time period 1976– 85, LDCs experienced net flight of nearly $200 billion, while simultaneously accumulating $450 billion additional foreign debt. This contradicts the standard view that a favourable investment climate in any country would not only attract foreign capital, but also retain domestic investment. Boyce (1992) and Chipalkatti and Rishi (2001) for India provide useful analyses for the Philippine and Indian experiences in this connection. Foreign capital enters the country as proceeds of foreign borrowings and then slips out of the country as capital flight. This is referred to as a financial revolving door. Thus capital outflows and external indebtedness are causally linked to each other.20 Many factors have been cited as reasons for capital flight. Differences in international tax regimes play an important role in determining capital flight. The US, for example, taxes the flow of income from the assets of non-residents, i.e., dividend and interest income but not capital gains. Furthermore, the US does not provide developing countries with information on such capital gains. Most developing countries lack the administrative capacity to collect information on such capital gains with the result that these gains remain untaxed. Capital flight—especially of the kind that is not associated with increased efficiency of allocation—robs developing countries of scarce resources and creates an atmosphere where the authority of economic decision making by the government is undermined.

446


Empirical studies show that capital flight is a substantial impediment to economic growth in developing countries (Varman-Schendier, 1991). For example, following the economic crisis and political instability in Mexico in 1983, capital outflows were 8.7 per cent of GDP partly as a consequence of which per capita GDP fell by 5.8 per cent. The following year capital flight fell to 3.1 per cent of GDP and per capita GDP grew at 6.1 per cent. How serious is the problem of capital flight? Surely, this problem takes very serious proportions when there is a sudden drop in the confidence of a developing country’s currency. This was essentially the case during the currency crises in Mexico and the East Asian countries. In the case of Russia, for example, the continued flight of capital poses one of the most serious problems facing the country. Estimates by the US Treasury Department, the IMF and other sources suggest that a net of well over $100 billion to $200 billion have left Russia—legally or illegally—during the past decade. In 1996 alone capital flight amounted to nearly $39 billion, for 1998 the figure was $25 billion and for 1999 it was $15 billion. In March 2000 according to a Russian government estimate, capital flight was running at $3 billion a month. Much of this money has ended up in western banks and securities markets and some toward the purchase of luxurious real estate in the US, Western Europe and tropical resorts. The resulting shortage of capital has choked economic reforms in Russia. Tax revenues have dried up and the resources for private and public investment are running low. On the other hand, the outflow of capital has led to the growth of criminal gangs—euphemistically called oligarchies—who coordinate large-scale transfers of money for generous commissions. The Russian situation is a bit extreme as the country is going through a difficult transition. However, even under ordinary circumstances this problem is quite serious. This is depicted in Tables A17.1 and A17.2. Table A17.1 Outward FDI stocks of developing countries as a percentage of GDP

* Indicates data for 1995 since data for 1997 are not available. Source: UNCTAD (1999).


447

Table A17.2 Estimates of misinvoicing for selected countries, 1998 ($m)

Source: Schneider (2000).

Some episodes of capital flight Major episodes of capital flight have typically been associated with political events. In Argentina episodes of political instability in the 1980s were associated with surges in capital flight. In 1981 General Jorge Rafael Videla became President but was ousted in December by Field Marshal Roberto Viola. In 1985 a state of emergency was declared to halt right-wing violence. In 1987 there was a military revolt successfully quelled by President Raul Alfonsin. Through all this, inflation was very high and the economy stagnated. In May 1989 President Carlos Menem imposed an austerity programme that sparked considerable unrest. Between 1976 and 1991 capital flight from Argentina averaged 3.4 per cent of GDP and peaked at 10 per cent in 1989. In Mexico the economic crisis in the mid-1980s challenged the authority of the ruling party. In the midst of the economic crisis in 1985 election fraud led to a massive protest by the opposition. Political reforms had to wait until 1988. Uprisings in Chiapas and Guerero demanding greater democratisation and lower inequality began in the early 1990s. Mexican capital flight averaged 3.11 per cent of GDP annually between 1976 and 1991, reaching a maximum of 8.7 per cent in 1983. In South Korea the mid-1980s were a period of greater turmoil leading to a transition to democracy in 1988. The political turmoil continued well into the 1990s. South Korea’s average capital flight averaged 1.92 per cent of GDP per year during 1976–91 but peaked to 6.6 per cent of GDP in 1987 just prior to the transition to democracy. Many sub-Saharan African countries are caught in what Londregan and Poole (1990) have called a ‘coup trap’. A typical case is Nigeria. Under authoritarian military regimes, organised labour and politicians were harassed. In 1985 General Ibrahim Babangida came to power through an internal military coup. In 1987 an attempted coup failed and the leaders of the coup were executed. There was another abortive coup attempt in 1991. Capital flight in Nigeria between 1976 and 1991 averaged 7.37 per cent of GDP and exceeded 31 per cent in 1987.

448


The unpopular Marcos regime in the Philippines faced upheavals in 1985. In 1986 Marcos called an election and was defeated by Corazan Aquino. Following this defeat, supporters of Marcos clashed with those of Aquino. For about two years there was considerable political uncertainty. From 1976– 91 capital flight from the Philippines averaged about 2.56 per cent a year but exceeded 8 per cent in 1985 prior to the presidential election. In the 1980s and 1990s, Côte d’Ivoire faced harsh political and economic conditions. Falling real wages led to increasing worker protests in 1986. In 1990 there was a threat of a coup as junior army officers and army conscripts seized the Abidjan airport. This led to substantial capital flight. From 1976– 91 average capital flight from Côte d’Ivoire was 2.98 per cent of GDP but it reached a peak of 16 per cent of GDP in 1985 and 1986. Preventing capital flight For some developing countries, then, capital flight seems to be a serious problem. All countries—developed as well as developing—have tried different measures to arrest capital flight. The option of advancing the economic reforms programme is easy to advance but begs the question of how economic reforms can take place in a country with high capital flight indicating loss of confidence in the economy. This then becomes circular reasoning. One method that has often been discussed is capital controls—which we discuss below. Capital controls Capital controls are an oft-used method to arrest capital flight. Article VI, Section 3 of the IMF permits capital controls. Article VIII of the IMF only forbids restrictions on current transactions—such as international trade in goods and services and the remittances of dividends, interest income and profits. Both developed and developing countries have used capital controls extensively at certain times. As would be expected developing countries have used this much more frequently. Such capital controls include regulation of direct investment and real estate transactions. Also included are credit operations and provisions specific to commercial banks as well as many types of portfolio investment including shares or other securities of a participating nature, bonds or other debt securities, money market instruments, collective investment securities and derivatives and other instruments. All such instruments involve inflows (purchase locally by non-residents or sale or issue abroad by residents) and outflows (sale or issue by non-residents, or purchase abroad by residents). In addition, institutional investors may be restricted from investing abroad. A considerable amount of attention is focused on restricting capital inflows in order to prevent the occurrence of speculative bubbles. Such speculative bubbles are usually ascribed to the development of significant gaps in the short-term return to capital in developed and developing countries. A sudden


449

reduction of this gap leads to a reversal of capital flows, which has devastating effects on stocks and currency values (Eichengreen and Mussa, 1998). Eichengreen and Mussa (1998) also note that capital controls may have a distortionary effect in that they may be used as a pretext to slacken macroeconomic discipline. Thus excessively expansionary monetary policy can be pursued under the assumption that the presence of capital controls will restrict pressures on the home currency. However, when domestic capital markets are imperfect and systems for financial supervision are not robust, there is a strong case for not liberalising capital account. Implementation issues Controls have typically involved limitations on (i) external asset and liability positions of domestic financial institutions, particularly banks; (ii) domestic operations of foreign financial institutions; (iii) external positions; and (iv) real estate holdings or direct investment of nonbank residents. The method of control typically involves restrictions on the types of instruments to be used, e.g., restricting the ability of domestic borrowers to issue bonds on overseas markets. Such quantitative restrictions have the disadvantage that their application will be uncertain and involve an unknown element of investor risk. Price-based controls such as taxes on overseas assets, on the other hand, are preferable to quantitative controls because these can easily be reflected in the investors’ risk-return calculations. Furthermore, they are credible insofar as they are backed by appropriate legislation. However, price-based controls act as only a weak brake during episodes of sudden surges in capital outflows. This, then, is a disadvantage of the price-based approaches. However, quantity-based approaches become porous over time as investors find ways of getting around these. For example, if capital controls are particularly effective, capital flight can take place through the traditional route of mis-invoicing of exports and imports. This route is harder to regulate because of existing international trade commitments. All said and done, however, recent experiences with capital controls in Chile and Malaysia have been encouraging, at least in the short run.

How the Bretton Woods Institutions were formed The International Monetary Fund (IMF) and the International Bank for Reconstruction and Development (or the World Bank or IBRD), are known as the Bretton Woods Institutions (BWIs). They were formed in Bretton Woods, New Hampshire in 1944 on the eve of the end of World War II. They were precursors to the United Nations and other

450


multilateral institutions formed after World War II and reflected the new spirit of cooperation between nations, especially in economic matters. At the founding of the IMF and the IBRD the founders were not directly concerned with ‘development’ as we know it today but with avoiding behaviour among nations that led to economic catastrophe and war and to prepare for the economic problems they believed would confront them after the end of World War II. Those who attended the Bretton Woods Conference in New Hampshire wanted to establish a monetary system that would prevent the repetition of the chaos during the inter-war period (1918–39), which had been marked by high inflation, restrictions on international trade and payments, speculation in the foreign exchange market, sharp movements in central banks’ foreign reserves, wildly fluctuating exchange rate movements, gold shortages and sharp recessions. The founders of the IMF agreed upon a ‘gold exchange standard’. The arrangement agreed upon at Bretton Woods was related to the gold standard period between 1870 and 1914. That international monetary system required that each unit of a country’s currency represent a certain weight of gold—i.e., central banks were required to keep an ‘official parity’ between the country’s currency and gold. Maintaining the official parity required central banks to keep adequate stocks of gold as ‘reserves’. The objective of the goldstandard system was to encourage countries to maintain sound economic policies. If they failed to do so, gold would flow from one country to another, which could cause deflation in the country losing gold or inflation in the country receiving the gold. For example, if a country was importing more than it was exporting—i.e., running a current account deficit—but not receiving sufficient loans from abroad to pay for all of the imports (the proceeds from exports help pay for some of the imports), the central bank would start losing gold reserves. This was supposed to lead to a decrease in prices and ultimately result in a reduction of imports and an increase in exports, which, in turn, would reduce the current account deficit. The country experiencing the corresponding current account surplus (for every country with a current account deficit there is a country with a matching current account surplus) would experience the opposite effect. In theory, the gold-standard system was supposed to operate automatically, requiring little government cooperation between countries or their central banks. In fact, the ‘rules of the game’ did not work as smoothly as expected (for example, prices did not drop easily) and countries frequently violated the rules because they did not want to risk unemployment or inflation. During the inter-war period, it was difficult for the European countries to return to the pre-1914 economic world because of the war debts they owed to each other and to the United States. The Allies had depleted their reserves during the war due to imports of food and


451

wartime supplies. After the war, reserves were further depleted in order to service wartime debts. Thus European nations could not maintain the gold standard after the end of World War I. Although several countries returned to the gold standard in the mid-1920s, the system ultimately fell apart permanently in 1931. During the same period, commodity prices had soared due to speculation—but these prices could not be maintained indefinitely so that eventually commodity prices plummeted and the US stock market crashed in 1929. The ensuing Great Depression intensified the demand for gold. Since commodity prices had sharply declined, export income had also declined. In order to protect reserves, countries began to hoard gold and reserve currencies such as the dollar. To make matters worse, country after country began to devalue its currency in order to make exports cheaper to foreign buyers. However, there was really no advantage to these competitive devaluations. All of this led to a severe drop in global economic activity and, in Germany, contributed to the rise of Hitler. The founders of the Bretton Woods Institutions were worried that after World War II the events of the inter-war period would repeat themselves. They sought to prevent this reoccurrence by setting up a system that would provide monetary order as well as facilitate investment in economies where little private capital was expected to flow, namely the European economies. They also hoped economic prosperity starting with reconstruction of war-ravaged Europe would promote peaceful relations. Given the inter-war experience, economic policy makers of the time viewed floating exchange rates—that is, rates mandated by market forces—as destabilising. They believed floating exchange rates produced speculation and large shifts in the value of the exchange rate. They also wanted to avoid competitive devaluations and the ensuing trade restrictions, which had brought trade to a virtual standstill. The gold standard provided the stability policy makers wanted to pursue at the end of World War II, but the standard’s negative aspects dissuaded the BWIs’ founders from adopting it. One of the principal founders, Lord John Maynard Keynes of Britain, was especially critical of the standard. Thus, the founders favoured a fixed exchange rate system tied to gold that would promote important domestic goals such as full employment and price stability. They also wanted a system that would allow countries to maintain sound external economic relations (healthy and balanced international trade and foreign investment-balance of payments equilibrium) without having to impose trade restrictions to, say, reduce imports. Most of the discussions related to figuring out how the IMF would operate, a subject that required much debate and analysis of technical issues relating to international monetary affairs. For example, although both the US and the UK agreed that some sort of mechanism would have to be created to supply countries experiencing

452


balance-of-payments deficits with temporary loans (liquidity), the two countries had serious disagreements over how resources would be made available. Keynes, representing a deficit country (the UK), wanted to create a system that would provide deficit countries with plenty of credit upon request. He therefore proposed an International Clearing Union (ICU) that would issue a new form of international money called ‘bancor’ and monitor lending from one country to another. Keynes wanted the ICU to issue about $26 billion worth of the new money. The United States, represented by Harry Dexter White, objected to Keynes’ proposal because it would amount to a huge loan from the United States to the rest of the world. In contrast to the Keynes plan, the White plan reflected the agenda of creditor nations, or more specifically the only expected creditor nation, the US. The US had the highest level of gold reserves, its infrastructure was not damaged by the war, and neither was its economy. So it was probably going to be the main contributor to the liquidity mechanism. The US wanted to avoid putting too many dollars into foreign hands. Consequently, it was important for White to limit the extent to which the United States would be liable for financing the post-war adjustments of other countries. The United States also wanted the international institution to have the power to require deficit countries to adopt policies that would restore balance of payments equilibrium. These concerns were reflected in White’s proposal for an international organisation called the Stabilisation Fund (SF). To address the liability concern, the SF plan, unlike the ICU proposal, required members to contribute their own currencies and gold to the Fund. Rather than borrowing directly from other countries (the ICU plan), deficit countries would have to obtain the currencies they needed from a fund established and operated by the SF. The United States’ other main concern—requiring countries to adopt sound economic policies—was reflected in a rule that required countries to comply with conditions set forth by the SF in exchange for accessing the pool of currencies. Each member country would contribute to the pool based on a formula reflecting their economic strength. The US contribution would be limited to just under $3.5 billion, far less than its liability under Keynes’ proposal. It was clear that the IMF would not be created without the approval of the US and the UK. The US negotiators knew the US Congress would not vote in favour of the ICU plan because elements of the plan were so unfamiliar to them. The British Parliament and public were not prepared to accept a plan based on a fixed exchange rate system that would completely deprive the nation of its sovereign right to control its monetary policy in order to protect its domestic economy. A series of compromises led to the creation of an international organisation, the International Monetary Fund, which closely resembled the White plan. Given the UK’s concerns, the


453

exchange rate system allowed countries to make a 10 per cent modification of the exchange rate. A larger change required IMF approval, although if the change was to correct a ‘fundamental disequilibrium’, then the IMF had to approve. Preparatory work for the World Bank reflected only a peripheral concern for development. Today the World Bank is at the centre of many development-related activities. But at the Bretton Woods Conference, creating the World Bank was an after-thought, and even when the participants did focus on it, ‘development’ was not their primary concern. At the Bretton Woods Conference the main point of contention was whether development or reconstruction would be the World Bank’s priority. The European nations were particularly interested in World Bank assistance for post-war reconstruction. The Soviet delegation emphasised, for example, that the main purpose of the World Bank was to assist in the reconstruction of infrastructure and the revitalisation of European economies destroyed by the war. The delegations from developing countries were more interested in the formation of the World Bank than the IMF. Mexico proposed language for the World Bank’s charter that would make development a priority. The proposal, not surprisingly, failed. However, developing nations did succeed in introducing the language found in Article III, Section 1(a), which states the Bank’s facilities would be used equitably.


2

The website http://www.imf.org/external/np/tre/quota/2001/eng/aqfc.pdf provides an overview of the IMF view on quotas. Attempt a critical evaluation of this document from the perspective of developing countries. In particular evaluate the report of the committee to oversee quota revaluations headed by Professor Richard Cooper of Harvard University. One of the areas in which the IMF has been evaluated is its response to financial crises. Supporters of the IMF contend that the IMF is not only relevant but more necessary than ever today, given the speed with which an economic crisis can develop and spread. They point to the role that the IMF played during the debt crisis of the 1980s, the Mexican peso crisis in 1994–95, and the Asian financial crisis. Specifically, the IMF helped arrange credit for the beleaguered nations, aided in negotiations with outside creditors, and assisted with structural adjustments to prevent further problems. In some of these ways the IMF was serving in the international economy one of the principal functions of a central bank in a country—that of a lender of last resort. According to some observers, there is no other multinational institution, financial or otherwise, that

454


could have performed this function. The IMF argues that member countries must endure the short-term pain of adjustment in order to enjoy long-term gain. Many critics have claimed that the IMF has placed most of the burdens of market-based adjustment on vulnerable groups in member countries. During the widespread structural adjustment programmes of the 1980s, the IMF frequently required member countries to reduce government spending. Governments responded by cutting back on programmes that the poor, women, children, and other vulnerable groups heavily relied on, e.g., food subsidies, health care and education. If you could rewrite the constitutions of the Bretton Woods Institutions, how would you ensure that their programmes do not end up hurting the poor in developing countries?

NOTES 1 Some countries have two-sided capital controls, i.e., foreign capital is neither allowed in nor is domestic capital allowed out. Some other countries have onesided capital controls with foreign capital being allowed in but domestic capital not allowed out. 2 The development finance requirements of the war-ravaged economies of Western Europe were covered largely by the Marshall Plan. The WTO is a direct descendant of the GATT. 3 They also experiment with the three components of capital flows, viz., FDI/ GDP, portfolio investment/GDP, and loans/GDP. 4 The difficulty in furthering this objective arises from the iniquitous distribution that political power to developed countries (with near veto power to the US in the Executive Board) vis-à-vis developing countries. Various quota reviews have not been able to correct this distortion and some have, instead, augmented the bias against non-oil developing countries. 5 The downside is a moral hazard problem: IMF bailouts provide some kind of insurance against possible payments defaults. In several cases, this may encourage imprudent macroeconomic policies by members. In recent times, this has encouraged the private sector to take cross-border investment positions way beyond the level that could be justified by pure risk-return calculations. This constitutes serious moral hazard. Therefore, we need to know the perceived liquidity requirement of the international economy as it grows and is subjected to real as well as financial shocks, and determine members’ quota contributions and access. 6 Economists have long emphasised the need for having as many independent instruments as targets; see Blinder (1997), Theil (1961) and Tinbergen (1952). 7 Initially the quotas were denominated in US dollars. In March 1992 they were converted to SDRs in a ratio of 1:1 with US dollars the original parity at which the SDRs were created in 1969. 8 Usable currencies are defined as currencies of member countries with ‘sufficiently strong’ balance of payments and gross external reserves position. The IMF maintains a confidential list of usable currencies. 9 The GRA is the principal account in the General Department of the Fund and is used for providing the balance of payments support to member countries.


10

11 12 13 14

15 16 17 18

19

20

455

Therefore, apart from the quota subscriptions, purchases and repurchases are also booked in this account. Eleven-member SDR 17 billion ($24 billion) GAB had seldom been invoked until very recently when calls amounting to about SDR 6.3 billion or 37 per cent of their total credit amount were made from the GAB members to finance the extended arrangements to Russia. Prior to that, GAB was invoked once to finance SDR 2.9 billion purchases by the UK and Italy under stand-by arrangements in 1977 and then to finance SDR 777 million single reserve tranche purchase by the United States in 1978. Following the Mexican crises the initiatives taken at the Halifax summit have seen a new agreement in the form of NAB, under which credit lines are available from developed and emerging markets to meet potential financial crises with a strong contagion effect. The 25-member NAB has increased the potential borrowing capacity of the Fund to nearly SDR 34 billion ($48 billion). But NAB could not be activated during the Asian currency crises, as the entry-into-force clause could not be invoked pending US Congressional approval. At times the IMF has made SDR allocations to member countries to provide additional liquidity. These allocations have been made linearly as per the quota shares of members. Focusing exclusively on MFN tariff rates is misleading since import duties are lower once account is taken of the preferences received by LDCs via GSP, CBI, Lomé Convention and other schemes. A comprehensive database on and analysis of trends in global and within-country inequality can be found on the website of the World Institute for Development Economics Research, www.wider.unu.edu. This classification corresponds roughly to what have been called first generation and second generation crisis models by Obstfeldt. First generation models can be typified by Agenor and Flood (1994) whereas second generation models are of the sort studied by Obsfeldt (1996). Third generation models allow considerable more role for contagion effects; see Eichengreen and Rose (2001). For detailed accounts of the GDDS and SDDS and progress in attending these, see the IMF website www.imf.org. See ‘Report to G7 Finance Ministers and Central Bank Governors on International Accounting Standards’ Basel Committee on Banking Supervision, Basel, April 2000. This document is available on the BIS website www.bis.org. See ‘International Standards and Fund Surveillance—Progress and Issues’, IMF, August 16, 1999. This document is available on the IMF website www.imf.org. It is interesting to note that while welcoming the WB’s involvement in this issue, the IMF is concerned about issues such as jurisdiction and delineation of responsibilities in this regard between the sister Bretton Woods Institutions. See ‘International Standards and Fund Surveillance—Progress and Issues’. This question becomes important in view of the fact that the IMF had issued some warnings on Thai banking problems to the government of Thailand a few months before the baht crisis. Some argue that public disclosure would have advanced the crisis. Whether the crisis would have been as severe with early disclosure as it actually was is a debatable issue. In a very recent paper Kant (2002) reinforces the argument that there should be a distinction between capital outflow and capital flight. His preferred definition of capital flight consists of two steps. In the first step, he regresses (an appropriately defined) measure of capital outflow upon macroeconomic variables such as interest rate differentials between the developing country in question and the US, differences in the GDP growth rates between the developing country and the exports. He argues that these are structural factors affecting capital outflows.

456


Capital flight is associated primarily with an increase in risk and uncertainty associated with the country in question and could partially be approximated as the difference between the actual capital outflow and that predicted by the above regression equation. To this must be added the term ‘errors and omissions’ as mentioned in the balance of payments accounts. Using this measure Kant computes capital flight for South Korea to be of the staggering magnitude of US$43 billion over the period 1980–99.

18

A final word

This is a book about the macroeconomic problems of developing countries. It is clear that there are alternative schools of thought on the determination of national income and employment, and the policy implications of these approaches are vastly different. We also learnt that the schools of thought that are alternative to the neoclassical school have gained considerable respectability. Both their theoretical underpinnings and empirical foundations have evolved and matured over the years. Macro problems of developing countries are much more pressing than those of developed countries. When a country with widespread poverty and few, if any, social safety nets for the poor and unemployed, is asked to slash public expenditures, the most vulnerable sections of society get affected. Thus Jha (2000) presents evidence that the initiation of the economic reform programme in India in 1991 was followed, in 1992, by a rise in the (already high) levels of rural poverty. The subsequent higher economic growth in the mid-1990s led to only a modest fall in poverty. This was a policy switch that was well thought out and debated yet there was a cost. Policy switches in other developing countries have involved even greater cost. Hence, it is imperative that policy errors be avoided. With lack of agreement on the basic model of macroeconomics for developing countries, this has been difficult. At the same time a concern for stabilisation in the standard sense of the term must necessarily be tempered with one for quick and equitable growth of economies that are economically very deprived. To quote Arida and Taylor (1989): The short and long run(s) have always avoided integration. The dilemma is most poignant in the Third World, where there is no agreement about an adequate vision of long-run circular flow and short-run stabilization problems led through orthodox policy into contractionary solutions. (Arida and Taylor, 1989:880) In this book we have highlighted the differences between the mainstream and alternative approaches. However, we have also tried to find common

458


ground, especially with regard to important policy questions. With regard to the problem of the international indebtedness of developing countries, for instance, there is almost complete agreement that debt relief for the most severely indebted countries is required. Similarly, there is some agreement on the need for financial liberalisation to stimulate economic growth in developing countries. In key areas, however, there is deep disagreement. Most variants of the alternative model argue that devaluation in developing countries leads to stagflation, whereas mainstream theory would have us believe that such devaluation is needed to bring LDC currency values more in line with their ‘true’ international worth. Some variants of the alternative model argue that credit and demand restraint in developing countries lead to shortages of working capital and, therefore, stifle industrial growth, whereas mainstream economists argue that such restraint is necessary to curb inflation and promote growth. Some supply-oriented versions of the alternative model agreed with some conclusions of the mainstream model—particularly with regard to the effect of demand restraint. Structuralists argue that the policy prescriptions of mainstream economics not only lead to stagflation in the short run but also dampen growth prospects in the long run apart from causing considerable suffering to the most vulnerable sections of LDC societies—the working poor. Empirical evidence seems to suggest that the IMF stabilisation packages have tended to reduce inflation and balance of payments difficulties at the cost of some sacrifice of the rate of economic growth in the short run. We have also discussed models that have tried to form an eclectic view and combine elements of mainstream thinking and the alternative model—in particular, structuralism. It may be the case that short-run stabilisation poses a dilemma for most developing countries. Credit restraint and devaluation by restricting demand may lead to a drop in the industrial growth rate with inflation adjusting only slowly and with considerable lag. The resulting stagflation may cause considerable hardship in economies where most people are poor and there are hardly any social insurance or welfare schemes. Reforms of output and factor prices (by reducing taxes and reducing subsidies) may be necessary to improve efficiency of supply but may cause considerable industrial restructuring. Nevertheless, there can be little doubt that such adjustments are necessary to ensure balanced and steady growth in the long run. Such are the choices that exist for developing countries. An immediate task of meaningful policy research in this area is to strike a balance between the requirements of short-term stabilisation and long-term rapid and equitable growth. This difference of opinion gets reflected in a number of ways. For instance, there are sharp differences between the IMF approach and the structuralist view of stabilisation policies. Indeed there are sharp differences between the neoclassical and structuralist views on the international economic

A final word

459

order—particularly the financial architecture. In the negotiations for a new WTO Round, for instance, there were sharp differences between the developed and developing countries in the Doha meeting in November 2001. Furthering of freer trade and investment is clearly an overriding economic concern in the wake of the continuing climate of uncertainty and impending global recession. This is true of many bilateral exchanges as well as talks within the institutional structure of the meetings in Shanghai and Doha. As the US trade representative at Doha repeatedly emphasised, the successful completion of a new WTO Round would send ‘a signal to the market, a sign that growth is on the way, that expansion is just around the corner’. After the immense shaking of confidence on September 11, 2001, the desire for such a sign is pressing, even desperate. From the OECD point of view it is more pressing than any possible problems with current WTO rules, problems mostly raised by many poor countries including India, who object to a system that has pushed them to drop their implicit and explicit trade barriers while rich countries largely kept theirs up. Surely the strong support that India obtained for its position at Doha would surprise only those habitually inclined to disregard or discount the developing country perspective. Ever since issues of international economic order started becoming discussed at the institutional level there has been a softening of the developing country position. For instance, in the 1970s the Brandt Commission ‘strongly demanded’ major concessions of a redistributive nature from the developed countries in favour of developing countries, apart from sweeping trade concessions. In 2001 individual developing countries were persuaded and/or coerced to agree to a new round even when major promises of the Uruguay Round remain unfulfilled. To be sure, these promises relate not to concessions of a redistributive nature but to tariffs, agricultural subsidies and dumping, and the patenting of life forms. The most contentious issue is drug patents. India, Brazil, Thailand and a coalition of African countries want clear language stating that patents can be overridden to protect public health. Developed countries, particularly the US and Canada, resisted hard. Their position became untenable after Bayer was successfully muscled into giving price discounts for Cipros in response to a perceived anthrax crisis using exactly the kind of pressure tactics that had been labelled unfair trade practices when it came to sub-Saharan African countries asking for discounts on HIVAIDS drugs. Substantial agreement on diluting the impact of drug patents was reached at Doha. The major potentially destabilising factors are lack of concessions by the US on textiles and EU insistence on competition, environment and other aspects of tied trade in return for significant easing of their position in key areas including agriculture. Thus signals emanating from Doha indicate imminent agreement for a new WTO Round. The most important reason for this is that despite disagreements there is a basic shared premise by all concerned that freer trade is good for all. A corollary believed by developing countries is that all

460


tied trade is bad. Are these valid assumptions? I consider some important exceptions and emphasise the importance of innovation in the agenda for the new Round. In the 1950s Hong Kong and Taiwan and, a little later, South Korea adopted an outward-oriented strategy and were well rewarded. In the aftermath of the collapse of the central planning model of development in the 1980s there was a rush to the outward-orientation strategy by more than 100 countries leading to sharp drops in developing country tariffs in the 1980s and 1990s—the most significant wave of global liberalisation. Most such reductions were unilateral and not negotiated multilaterally. But this strategy is now being questioned. If more than 100 developing countries suddenly switch to outward-looking strategies, is there room in the OECD countries for rising imports from each of these countries? Or is it likely that at some level there is a zero-sum game with only some developing countries increasing their export shares to the OECD economies, whereas the rest lose out? Will world trade growth forever outpace world GDP growth? Also many of the poorest (commodity exporting) countries, especially in the sub-Saharan region, have export revenues equal to only a small fraction of import costs, with the balance financed by aid flows. Falling commodity prices have brought to these countries many problems including civil unrest, rising poverty and HIV-AIDS epidemics as well as sharp declines in their shares in world trade—in some cases, absolute declines in trade in value terms. To these countries the ‘trade as an engine of growth’ strategies of Korea and Taiwan would hold little relevance. Some authors argue that these countries should diversify trade. But this is easier said than done. It is not a simple matter to establish a stable proinvestment policy regime and even if one is established, chances for its sustenance and success are limited. FDI to developing countries has been highly concentrated almost exclusively on Asia (South-East Asia before the crisis of 1997 and China). Thus, an alternative strategy emphasising rapid expansion of South-South trade may be more suitable for the poorest countries. The second assumption may not always be valid either. From the point of view of developing countries tied trade could have its pitfalls—say, labour standards. However, there could be new opportunities. Consider the broadening of intellectual property rights regulations to cover other forms of property rights, including environmental rights such as those to emit carbon into the upper atmosphere. If there could be an agreement that developed countries had largely used up their emission rights cumulatively over their period of industrialisation, they would have to buy polluting rights from developing countries. This could result in significant new financial flows to developing countries. Particularly benefited would be countries with large forest cover. There could be other such areas. Thus imaginative articulation

A final word

461

of the agenda is central to developing countries’ success in the new WTO Round. The area of international finance is another major area of concern for developing countries. Surely there is need for a review of institutions like the International Monetary Fund. As this book has emphasised, there is need to examine the structure of country rights and voting patterns within the IMF. Also perhaps a review of some of its stabilisation policies is warranted. The weight of their accumulated debt is stunting growth prospects in many developing countries. Large fractions of their foreign exchange earnings remain earmarked for debt servicing. The international community needs to design an appropriate policy response to the problems of the HIPC countries. The economies of many LDCs are in crisis today. In fact, such a state of affairs has prevailed for a long time, although there appears to be considerable scope for long-term growth if home and transnational policies are better coordinated. In this book we have tried to understand the various policy dilemmas faced by less developed countries. However, an important point is that several of these policy measures are unlikely to succeed unless these are supported by a suitable international arrangement. This book has also highlighted some of the deficiencies in international arrangements that need to be addressed in order to facilitate sustained economic development in LDCs. Surely it is in the collective long-term self-interest of developed countries as well to pursue this route because enhanced prosperity in developing countries would be beneficial to developed countries as well. However, this may not be in the short-term self-interest of individual developed countries. It remains to be seen whether the international community can grasp the initiative and set the ball rolling on these important institutional reforms.

Bibliography

Agenor, P. and Flood, R. (1994) ‘Macroeconomic Policy, Speculative Attacks and Balance of Payments Crises’, in F.van der Ploeg (ed.) The Handbook of International Macroeconomics, Oxford: Basil Blackwell. Agenor, P. and Montiel, P. (1996) Development Macroeconomics, Princeton, NJ: Princeton University Press. Aghion, P. and Howitt, P. (1992) ‘A Model of Growth through Creative Destruction’, Econometrica 60:323–51. Agosin, M. (1994) ‘Saving and Investment in Latin America’, UNCTAD Discussion Papers, no. 90. Aguilar, R. (2000) ‘Angola’s Incomplete Transition’, in T.Addison (ed.) From Conflict to Reconstruction in Africa, Helsinki, UNU/WIDER. Ahimud, Y. (1981) ‘Price-Level Uncertainty, Indexation and Employment’, Southern Economic Journal 47 (3):776–87. Aiyagari, S.R. (1991) ‘Response to a Defense of Zero Inflation’, Federal Reserve Bank of Minneapolis Quarterly Review (Spring): 21–4. Ajayi, I. (1997) ‘An Analysis of External Debt and Capital Flight in the Severely Indebted Low Income Countries in Sub-Saharan Africa’, IMF Working Paper WP/ 97/68. Alesina, A. and Rodrik, D. (1994) ‘Distributive Politics and Economic Growth’, Quarterly Journal of Economics 109 (2):465–90. Arida, P. and Taylor, L. (1989) ‘Short-Run Macroeconomics’, in H.Chenery and T.N.Srinivasan (eds), Handbook of Development Economics, Amsterdam: North Holland. Aronsson, T., Johansson, P. and Lofgren, K. (1997) ‘Welfare Measurement, Sustainability and Green National Accounting: A Growth Theoretical Approach’, in New Horizons in Environmental Economics Series, Cheltenham, UK: Elgar. Arrow, K. (1964) ‘The Role of Securities in the Optimal Allocation of Risk Bearing’, Review of Economic Studies 91–6. Backus, D. and Driffill, J. (1985) ‘Inflation and Reputation’, American Economic Review 75 (3):530–8. Bagehot, W. (1873) Lombard Street: A Description of the Money Market, London: William Clowes and Sons. Bailey, M.J. (1956) ‘The Welfare Cost of Inflationary Finance’, Journal of Political Economy 63 (1):93–110. Barro, R. (1974) ‘Are Government Bonds Net Wealth?’, Journal of Political Economy 82(6):1095–117. Barro, R. (1986) ‘Reputation in a Model of Monetary Policy with Incomplete Information’, Journal of Monetary Economics 17 (1):3–20.

464

Bibliography

Barro, R. (1991) ‘Economic Growth in a Cross Section of Countries’, Quarterly Journal of Economics 106 (2):407–43. Barro, R. (1995) ‘Optimal Debt Management’, Working Paper 5327, NBER. Barro, R. (1997) Determinants of Economic Growth, Cambridge, MA: MIT Press. Barro, R. and Gordon, D. (1983) ‘A Positive Theory of Monetary Policy in a Natural Rate Model’, Journal of Political Economy 91 (3):589–610. Barro, R. and Sala-i-Martin, X. (1995) Economic Growth. New York: McGraw Hill. Barro, R., Mankiw, N. and Sala-i-Martin, X. (1995) ‘Capital Mobility in Neoclassical Models of Growth’, American Economic Review 85 (1):103–15. Bayoumi, T. and Masson, P. (1998) ‘Liability-Creating Versus Non-Liability Creating Fiscal Stabilization Policies: Ricardian Equivalence, Fiscal Stabilization and EMU’, IMF Working Paper WP/98/112, Washington, DC. Bayoumi, T., Clark, P., and Symansky, S. (1994) ‘Robustness of Equilibrium Exchange Rate Calculations to Alternative Assumptions and Methodologies’, IMF Working Paper IMF/WP/94/17. Benabou, R. (1991) ‘“Comment” on the Welfare Cost of Moderate Inflations’, Journal of Money, Credit, and Banking 23 (3):504–13. Benhabib, J. and Spiegel, M. (1994) ‘The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data’, Journal of Monetary Economics 34 (2):143–73. Benhabib, J. and Spiegel, M. (2000) ‘The Role of Financial Development in Growth and Investment’, Journal of Economic Growth 5 (4):341–60. Bergsten, C.F. (2000) ‘Reforming the International Financial Institutions: A Dissenting View’, Testimony before the Senate Committee on Banking, Housing, and Urban Affairs, March 9. Bernanke, B. and Gertler, M. (1989) ‘Agency Costs, Net Worth and Business Fluctuation’, American Economic Review 79:14–27. Bernanke, B. and Gertler, M. (1990) ‘Financial Fragility and Economic Performance’, Quarterly Journal of Economics 105:87–114. Beseler, J. and Williams, A. (1986) Anti-Dumping and Anti-Sub sidy Law: The European Communities, London. Bhagwati, J. (1982) ‘Directly Unproductive Profit-Seeking (DUP) Activities’, Journal of Political Economy 90 (5):988–1002. Bhagwati, J. (1984) ‘Why are Services Cheaper in Developing Countries?’, Economic Journal 94 (2):279–86. Bhagwati, J. (1986) ‘Ideology and North-South Relations’, World Development 14 (6):767–74. Bhagwati, J. (1987) ‘Protectionism’, 1987 Ohlin Lecture (published 1988), Cambridge, MA: MIT Press. Bhagwati, J. (1989) ‘The Rise of Protectionism’, Economic Impact 2 (1):6–12. Bhagwati, J. and Srinivasan, T. (1979) ‘Domestic Resource Costs, Effective Rates of Protection, and Project Analysis in Tariff-Distorted Economies’, Quarterly Journal of Economics 94 (1):205–9. Bhagwati, J., Panagariya, A. and Srinivasan, T. (1998) Lectures on International Trade, Cambridge MA and London: MIT Press, 2nd edn. Bird, G. (1995) IMF Lending to Developing Countries: Issues and Evidence, London: Routledge. Blanchard, O. and Fischer, S. (eds) (1991) NBER Macroeconomics Annual, Cambridge, MA: MIT Press. Blanchard, O. and Kremer, M. (1997) ‘Disorganization’, Quarterly Journal of Economics 112–14:1091–126. Blanchard, O.J. and Summers, L.H. (1986) ‘Hysteresis and the European

Bibliography

465

Unemployment Problem’, in S.Fischer (ed.), NBER Macroeconomic Annual, Cambridge, MA: MIT Press, pp. 15–78. Blinder, A. (1997) ‘Is Government Too Political?’, Foreign Affairs 76 (6):115–26. Bloem, A. Dippelsman, R. and Maehle, N. (2001) Quarterly National Accounts Manual: Concepts, Data Sources and Compilation, Washington, DC: IMF. Board of Governors of the Federal Reserve System (1994) Money Stock Revisions, April, Washington DC: Board of Governors. Bordo, M. (1980) ‘The Effects of Monetary Change on Relative Commodity Prices and the Role of Long-Term Contracts’, Journal of Political Economy 88 (4): 1088–109. Bordo, M. and Flandreau, M. (2001) ‘Core, Periphery, Exchange Rate Regimes and Globalization’, NBER, mimeo. Bordo, M. and Schwartz, A. (2000) ‘Measuring Real Economic Effects of Bailouts: Historical Perspectives on How Countries in Financial Distress have Fared with and without Bailouts’, Carnegie Rochester Series on Public Policy 53:81–167. Borensztein, E. (1990) ‘Debt Overhang, Credit Rationing and Investment’, Journal of Development Economics 32 (April):315–35. Borensztein, E. (1994) ‘Structural Policies in Developing Countries’, IMF Papers on Policy Analysis and Assessment, PP/AA/94/19. Bosworth, B. and Collins, S. (1999) ‘Capital Flows to Developing Economies: Implications for Saving and Investment’, Brookings Papers on Economic Activity, pp. 143–80. Boyce, K. (1992) ‘The Revolving Door? External Debt and Capital Flight: A Philippine Case Study’, World Development 20:335–49. Brealey, R. and Kaplanis, E. (1999) ‘The Impact of IMF Assistance on Asset Values’, Working Paper, London: Bank of England. Brooks, R., Cortes, M., Fornasari, F., Ketchekmen, B., Metzgen, Y., Powell, R., Rizavi, S., Ross, D. and Ross, K. (1998) ‘External Debt Histories of Ten Low-Income Developing Countries: Lessons from their Experiences’, International Monetary Fund, Working Paper Series WP/98/72. Browning, E. (1987) ‘On the Marginal Welfare Cost of Inflation’, American Economic Review 77 (1):11–23. Buffie, E. (1999) ‘Public Sector Wage-Cycles and the Co-Movement of the Fiscal Deficit and Inflation in Less-Developed Countries’, Journal of Money, Credit and Banking 31 (4):785–810. Buiter, W. (1985) ‘A Guide to Public Sector Debts and Deficits’, Economic Policy 1 (1):13–79. Buiter, W. (1993) ‘Measurement of the Public Sector Deficit and its Implications for Policy Evaluation and Design’, in M.I.Blejer and A.Cheasty (eds), How to Measure the Fiscal Deficit, Washington, DC: IMF, pp. 297–344. Burmeister, E. and Dobell, R. (1970) Mathematical Theories of Economic Growth, New York: Macmillan. Burns, A. (1990) ‘Unemployment in Canada: Frictional, Structural and Cyclical Aspects’, Working Paper 1, Ottawa: Economic Council of Canada. Burnside, C. and Dollar, D. (1997) ‘Aid Spurs Growth—in a Sound Policy Environment’, Finance and Development 34 (4):4–7. Cagan, P. (1956) ‘The Monetary Dynamics of Hyperinflation’, in M.Friedman (ed.), The Quantity Theory of Money and Other Essays, Chicago, IL: University of Chicago Press. Cagan, P. (1965) Determinants and Effects of Changes in the Stock of Money, 1875– 1960, New York: Columbia University Press for NBER. Calvo, G. (1991) ‘The Perils of Sterilization’, IMF Staff Papers 38 (4):921–6. Calvo, G. (1997) ‘Varieties of Capital Market Crises’, in G.Calvo and M.King (eds),

466

Bibliography

The Debt Burden and its Consequences for Monetary Policy, New York: St Martin’s Press. Calvo, G. (1998) ‘Capital Flows and Capital Market Crises: The Simple Economics of Sudden Stops’, Journal of Applied Economics (CEMA) 1 (1):35–54, November. Calvo, G. and Reinhart, C. (2000) ‘Fear of Floating’, Mimeo. College Park, MD: University of Maryland, January. Calvo, G., Leiderman, L. and Reinhart, C. (1996) ‘Inflows of Capital to Developing Countries’, Journal of Economic Perspectives 10 (2):123–40. Camdessus, M. (1987) ‘Opening Remarks’, in V.Corbo, M.Goldstein, and M. Khan (eds), Growth-Oriented Adjustment Programmes, Washington, DC: IMF and World Bank. Camdessus, M. (1995) ‘Entrevista con el Director Gerente del FMI’: PensamientoIberoamericano: Revista-de-Economia-Politica 27:181–4. Canning, D., Fay, M. and Perotti, R. (1994) ‘Infrastructure and Growth’, in M.Baldassarri, L.Paganetto and E.S.Phelps (eds), International Differences in Growth Rates: Market Globalization and Economic Areas, Central Issues in Contemporary Economic Theory and Policy Series, New York: St. Martin’s Press; London: Macmillan Press in association with Rivista di Politica Economica and CEIS, University ‘TorVergata’, Rome, pp. 113–47. Cardoso, E. (1992) ‘Deficit Finance and Monetary Dynamics in Brazil and Mexico’, Journal of Development Economics 37 (November):173–97. Chalk, N. and Hemming, N. (2000) ‘Assessing Fiscal Sustainability in Theory and Practice’, IMF Working Paper WP/00/81, IMF, Washington, DC. Chenery, H. and Srinivasan, T. (1989) Handbook of Development Economics, Amsterdam: North Holland. Chipalkatti, N. and Rishi, M. (2001) ‘External Debt and Capital Flight in the Indian Economy’, Oxford Development Studies 29 (1):31–44. Clark, P.K. (1982) ‘Inflation and the Productivity Decline’, American Economic Review 72 (2):149–54. Coe, D.T. (1990) ‘Structural Determinants of the Natural Rate of Unemployment in Canada’, International Monetary Fund Staff Papers 37 (1):94–115. Conable, B. (1987) ‘Opening Remarks’, in V.Corbo, M.Goldstein and M.Khan (eds), Growth-Oriented Adjustment Programmes, Washington, DC: IMF and World Bank. Conning, J. and Kevane, M. (2002) ‘Why isn’t there more Financial Intermediation in Developing Countries’ , WIDER Discussion Paper 2002/28. Conway, P. (1994) ‘IMF Lending Programmes: Participation and Impact’, Journal of Development Economics 45 (2):365–91. Cooley, T. and Hansen, G. (1989) ‘The Inflation Tax in a Real Business Cycle Model’, American Economic Review 79 (4):733–48. Cooley, T. and Hansen, G. (1991) ‘The Welfare Costs of Moderate Inflations’, Journal of Money, Credit, and Banking 23 (3):483–503. Corbo, V. and Schmidt-Hebbel, K. (1991) ‘Public Policies and Saving in Developing Countries’, Journal of Development Economics 36 (July):89–115. Corden, W.M. (1977) Inflation, Exchange Rates and the World Economy, Oxford: Oxford University Press. Corden, W.M. (1984) ‘Booming Sector and Dutch Disease Economics: Survey and Consolidation’, Oxford Economic Papers 36 (3):359–80. Corden, W.M. (1994) Economic Policy, Exchange Rates and the International System, Oxford: Oxford University Press. Cottani, J., Cavallo, D. and Khan, S. (1990) ‘Real Exchange Rate Behavior and Economic Performance in LDCs’, Economic Development and Cultural Change 39 (1):61–76. Council on Foreign Relations Independent Task Force (1999) ‘Safeguarding Prosperity’, in C.Hills and P.Peterson (co-chairs), M.Goldstein (project director), A Global

Bibliography

467

Financial System: The Future International Financial Architecture, Washington DC: Institute for International Economics. Cozier, B. and Selody, J. (1991) ‘Inflation and Macroeconomic Performance: Some Cross-Country Evidence’, Mimeo, Bank of Canada, August. Cozier, B. and Wilkinson, G. (1991) ‘Some Evidence on Hysteresis and the Costs of Disinflation in Canada’, Bank of Canada, Technical Report, 55. Ottawa. Crow, J. (chairman), Arriazu, R. and Thygesen, N. (1999) External Evaluation of IMF Surveillance, Washington DC: IMF, August. Referred to as the ‘Crow Report’. Cukierman, A. (1984) Inflation, Stagflation, Relative Prices and Imperfect Information. Cambridge: Cambridge University Press. Danziger, L. (1987) ‘On Inflation and Relative Price Variability’, Economic Inquiry 25 (2):285–98. Deardorff, A. and Stern, R. (1998) Measures of Non-Tariff Barriers, Ann Arbor, MI : University of Michigan Press. Debreu, G. (1959) Theory of Value, New York: Wiley. De Gregorio, J. (1992a) ‘Economic Growth in Latin America’, Journal of Development Economics 39 (1):59–84. De Gregorio, J. (1992b) ‘The Effects of Inflation on Economic Growth: Lessons from Latin America’, European Economic Review 36 (3):417–25. De Gregorio, J., Edwards, S. and Valdes, R.O. (2000) ‘Controls on Capital Inflows: Do They Work?’, Journal of Development Economics 63 (1):59–83. De Gregorio, J., Eichengreen, B., Takatoshi, I. and Wyplosz, C. (1999) An Independent and Accountable IMF, Geneva: International Centre for Monetary and Banking Studies and Centre for Economic Policy Research. De Haan, J. and Zelhorst, D. (1990) ‘The Impact of Government Deficits on Money Growth in Developing Countries’, Journal of International Money and Finance 9 (3):455–69. De Haan, W.J. (1990) ‘The Optimal Inflation Path in a Sidrauski-Type Model with Uncertainty’, Journal of Monetary Economics 25 (3):389–409. Deininger, K. and Squire, L. (1998) ‘New Ways of Looking at Old Issues: Inequality and Growth’, Journal of Development Economics 57 (2):259–87. Detragiache, E. (1999) ‘Bank Fragility and International Capital Mobility’, IMF Working Paper 99/113. Devarajan, S., Lewis, J. and Robinson, S. (1993) ‘External Shocks, Purchasing Power Parity, and the Equilibrium Real Exchange Rate’, World Bank Economic Review 7 (1):45–63. De Vries, M. (1987) Balance of Payments Adjustment, 1945 to 1986: The IMF Experience, Washington, DC: International Monetary Fund. Diamond, D. and Dybvig, P. (1983) ‘Bank Runs, Deposit Insurance, and Liquidity’, Journal of Political Economy 91 (3):401–19. Diamond, D. and Verrecchia, R. (1982) ‘Optimal Managerial Contracts and Equilibrium Security Prices’, Journal of Finance 37:275–87. Diamond, P.A. and Mirrlees, J.A. (1971a) ‘Optimal Taxation and Public Production, I: Production Efficiency’, American Economic Review 61 (2):8–27. Diamond, P.A. and Mirrlees, J.A. (1971b) ‘Optimal Taxation and Public Production, II: Tax Rules’, American Economic Review 61 (2):261–78. Diaz-Alejandro, C. (1966) Exchange Devaluation in a Semi-Industrial Country, Cambridge, MA: MIT Press. Diaz-Alejandro, C. (1981) ‘Latin America and the World Economy in the 1980s’, Economic Forum 11 (2):38–53. Diaz-Alejandro, C. (1984) ‘Latin American Debt: I don’t think we are in Kansas anymore’, Brookings Papers on Economic Activity, 2:335–89.

468

Bibliography

Dicks-Mireaux, L., Mecagni, M. and Schadler, S. (2000) ‘Evaluating the Effect of IMF Lending to Low-Income Countries’, Journal of Development Economics 61 (2):495–526. Dijkstra, G. and Hermes, N. (2001) ‘The Uncertainty of Debt Service Payments and Economic Growth of Highly Indebted Poor Countries: Is there a case for debt relief?’, WIDER Working Paper, Helsinki, Finland. Dobb, M. (1937) Studies in the Development of Capitalism, Cambridge: Cambridge University Press. Dollar, D. (1992) ‘Outward-Oriented Developing Economies Really Do Grow More Rapidly: Evidence from 95 LDCs, 1976–1985’, Economic Development and Cultural Change 40 (April):523–44. Dooley, M. (1986) ‘An Analysis of the Debt Crisis’, Working Paper 86/14, International Monetary Fund (December). Dooley, M. (1988) ‘Capital Flight: A Response to Differences in Financial Risk’, IMF Staff Papers 35 (3):422–36. Dooley, M. (1996) ‘A Survey of Literature on Controls over International Capital Transactions’, IMF Staff Papers 43 (4):639–87. Dooley, M. and Isard, P. (1980) ‘Capital Controls, Political Risk, and Deviations from Interest Parity’, Journal of Political Economy 88 (2):370–84. Dooley, M. and Kletzer, K. (1994) ‘Capital Flight, External Debt, and Domestic Policies’, Federal Reserve Bank of San Francisco Economic Review 3:29–37. Dooley, M. and Stone, M. (1993) ‘Endogenous Creditor Seniority and External Debt Values’, IMF Staff Papers 40 (3):395–413. Dooley, M., Fernandez-Arias, E. and Kletzer, K. (1994) ‘Is the Debt Crisis History? Recent private capital inflows to developing countries’, Working Paper 4792, National Bureau of Economic Research (July). Dooley, M., Frankel, J. and Mathieson, D. (1987) ‘International Capital Mobility: What do saving investment correlations tell us?’, IMF Staff Papers 34 (3): 503–30. Dornbusch, R. (1976) ‘Expectations and Exchange Rate Dynamics’, Journal of Political Economy 84 (4):1161–76. Dornbusch, R. (1980) Open-Economy Macroeconomics, New York: Basic Books. Dornbusch, R. (1982) ‘PPP Exchange-Rate Rules and Macroeconomic Stability’, Journal of Political Economy 90 (1):158–65. Dornbusch, R. (1983) ‘Real Interest Rates, Home Goods, and Optimal External Borrowing’, Journal of Political Economy 91 (1):141–53. Dornbusch, R. (1991) ‘Credibility and Stabilization’, Quarterly Journal of Economics 106 (3):837–50. Dornbusch, R. (1993a) ‘Lessons from Experiences with High Inflation’, in R.Dornbusch (ed.) Stabilization, Debt, and Reform, Englewood Cliffs, NJ: Prentice Hall. Dornbusch, R. (1993b) ‘Moderate Inflation’, World Bank Economic Review 7 (1): 1–44. Dornbusch, R. (1998) ‘After Asia: New Directions for the International Financial System’, July, http://www.mit.edu/~rudi. Dornbusch, R. (2000) Keys to Prosperity, Cambridge, MA: MIT Press. Dornbusch, R. (2001) ‘A Primer on Emerging Market Crises’, mimeo, MIT, Department of Economics. Dornbusch, R. and Edwards, S. (1990) ‘Macroeconomic Populism’, Journal of Development Economics 32 (1):247–77. Dornbusch, R. and Fischer, S. (1986) ‘Stopping Hyperinflations, Past and Present’, Weltwirtschaftliches Archiv 122 (1):1–47. Dornbusch, R. and Helmers, F.L. (eds) (1988) The Open Economy: Tools for Policymakers in Developing Countries, New York: Oxford University Press.

Bibliography

469

Dornbusch, R. and Mussa, M. (1975) ‘Consumption, Real Balances and the Hoarding Function’, International Economic Review 16 (June):415–21. Dornbusch, R. and de Pablo, J. (1989) ‘Debt and Macroeconomic Instability in Argentina’ , in J.D.Sachs (ed.), Developing Country Debt and the World Economy, Chicago, IL: University of Chicago Press. Dornbusch, R. and Reynoso, A. (1993) ‘Financial Factors and Economic Development’, in R.Dornbush (ed.), Policymaking in the Open Economy, New York: Oxford University Press. Dornbusch, R. and Simosen, M. (1988) ‘Inflation Stabilization: The Role of Incomes Policy and Monetization’, in R.Dornbusch (ed.), Exchange Rates and Inflation, Cambridge, MA: MIT Press. Dornbusch, R. and Wener, A. (1994) ‘Mexico: Stabilization, Reform, and No Growth’, Brookings Papers on Economic Activity 1:253–315. Dornbusch, R., Dantas, D.V., Pechman, C., Rocha, R. and Simoes D. (1983) ‘The Black Market for Dollars in Brazil’, Quarterly Journal of Economics 98 (1):25–40. Dornbusch, R., Sturzenegger, F. and Wolf, H. (1990) ‘Extreme Inflation: Dynamics and Stabilization’, Brookings Papers on Economic Activity 1 (March):1–84. Doroodian, K. (1993) ‘Macroeconomic Performance and Adjustment under Policies Commonly Supported by the International Monetary Fund’, Economic Development and Cultural Change 41 (4):849–64. Dowd, K. and Mizen, P. (1994) ‘Output Hysteresis and the Costs of Disinflation Inside and Outside the EMS’, Mimeo, Sheffield Hallam University and University of Nottingham. Dreher, A. and Vaubel, R. (2001) ‘Does the IMF Cause Moral Hazard and Political Business Cycles? Evidence from Panel Data’, Institut für Volkswirtschaftslehre und Statistik no. 598–01, Universität Mannheim. Driffil, J., Mizon, G.E. and Ulph, A. (1990) ‘Costs of Inflation’, in B.F.Friedman and F.H.Hahn (eds), Handbook of Monetary Economics, Vol. 2, Amsterdam: North Holland, pp. 1013–66. Durlauf, S. and Quah, D. (1999) ‘The New Empirics of Economic Growth’, in J.Taylor and M.Woodford (eds), Handbook of Macroeconomics, Amsterdam: North Holland, pp. 235–308. Easterly, W. (1999) ‘When is Fiscal Adjustment an Illusion?’, Economic Policy 28 (1):57–86. Easterly, W. (2001) ‘The Lost Decades: Developing Countries’ Stagnation in Spite of Policy Reform 1980–98’, Journal of Economic Growth 6 (2):135–57. Easterly, W. and Levine, R. (1997) ‘Africa’s Growth Tragedy: Policies and Ethnic Divisions’, Quarterly Journal of Economics 112(4):1203–50. Easterly, W. and Schmidt-Hebbel, K. (1993) ‘Fiscal Deficits and Macroeconomic Performance in Developing Countries’, World Bank Research Observer 8 (2):211– 37. Easterly, W., Kremer, M., Pritchett, L. and Summers, L. (1993) ‘Good Policy or Good Luck? Country Growth Performance and Temporary Shocks’, NBER Working Paper 4474, National Bureau of Economic Research, Inc. Eaton, J. and Fernández, R. (1995) ‘Sovereign Debt’, in G.Grossman and K.Rogoff (eds), Handbook of International Economics, Amsterdam: North Holland, pp. 2031–77. Eckstein, Z. and Leiderman, L. (1992) ‘Seigniorage and the Welfare Cost of Inflation: Evidence from an Intertemporal Model of Money and Consumption’, Journal of Monetary Economics 29 (4):389–410. Edwards, E., Gregorio, J. and Valdes, R. (2000) ‘Controls on Capital Inflows: Do They Work?’ , Journal of Development Economics 63 (1):59–83. Edwards, S. (1983) ‘The Short-Run Relation between Growth and Inflation in Latin America’, American Economic Review 73 (June):477–82.

470

Bibliography

Edwards, S. (1984) The Order of Liberalization of the External Sector in Developing Countries. Princeton Essay in International Finance no. 156, Princeton, NJ: Princeton University. Edwards, S. (1986) ‘Are Devaluations Contractionary?’, Review of Economics and Statistics 68 (3):501–8. Edwards, S. (1988) ‘Terms of Trade, Tariffs and Labor Market Adjustment in Developing Countries’, World Bank Economic Review 2 (1):165–85. Edwards, S. (1989a) Real Exchange Rates, Devaluation and Adjustment: Exchange Rate Policies in Developing Countries. Cambridge, MA: MIT Press. Edwards, S. (1989b) ‘On the Sequencing of Structural Reforms’, Working Paper 3138, National Bureau of Economic Research (October). Edwards, S. (1995) ‘Real and Monetary Determinants of Real Exchange Rate Behavior: Theory and Evidence from Developing Countries’, in J.Williamson (ed), Estimating Equilibrium Real Exchange Rates, Washington, DC: Institute for International Economics. Edwards, S. (2001) ‘Does the Current Account Matter?’, presented at NBER Conference on Currency Crises Prevention, forthcoming in an NBER conference volume. Edwards, S. and Vegh, C. (1997) ‘Banks and Macroeconomic Disturbances under Predetermined Exchange Rates’, Journal of Monetary Economics 40 (2):239–78. Edwards, S. and van Wijnbergen, S. (1986) ‘The Welfare Effects of Trade and Capital Market Liberalization’, International Economic Review 27 (1):141–8. Edwards, S. and van Wijnbergen, S. (1989) ‘Disequilibrium and Structural Adjustment’, in H.Chenery and T.N.Srinivasan (eds), Handbook of Development Economics, Amsterdam: North Holland. Eichengreen, B. (1999) ‘Toward a New International Financial Architecture: A Practical Post-Asia Agenda’, Washington, DC: Institute for International Economics, February. Eichengreen, B. (2000) ‘Can the Moral Hazard Created by IMF Loans Be Reduced?’, Geneva Report on the World Economy, No. 1, London: Centre for Economic Policy Research. Eichengreen, B. and Hausman, R. (2000) ‘Exchange Rates and Financial Fragility’, in Federal Reserve Bank of Kansas City, New Challenges for Monetary Policy, Kansas City, KS: Federal Reserve Bank of Kansas City. Eichengreen, B. and Mussa, M. (1998) ‘Capital Account Liberalization and the IMF’, Finance and Development 35 (4):16–19. Eichengreen, B. and Rose, A. (2001a) ‘The Empirics of Currency and Banking Crises’, Wirtschaftspolitische-Blatter 47 (4):395–402. Eichengreen, B. and Rose, A. (2001b) ‘Does it Pay to Defend Against a Speculative Attack?’, NBER Working Paper. EIU (2000) ‘Zimbabwe Country Report’, London: Economist Intelligence Unit, September. Elbadawi, I. and Soto, R. (1997) ‘Real Exchange Rates and Macroeconomic Adjustment in Sub Saharan Africa and Other Developing Countries’, Journal of African Economies 6 (3), Supplement, 74–120. Evans, P. (1983) ‘Price-Level Instability and Output in the US’, Economic Inquiry 21 (2):172–87. Faig, M. (1988) ‘Characterizations of the Optimal Tax on Money When it Functions as a Medium of Exchange’, Journal of Monetary Economics 22 (2):137–48. Feldstein, M.S. (1976) ‘On the Theory of Tax Reform’, Journal of Public Economics 6 (1):77–104. Feldstein, M.S. (1979) ‘The Welfare Cost of Permanent Inflation and Optimal ShortRun Economic Policy’, Journal of Political Economy 87 (3):749–68. Feldstein, M.S. (1982) ‘Inflation, Tax Rules and Investment: Some Econometric Evidence’, Econometrica 50 (3):825–62.

Bibliography

471

Feldstein, M.S. (1983) Inflation, Tax Rules, and Capital Formation, Chicago, IL and London: University of Chicago Press. Feldstein, M.S. (1998) ‘Refocusing the IMF’, Foreign Affairs, pp. 20–33. Feldstein, M.S. and Slemrod, J. (1978) ‘Inflation and the Excess Taxation of Capital Gains on Corporate Stock’, National Tax Journal 31 (2):107–18. Feldstein, M.S. and Summers, L. (1979) ‘Inflation and the Taxation of Capital Income in the Corporate Sector’, National Tax Journal 32 (3):445–70. Feldstein, M.S., Poterba, J. and Dicks-Mireaux, L. (1981) ‘The Effective Tax Rate and the Pretax Rate of Return’, NBER Working Paper 740, August. Fidler, S. (2000) ‘IMF Agrees Need to End Mission Creep’, Financial Times, 19 July. Fischer, S. (1981) ‘Toward an Understanding of the Costs of Inflation: II’, originally published in K.Brunner and A.H.Meltzer (eds), The Costs and Consequences of Inflation, pp. 5–42. Amsterdam: North Holland. Fischer, S. (1982) ‘Adapting to Inflation in the United States Economy’, in R.E.Hall (ed.), Inflation: Causes and Effects, Chicago, IL: Chicago University Press. Fischer, S. (1986) ‘Exchange Rate Versus Money Targets in Disinflation’, in Indexing, Inflation, and Economic Policy, Cambridge, MA: MIT Press Fischer, S. (1993) ‘The Role of Macroeconomic Factors in Growth’, Journal of Monetary Economics 32 (3):485–512. Fischer, S. (1999) On the Need for an International Lender of Last Resort, Washington DC: International Monetary Fund, May. Fischer, S. (2000a) ‘Presentation to the International Financial Institution Advisory Commission (Meltzer Commission)’, Washington, DC: International Monetary Fund, February. Fischer, S. (2000b) ‘Remarks by Stanley Discher, Acting Managing Director’, Remarks to the Bretton Woods Committee Meetings on 4 April, http://www.imf.org/external/ np/speeches/2000/040400.htm. Fischer, S. and Modigliani, F. (1978) ‘Towards an Understanding of the Real Effects and Costs of Inflation’, Weltwirtschaftliches Archiv 114 (4):810–33. Fisher, I. (1930) The Theory of Interest, New York: Macmillan. Forbes, K. and Rigobon, R. (1999) ‘Measuring Contagion: Conceptual and Empirical Issues’, unpublished manuscript, MIT. Ford, R. and Rose, D. (1989) ‘Estimates of the NAIRU Using an Extended Okun’s Law’, Working Paper 89–3, Ottawa: Bank of Canada. Fortin, P. (1990) ‘Can the Costs of an Anti-Inflation Policy be Reduced?’, in R.C. York (ed.), Taking Aim: The Debate on Zero Inflation, Toronto: C.D.Howe Institute, pp. 135–72. Foster, E. (1972) ‘Costs and Benefits of Inflation’, Studies in Monetary Economics, 1, Minneapolis, MN: Federal Reserve Bank of Minneapolis. Foster, E. (1978) ‘The Variability of Inflation’, Review of Economics and Statistics 60 (3):346–50. Frankel, J. (1999) ‘No Single Currency Regime is Right for all Countries or at all Times’, NBER Working Paper 7338. Frankel, J. and Rose, A. (1996) ‘Currency Crashes in Emerging Markets: Empirical Indicators’, NBER Working Paper 5437, National Bureau of Economic Research, Cambridge, MA. Frenkel, J., Goldstein, M. and Masson, P. (1988) ‘International Coordination of Economic Policies: Scope, Methods, and Effects’, in W.Guth (ed.), Economic Policy Coordination: Proceedings of an International Seminar held in Hamburg, Washington, DC: International Monetary Fund, Hamburg: HWWA-Institut für Wirtschaftsforschung, pp. 149–92. Frenkel, J., Goldstein, M. and Masson, P. (1991) ‘Key Issues in the International Coordination of Economic Policies’, in H.J.Blommestein (ed.), The Reality of

472

Bibliography

International Economic Policy Coordination, Contributions to Economic Analysis no. 199, Amsterdam, Oxford and London: North-Holland; distributed in the US and Canada by Elsevier Science, New York, pp. 27–54. Friedman, B. (2000) ‘Monetary Policy’, NBER Working Paper 8057. Friedman, M. (1968) ‘The Role of Monetary Policy’, American Economic Review 58 (1):1–17. Friedman, M. (1977) ‘Nobel Lecture: Inflation and Unemployment’, Journal of Political Economy 85 (3):451–72. Froyen, R.T. and Waud, R.N. (1987) ‘An Examination of Aggregate Price Uncertainty in Four Countries and Some Implications for Real Output’, International Economic Review 28 (3):53–73. Fry, M. (1995) Money, Interest and Banking in Economic Development, Baltimore, MD: Johns Hopkins University Press. Garber, P. (2000) ‘Notes on Market-Based Bank Regulation’, in J.Bisignano, W.Hunter and G.Kaufman (eds), Global Financial Crises: Lessons from Recent Events, Boston, MA: Kluwer Academic Publishers. Garfinkel, M. (1989) ‘What is an “Acceptable” Rate of Inflation?—A Review of the Issues’, Federal Reserve Bank of St. Louis Review, July-August, pp. 3–15. Gavin, W.T. (1990) ‘In Defence of Zero Inflation’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Cleveland, Ohio: Federal Reserve Bank of Cleveland, pp. 43–62. Gavin, W.T. and Stockman, A.C. (1988) ‘The Case for Zero Inflation’, Federal Reserve Bank of Cleveland Economic Commentary, September 15. Gelb, A. (1989) ‘Financial Policies, Growth, and Efficiency’, Working Paper WPS 202, World Bank (June). Gemmell, N. (2000) ‘Fiscal Policy in a Growth Framework’, paper for the WIDER project on New Fiscal Policies for Growth and Poverty Reduction. Ghura, D. and Grennes, T. (1993) ‘The Real Exchange Rate and Macroeconomic Performance in Sub-Saharan Africa’, Journal of Development Economics 42 (1): 155–74. Gilbert, C. (1987) ‘The Impact of Exchange Rates and Developing Country Debt on Commodity Prices’, Oxford University Applied Economics Discussion Paper no. 30. Gillman, M. (1990) ‘Standardizing Estimates of the Welfare Costs of Inflation’, Mimeo, Atlanta, GA: Emory University. Gillman, M. (1992) ‘The Welfare Cost of Inflation in a Cash-in-Advance Economy with Costly Credit’, Mimeo, New Zealand: University of Otago. Goldstein, M. (1997) ‘The Case for an International Banking Standard’, Policy Analyses in International Economics 47, Washington, DC: Institute for International Economics, April. Goldstein, M. (1998) ‘The Asian Financial Crisis: Causes, Cures, and Systemic Implications’, Policy Analyses in International Economics 55, Washington DC: Institute for International Economics, June. Goldstein, M. (2000) Intrahousehold Allocation and Farming in Ghana, PhD Dissertation, Berkeley, CA: University of California. Goldstein, M. and Turner, P. (1996) ‘Banking Crises in Emerging Economies: Origins and Policy Options’, BIS Economic Papers 46, October. Goldstein, M., Kaminsky, G. and Reinhart, C. (2000) Assessing Financial Vulnerability: An Early Warning System for Emerging Economies, Washington, DC: Institute for International Economics, June. Gordon, R. (1989) ‘Hysteresis in History: Was there ever a Phillips Curve?’, American Economic Review Papers and Proceedings 79:220–5. Graff, M. (1999) ‘Financial Development and Economic Growth—A New Empirical Analysis’, Dresden Discussion Papers in Economics 5/99.

Bibliography

473

Granger, C. (1980) ‘Testing for Causality: A Personal Viewpoint’, Journal of Economic Dynamics and Control 2 (4):329–52. Gray, J. (1976) ‘Wage Indexation: A Macroeconomic Approach’, Journal of Monetary Economics 2 (2):221–35. Greenwood, J. and Jovanovic, B. (1990) ‘Financial Development, Growth and the Distribution of Income’, Journal of Political Economy 98:1076–107. Gregory, A., Hansen, B. and Bruce, E. (1996) ‘Residual-Based Tests for Cointegration in Models with Regime Shifts’, Journal of Econometrics 70 (1):99–126. Grier, K.B. and Tullock, G. (1989) ‘An Empirical Analysis of Cross-National Economic Growth’, Journal of Monetary Economics 24 (2):259–76. Griffith-Jones, A. and Spratt, S. (2001) ‘Will the Proposed New Basel Capital Accord have a Net Negative Effect on Developing Countries?’, Institute of Development Studies, Discussion Paper, University of Sussex. Grimes, A. (1992) ‘The Effects of Inflation on Growth: Some International Evidence’, Weltwirtschaftliches Archiv 631–44. Grossman, G. and Helpman, E. (1991) ‘Quality Ladders in the Theory of Growth’, Review of Economic Studies 58:43–61. Grossman, G. and Stiglitz, J. (1980) ‘On the Impossibility of Informationally Efficient Markets’, American Economic Review 70 (3):393–408. Group of Seven Finance Ministers (2000) ‘Strengthening the International Financial Architecture’, Report to the Heads of State and Government, Fukuoka, Japan, July 8. Gundlach, E. (1999) ‘The Economic Growth of Nations in the Twentieth Century’, in Economics: A Biannual Collection of Recent German Studies 60:7–30. Gupta, K.L. (1984) Finance and Economic Growth in Developing Countries, London: Croom Helm. Gylfason, T. (1987) ‘Credit Policy and Economic Activity in Developing Countries with IMF Stabilization’, Princeton Studies in International Finance 60, Princeton, NJ. Gylfason, T. (2000) ‘Fix or Flex? Alternative Exchange Rate Regimes in an Era of Global Capital Mobility’, North American Journal of Economics and Finance 11 (2):173–89. Hafer, R.W. (1986) ‘Inflation Uncertainty and a Test of the Friedman Hypothesis’, Journal of Macroeconomics 8 (3):365–72. Hahn, F. (1977) ‘Keynesian Economics and General Equilibrium Theory: Reflections on some Current Debates’, in G.Harcourt (ed.), The Microeconomic Foundations of Macroeconomics, London: Macmillan. Hahn, F.H. (1969) ‘On Money and Growth’, Journal of Money, Credit and Banking 1 (2):171–87. Hakkio, G. and Rush, M. (1991) ‘Is the Budget Deficit “Too Large”?’ Economic Inquiry 29 (3):429–45. Hamilton, J. and Flavin, M. (1986) ‘On the Limitations of Government Borrowing: A Framework for Empirical Testing’, American Economic Review 76 (4):808–19. Hansen, B. (1973) A Survey of General Equilibrium Systems, New York: McGraw Hill. Haque, N. (1988) ‘Fiscal Policy and Private Sector Saving Behavior in Developing Economies’, IMF Staff Papers 35 (2):316–35. Haque, N., Lahiri, K. and Montiel, P. (1990) ‘A Macroeconometric Model for Developing Countries’, IMF Staff Papers 37 (3):537–59. Harberger, A. (1985) ‘Lesson for Debtor-Country Managers and Policymakers’, in G.Smith and J.Cuddington (eds), International Debt and the Developing Countries, Washington DC: The World Bank, pp. 236–57. Heady, C. (2000) ‘Taxation Policy in Low-Income Countries’, paper for the WIDER project on New Fiscal Policies for Growth and Poverty Reduction.

474

Bibliography

Helleiner, G. (1987) ‘Comment on Adjustment and Economic Growth’, in V.Corbo, M.Goldstein and M.Khan (eds), Growth-Oriented Adjustment Programmes, Washington, DC: IMF and World Bank. Hemphill, W. (1974) ‘The Effect of Foreign Exchange Receipts on Imports of Less Developed Countries’, IMF Staff Papers 21 (3):637–77. Hermes, N. and Lensink, R. (2000) ‘Fiscal Policy and Private Investment in Less Developed Countries’, paper for the WIDER project on New Fiscal Policies for Growth and Poverty Reduction. Hicks, J. (1969) A Theory of Economic History, Oxford: Clarendon Press. Hjertholm, P. (2000) ‘Analytical History of Heavily Indebted Poor Country (HIPC) Debt Sustainability Targets’, University of Copenhagen, Institute of Economics Discussion Paper 00/03. Holland, S. (1986) ‘Wage Indexation and the Effect of Inflation Uncertainty on Employment: An Empirical Analysis’, American Economic Review 76 (2):235–43. Holland, S. (1988) ‘Indexation and the Effect of Inflation Uncertainty on Real GNP’, Journal of Business 61 (3):473–84. Holmstrom, B. and Tirole, J. (1993) ‘Markets, Liquidity and Performance Monitoring’, Journal of Political Economy 101:678–709. Hoskins, W.L. (1990) ‘A US Perspective on Zero Inflation’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Toronto: C.D.Howe Institute, pp. 34–42. Hoskins, W.L. (1992) ‘Views on Monetary Policy’, paper presented to the Annual Monetary Policy Conference, Federal Reserve Bank of St Louis, MO: October. Howitt, P. (1986) ‘Indexation and the Adjustment to Inflation in Canada’, in T.J. Sargent (ed.), Postwar Macroeconomic Developments, Toronto: University of Toronto Press. Howitt, P. (1990a) ‘Zero Inflation as a Long-Term Target for Monetary Policy’, in R.G.Lipsey (ed.), Zero Inflation: The Goal of Price Stability, Toronto: C.D. Howe Institute, pp. 67–108. Howitt, P. (1990b) ‘A Comment’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Toronto: C.D. Howe Institute, pp. 104–8. Humphreys, C. and Underwood, J. (1989) ‘The External Debt Difficulties of LowIncome Africa’, in I.Husain and I.Diwan (eds), Dealing with the Debt Crisis. A World Bank Symposium , Washington, DC: World Bank, pp. 45–65. Husain, I. and Underwood, J. (1991) ‘African External Finance in the 1990s’, A World Bank Symposium, Washington, DC. Hutchinson, M. (2002) ‘A Cure Worse than the Disease? Currency Crises and Output Costs of Supported Stabilization Programmes’, in M.Dooley and J.Frankel (eds), Managing Currency Crises in Emerging Markets, Chicago, IL: University of Chicago Press. Ibi, A. (1997) ‘An Analysis of External Debt and Capital Flight in the Severely Indebted Low Income Countries in Sub-Saharan Africa’, IMF Working Paper WP/97/68, June. Imrohoroglu, A. (1992) ‘The Welfare Cost of Inflation under Imperfect Insurance’, Journal of Economic Dynamics and Control 6 (1):79–91. International Development Association and IMF (2001) ‘The Challenge of Maintaining Long-Term External Debt Sustainability’, Washington, DC. International Institute of Finance (IIF) (1999) ‘Testing for “Moral Hazard”, in Emerging Markets Lending’, IIF Research Paper 99–1 (August 1999) International Financial Institution Advisory Commission, Allan H.Meltzer, Chairman. Washington DC, referred to as the ‘Meltzer Report’ and the ‘Meltzer Commission’, March, 2000. International Monetary Fund (1995) World Economic Outlook 1995, Washington, DC: IMF.

Bibliography

475

International Monetary Fund (1996) World Economic Outlook 1996, Washington, DC : IMF. International Monetary Fund (1999a) ‘HIPC Initiative—A Progress Report’. Available on website http://www.imf.org/external/np/hipc/prog 3/092299.htm (September 1999). International Monetary Fund (1999b) World Economic Outlook 1999, Washington, DC: IMF. International Monetary Fund (2000a) ‘Report of the Acting Managing Director to the International Monetary and Financial Committee on Progress in Reforming the IMF and Strengthening the Architecture of the International Financial System,’ Washington DC, IMF, April 12. International Monetary Fund (2000b) ‘IMF Board Agrees on Changes to Fund Financial Facilities’, Public Information Notice no. 00/79, International Monetary Fund, Washington, DC. International Monetary Fund (2000c) Review of the Burkina Faso Economy, available on the website www.imf.org. International Monetary Fund (2000d) Report to the IMF Executive Board of the Quota Formula Review Group. International Monetary Fund (2001) ‘Conditionally in Fund Supported Programmes— Policy Issues’, on www.imf.org. Issler, J. and Lima, L. (2000) ‘Public Debt Sustainability and Endogenous Seignorage in Brazil: Time Series Evidence from 1947–92’, Journal of Development Economics 62 (1):131–47. Jansen, D.W. (1989) ‘Does Inflation Uncertainty Affect Output Growth? Further Evidence’, Federal Reserve Bank of St Louis Review, July-August: pp. 43–54. Jarrett, J.P. and Selody, J.G. (1982) ‘The Productivity-Inflation Nexus in Canada, 1963–79’, Review of Economics and Statistics 64 (3):361–7. Jeanne, O. (2000) ‘Currency Crises: A Perspective on Recent Theoretical Developments’, Special Papers in International Economics, no. 20. International Finance Section, Princeton, NJ: Princeton University. Jha, R. (1998) Modern Public Economics, London and New York: Routledge. Jha, R. (2000) ‘Reducing Poverty and Inequality in India: Has Liberalization Helped?’, Working Paper 204, WIDER, Helsinki. Jha, R. and Rath, D. (2002) ‘On the Endogeneity of the Money Multiplier in India’, Australia South Asia Research Centre Working Paper, Australian National University. Jha, R. and Saggar, M. (2000) ‘Towards a More Rational IMF Quota Structure: Suggestions for the Creation of a New International Financial Architecture’, Development and Change 31 (3):579–604. Jha, R. and Sahni, B. (1993) Industrial Efficiency: An Indian Perspective, New Delhi: Wiley Eastern. Jha, R. and Schatan, C. (2001) ‘Debt for Nature: A swap whose time has gone?’, ECLAC, United Nations no. 4, Mexico City. Jha, R., Sahu, A. and Meyer, L. (1990) ‘The Fisher Equation Controversy: A Reconciliation of Contradictory Results’, Southern Economic Journal 57 (1):106–13. Jovanovic, B. (1989) ‘Observable Implications of Models with Multiple Equilibria’, Econometrica 57 (6):1431–7. Kamin, S. and Wood, P. (1997) ‘Capital Flows, Financial Intermediation and Aggregate Demand: Evidence from Mexico and other Pacific Basin Countries’, United States Board of Governors of the Federal Reserve System: International Finance Discussion Papers 583, Washington, DC. Kaminsky, G. (1999) ‘Currency and Banking Crises: The Early Warnings of Distress’, unpublished IMF Working Paper WP/99/178. December.

476

Bibliography

Kaminsky, G. and Reinhart, C. (1996) ‘The Twin Crises: The Causes of Banking and Balance-of-Payments Problems’, Board of Governors of the Federal Reserve System, International Finance Discussion Papers 544, March. Kaminsky, G. and Reinhart, C. (1998) ‘The Twin Crises: The Causes of Banking and Balance-of-Payments Problems’, unpublished, College Park, MD: University of Maryland. Kaminsky, G. and Reinhart, C. (1999) ‘The Twin Crises: The Causes of Banking and Balance of Payments Problems’, American Economic Review 89, June. Kant, C. (2002) ‘What is Capital Flight?’, The World Economy 25 (3):341–58. Kantor, L.G. (1986) ‘Inflation Uncertainty and Real Economic Activity: An Alternative Approach’, Review of Economics and Statistics 68 (3):493–500. Kaplan, E. and Rodrik, D. (2001) ‘Did the Malaysian Capital Controls Work?’, NBER Working Paper 8142. Kapur, B. (1976) ‘Alternative Stabilization Policies for Less Developed Economies’, Journal of Political Economy 84 (4):777–95. Kapur, B. (1992) ‘Formal and Informal Financial Markets, and the Neo-Structuralist Critique of the Financial Liberalization Strategy in Less Developed Countries’, Journal of Development Economics 38 (1):63–77. Karnik, A.V. (1990) ‘Elections and Government Expenditures: The Indian Evidence’, Journal of Quantitative Economics 6 (January):203–12. Kendrick, D. (1976) ‘Applications of Control Theory to Macroeconomics’, Annals of Economic and Social Measurement 5 (2):171–90. Kenny, C. and Williams, D. (2001) ‘What do we Know about Economic Growth? Or, why don’t we know very much?’, World Development 29 (1):1–22. Keynes, J. (1936) The General Theory of Employment, Interest and Money, London: Macmillan. Keynes, J. (1973) The Collected Writings of John Maynard Keynes, D.Moggridge (ed.) London: Macmillan. Khan, M. (1980) ‘Monetary Shocks and the Dynamics of Inflation’, IMF Staff Papers 27 (2):250–84. Khan, M. (1990) ‘Evaluating the Effects of IMF-Supported Adjustment Programmes: A Survey’, in K.Phylaktis and M.Pradham (eds), International Finance and the Less Developed Countries, New York: St Martin’s Press. Khan, M. and Knight, M. (1981) ‘Stabilization Programmes in Developing Countries: A Formal Framework’, IMF Staff Papers 28 (1):1–53. Khan, M. and Kumar, M. (1994) ‘Determinants of the Current Account in Developing Countries 1970–90’, unpublished, International Monetary Fund (March). Khan, M. and Kumar, M. (1997) ‘Public and Private Investment and the Growth Process in Developing Countries’, Oxford Bulletin of Economics and Statistics 59 (1):69–88. Khan, M. and Lizondo, J.S. (1987) ‘Devaluation, Fiscal Deficits, and the Real Exchange Rate’, World Bank Economic Review 1 (3):357–74. Khan, M. and Montiel, P. (1987) ‘Real Exchange Rate Dynamics in a Small PrimaryExporter Country’, IMF Staff Papers 34 (3):687–710. Khan, M. and Ramirez-Rojas, C.L. (1986) ‘Currency Substitution and Government Revenue from Inflation’, Revista de Analisis Economico 1 (1):79–88. Khan, M. and Senhadji, A. (2000) ‘Financial Development and Economic Growth: An Overview’, IMF Working Paper 00/209, December. Khan, M. and Zahler, R. (1983) ‘The Macroeconomic Effects of Changes in Barriers to Trade and Capital Flows’, IMF Staff Papers 30 (2):223–82. Khan, M. and Zahler, R. (1985) ‘Trade and Financial Liberalization Given External Shocks and Inconsistent Domestic Policies’, IMF Staff Papers 32 (1): 22–55.

Bibliography

477

Khan, M., Montiel, P. and Haque, N. (1990) ‘Adjustment with Growth: Relating the Analytical Approaches of the IMF and the World Bank’, Journal of Development Economics 32 (1):155–79. Killick, T. (1995) IMF Programmes in Developing Countries, London: Routledge. Kimbrough, K.P. (1986) ‘The Optimal Quantity of Money Rule in the Theory of Public Finance’, Journal of Monetary Economics 18 (2):277–84. King, R. and Levine, R. (1993a) ‘Finance and Growth: Schumpeter Might be Right’, Quarterly Journal of Economics 108:717–37. King, R. and Levine, R. (1993b) ‘Finance, Entrepreneurship, and Growth: Theory and Evidence’, Journal of Monetary Economics 32 (3):513–42. Klein, B. (1975a) ‘The Impact of Inflation on the Term Structure of Corporate Financial Investments: 1900–72’, in W.L.Silber (ed.), Financial Innovation, Lexington, MA: Lexington, pp. 125–49. Klein, B. (1975b) ‘Our New Monetary Standard: Measurement and Effects of Price Uncertainty, 1880–1973’, Economic Inquiry 13 (3):461–84. Klein, B. (1976) ‘The Social Costs of the Recent Inflation: The Mirage of Steady Anticipated Inflation’, Carnegie-Rochester Conference Series on Public Policy 3: Amsterdam: North Holland, pp. 185–212. Knack, S. and Keefer, P. (1995) ‘Institutions and Economic Performance: Cross-country Tests Using Alternative Institutional Measures’, Economics and Politics 7 (3):207–27. Kohler, H. (2000) ‘Address to the Board of Governors of the Fund’, Annual Meetings of the IMF and World Bank Group, Prague, September 26. Kormendi, R.C. and Meguire, P.G. (1985) ‘Macroeconomic Determinants of Growth: Cross-Country Evidence’, Journal of Monetary Economics 16 (2):141–63. Kouri, P. (1976) ‘The Exchange Rate and the Balance of Payments in the Short Run and in the Long Run: A Monetary Approach’, Scandinavian Journal of Economics 78 (2):280–304. Kouri, P. (1984) The IMF and Stabilization, New York: St Martin’s Press. Krueger, A. (1978) Liberalization Attempts and Consequences, Cambridge, MA: NBER. Krugman, P. (1995) ‘Technology, Trade, and Factor Prices’, NBER Working Paper 5355. Krugman, P. (1998) ‘Ricardo’s Difficult Idea: Why Intellectuals don’t Understand Comparative Advantage’, in G.Cook (ed.), Freedom and Trade, Vol. 2. The economics and politics of international trade. Studies in the Modern World Economy 10, London and New York: Routledge. Krugman, P. and Taylor, L. (1978) ‘Contractionary Effects of Devaluation’, Journal of International Economics 8 (3):445–56. Kydland, F. and Prescott, E. (1977) ‘Rules Rather than Discretion: The Inconsistency of Optimal Plans’, Journal of Political Economy 87 (2):473–92. Lancaster, C. and Williamson, J. (eds) (1986) African Debt and Financing, Special Reports Series no. 5, Institute of International Economics, Washington, DC. Lane, T. and Phillips, S. (2000) ‘Does IMF Financing Result in Moral Hazard?’, IMF Working Paper WP/00/168. Leach, J. (1983) ‘Inflation as a Commodity Tax’, Canadian Journal of Economics 16 (3):508–16. Leech, D. (2001) ‘Members’ Voting Power in the Governance of the International Monetary Fund’, CSGR Working Paper 68/01. Leff, N. and Sato, K. (1980) ‘Macroeconomic Adjustment in Developing Countries: Instability, Short Run Growth and External Dependency’, Review of Economics and Statistics 63 (1):170–9. Levi, M.D. and Makin, J. (1980) ‘Inflation Uncertainty and the Phillips Curve: Some Empirical Evidence’, American Economic Review 70 (4):1022–7.

478

Bibliography

Levine, R. (1997) ‘Financial Development and Economic Growth: Views and Agenda’, Journal of Economic Literature 35:688–726. Levine, R. (2001) ‘International Financial Liberalization and Economic Growth’, Review of International Economics 9 (4):688–702. Levine, R. and Renelt, D. (1992) ‘A Sensitivity Analysis of Cross-Country Growth Regressions’, American Economic Review 82 (4):942–63. Levine, R. and Zervos, S. (1993) ‘What have we Learned about Policy and Growth from Cross-Country Regressions?’, American Economic Review 83 (2): 426–30. Levine, R. and Zervos, S. (1996) ‘Stock Market Development and Long-Run Growth’, World Bank Economic Review 10 (May):323–39. Levine, R., Loayza, N. and Beck, T. (2000) ‘Financial Intermediation and Growth: Causality and Causes’, Journal of Monetary Economics 46 (1):31–77. Lewis, W.A. (1954) ‘Economic Development with Unlimited Supplies of Labour’, The Manchester School 18 (1):1–32. Lim, E. (2001) ‘Determinants of and Relation between Foreign Direct Investment and Growth: A Summary of the Recent Literature’, IMF Working Paper WP/01/ 175. Londregan, J. and Poole, K. (1992) ‘The Seizure of Executive Power and Economic Growth: Some Additional Evidence’, in A.Cukierman, S.Hercowitz and L.Leiderman (eds), Political Economy, Growth, and Business Cycles, Cambridge, MA and London: MIT Press. Lucas, R.E. (1972) ‘Expectations and the Neutrality of Money’, Journal of Economic Theory 4 (1):115–38. Lucas, R.E. (1981) ‘Discussion’, in K.Brunner and A.M.Meltzer (eds), The Costs and Consequences of Inflation, Carnegie-Rochester Conference Series on Public Policy 15. Amsterdam: North Holland, pp. 43–52. Lucas, R.E. (1988) ‘On the Mechanics of Economic Development’, Journal of Monetary Economics 22:99–122. Lucas, R.E. (1990) ‘Supply-Side Economics: An Analytical Review’, Oxford Economic Papers 42 (2):293–316. Lucas, R.E. (2000a) ‘Inflation and Welfare’, Econometrica 68 (2):247–74. Lucas, R.E. (2000b) ‘Some Macroeconomics for the 21st Century’, Journal of Economic Perspectives 41 (1):159–68. Lucas, R.E. and Stokey, N. (1983) ‘Optimal Fiscal and Monetary Policy in a World Without Capital’, Journal of Monetary Economics 12 (1):55–93. Lucas, R.F. (1989) ‘The Bank of Canada and Zero Inflation: A New Cross of Gold?’, Canadian Public Policy 15 (1):84–93. Lucas, R.F. (1990) ‘The Case for Stable, But Not Zero, Inflation’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Toronto: C.D.Howe Institute, pp. 65–80. Lutz, M. (2001) ‘Globalization, Convergence and the Case for Openness in Developing Countries: What do we learn from open economy growth theory and empirics?’, CSGR Working Paper 72/01, University of Warwick. Makin, J.H. (1982) ‘Anticipated Money, Inflation Uncertainty, and Real Economic Activity’, Review of Economics and Statistics 65 (3):379–84. Mankiw, N., Romer, D. and Weil, D. (1992) ‘A Contribution to the Empirics of Economic Growth’, Quarterly Journal of Economics 107 (2):407–37. McCallum, B. (1989) Monetary Economics: Theory and Policy. New York: Macmillan. McCallum, J. (1988) ‘Les taux de chomage canadien et americain dans les années 1980: un test de trois hypotheses’, L’Actualité Economique 64 (3):495–508. McKinnon, R. (1973) Money and Capital in Economic Development, Washington, DC: The Brookings Institution. McKinnon, R. (1976) ‘Saving Propensities and the Koran Monetary Reform in Retrospect’, in Ronald McKinnon (ed.), Finance in Growth and Development, New York: Marcel Dekker.

Bibliography

479

McKinnon, R. (1984) ‘Financial Liberalization and the Debt Crisis in LDCs: The International Misregulation of Commercial Banks’, Stanford, CA: Stanford Centre for Research in Economic Growth, Memo no. 265. McKinnon, R. (1993) The Order of Economic Liberalization. 2nd edn, Baltimore, MD: Johns Hopkins University Press. McKinnon, I.R. and Mathieson, D.J. (1981) How to Manage a Repressed Economy, Essay in International Finance no. 145, Princeton, NJ: Princeton University. McKinnon, R. and Pil, H. (1995) ‘Credible Liberalization and International Capital Flows’, Stanford University, CA, USA (mimeo). McPherson, M. (1999) ‘Seignorage in Highly Indebted Developing Countries’, Development Discussion Paper, no. 696, Cambridge, MA: Harvard Institute for International Development. Meltzer, A. (1999) ‘What’s Wrong with the IMF? What Would Be Better?’, in P.Hunter, A.Kauffman and A.Krueger (eds), The Asian Financial Crisis, Boston, MA: Kluwer Academic Publishers. Ministry of Finance, Government of India (1997) Economic Survey, New Delhi. Mishkin, F. (2000) ‘Inflation Targeting in Emerging Market Countries’, Working Paper 7618, NBER. Mistry, P. (1992) ‘Fragile Finance: Rethinking the International Monetary System: Summary of Workshop Discussion’, in J.Teunissen (ed.), Fragile Finance: Rethinking the International Monetary System, The Hague: Forum on Debt and Development, pp. 121–5. Mohamad, M. (1999) A New Deal for Asia, Malaysia: Pelanduk. Montiel, P. (1999) ‘Determinants of the Long-Run Equilibrium Real Exchange Rate: An Analytical Model’, in L.Hinkle and P.Montiel (eds), Exchange Rate Misalignment: Concepts and Measurement for Developing Countries, Oxford and New York: Oxford University Press for the World Bank, pp. 264–90. Moran, C. (1988) ‘A Structural Model for Developing Countries’ Manufactured Exports’, World Bank Economic Review 2 (3):321–40. Motley, B. (1990) ‘Has There Been a Change in the Natural Rate of Unemployment?’ Federal Reserve Bank of San Francisco Economic Review, pp. 3–16. Muhleisen, M. (1997) ‘Improving India’s Saving Performance’, IMF Working Paper WP/97/04. Mullineaux, D. (1980) ‘Unemployment, Industrial Production, and Inflation Uncertainty in the United States’, Review of Economics and Statistics 62 (2):163–9. Mundell, R. (1968) International Economics, New York: Macmillan. Mundell, R. (2000) ‘Threat to Prosperity’, Wall Street Journal, March. Mussa, M. (1977) ‘The Welfare Cost of Inflation and the Role of Money as a Unit of Account’, Journal of Money, Credit, and Banking 9 (2):276–86. Mussa, M. (1994) ‘Exchange Rate Policy’, in M.Feldstein (ed.), American Policy in the 1980s, Chicago, IL: University of Chicago Press for NBER, pp. 349–60. Mussa, M. (1999) ‘Reforming the International Financial Architecture: Limiting Moral Hazard and Containing Real Hazard’, Washington, DC: International Monetary Fund, August. Nafziger, E.W. (1993) ‘Debt, Adjustment, and Economic Liberalization in Africa’, Journal of Economic Issues 27 (2):427–39. Nam, C. (1986) ‘Export Promoting Policies under Countervailing Threats: GATT Rules and Practices’, Report Number VPERS 9, World Bank. Ndulu, B.J. (1991) ‘Growth and Adjustment in Sub-Saharan Africa’, in A.Chibber and S.Fischer (eds), Economic Reform in Sub-Saharan Africa, Washington, DC: World Bank, pp. 287–302. Nordhaus, W. (1975) ‘The Political Business Cycle’, Review of Economic Studies 42 (1):169–90.

480

Bibliography

Nordhaus, W. (1989) ‘Alternative Models of Political Business Cycles’, Brookings Papers on Economic Activity, no. 1, pp. 1–68. North, D. (1990) The Political Economy of Institutions and Decisions Series, Cambridge: Cambridge University Press. North, D. (1995) ‘The New Institutional Economics and Third World Development’, in J.Harriss, J.Hunter and C.Lewis (eds), The New Institutional Economics and Third World Development, London and New York: Routledge, pp. 17–26. Novin, F. (1991) ‘The Productivity-Inflation Nexus Revisited: Canada, 1969–1988’, Bank of Canada Working Paper 1991–1, February. Obsfeldt, M. (1996) ‘Models of Currency Crises with Self-Fulfilling Features’, European Economic Review 40 (4):1037–47. Obsfeldt, M. (1998) ‘The Global Capital Market: Benefactor or Menace?’ NBER Working Paper 6559, Cambridge, MA. Obsfeldt, M. and Rogoff, K. (1996) Foundations of International Macroeconomics, Cambridge, MA: MIT Press. OECD’s DAC Journal (2000) Development Co-operation, 2000 Report, 2 (1): 178–9. Okun, A.M. (1975) ‘Inflation: Its Mechanics and Welfare Costs’, Brookings Papers on Economic Activity 2:351–90. Palley, T. (1987) ‘Bank Lending, Discount Window Borrowing, and the Endogenous Money Supply: A Theoretical Framework’, Journal of Post Keynesian Economics 10 (2):282–303. Papadopoulos, A. and Sidiropoulos, M. (1999) ‘The Sustainability of Fiscal Deficits in the European Union’, International Advances in Economic Research 5 (3): 289– 307. Parfitt, T. and Riley, S. (1989) The African Debt Crisis, London: Routledge. Pastor, M. (1987) The International Monetary Fund and Latin America: Economic Stabilization and Class Conflict, Series in Political Economy and Economic Development in Latin America, Boulder, CO and London: Westview Press. Patel, U. and Srivastava, P. (1998) ‘The Real Exchange Rate in India: Determinants and Targeting’, LSE Discussion Paper 323. Paul, M. (2001) ‘Indian Real Exchange Rate Behaviour and Durable Good Market under Exchange Rate Uncertainty’, PhD Dissertation, IGIDR, Mumbai. Pearce, D.K. (1979) ‘Comparing Survey and Rational Measures of Expected Inflation: Forecast Performance and Interest Rate Effects’, Journal of Money, Credit, and Banking 11 (3):447–56. Persson, T. and Tabellini, G. (1994) ‘Is Inequality Harmful for Growth?’, American Economic Review 84 (2):600–21. Peters, D. (1990) ‘Fallacies and Detriments of a Zero-Inflation Policy’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Cleveland, Ohio: Federal Reserve Bank of Cleveland, pp. 111–20. Phelps, E. (1966) Golden Rules of Economic Growth, New York: Norton. Phelps, E. (1967) ‘Phillips Curves, Expectations of Inflation and Optimal Unemployment Over Time’ Economica 33 (1):309–37. Phelps, E. (ed.) (1969) Microeconomic Foundations of Employment and Inflation Theory, New York: Norton. Phelps, E.S. (1973) ‘Inflation in the Theory of Public Finance’, Swedish Journal of Economics 75 (1):67–82. Pollak, J. (1957) ‘Monetary Analysis of Income Formation and Payments Problems’, IMF Staff Papers 6 (November):1–50. Pollak, J. (1989) Financial Policies and Development, Paris: Organisation for Economic Cooperation and Development. Poole, W. (1970) ‘Optimal Choice of Monetary Instruments in a Simple Stochastic Macro Model’, Quarterly Journal of Economics, 84 (2):197–216.

Bibliography

481

Porter, R. and Ranney, S. (1982) ‘An Eclectic Model of Recent LDC Macroeconomic Analyses’, World Development 10 (9):751–65. Prachowny, M.J. (1984) Macroeconomic Analysis for Small Open Economies, Oxford: Clarendon Press. Praetzel, G.D. (1981) ‘How Inflation has Affected Investment Decisions’, Voice of the Federal Reserve Bank of Dallas, June, 1–9. Pritchett, L. (1996a) ‘Measuring Outward Orientation in LDCs: Can It Be Done?’, Journal of Development Economics 49 (2):307–35. Pritchett, L. (1996b) ‘Forget Convergence: Divergence Past, Present, and Future’, Finance and Development 33 (2):40–3. Pritchett, L. (1997) ‘Divergence, Big Time’, Journal of Economic Perspectives 11 (3): 3–17. Przeworski, A. and Vreeland, J. (2000) ‘The Effect of IMF Programmes on Economic Growth’, Journal of Development Economics 62 (2):385–421. Ranis, G. (1989) ‘Analytics of Development: Dualism’, in H.Chenery and T.N. Srinivasan (eds) Handbook of Development Economics 1, Amsterdam: North Holland. Rath, D. (2001) ‘Endogeneity Issues in the Money Supply Process of India’, unpublished PhD dissertation, IGIDR, Mumbai, India. Rattso, J. (1989) ‘Macrodynamic Adjustment Mechanisms in a Dual SemiIndustrialized Economy’, Journal of Development Economics 30 (1):47–69. Razin, O. and Collins, S. (1997) ‘Real Exchange Rate Misalignments and Growth’, NBER Working Paper 6174. Rebelo, S. (1991) ‘Long-Run Policy Analysis and Long-Run Growth’, Journal of Political Economy 99:500–21. Reisen, H. (1999) ‘Sustainable and Excessive Current Account Deficits’, in J.Gacs, R.Holzman and M.Wyzan (eds), The Mixed Blessing of Financial Flows, Cheltenham, UK: Edward Elgar. Reserve Bank of India (2000) Reserve Bank of India Bulletin, Mumbai. Rodrik, D. (1999) ‘Globalisation and Labour, or: If globalisation is a bowl of cherries, why are there so many glum faces around the table?’, in R.Baldwin (ed.), Market Integration, Regionalism and the Global Economy, Cambridge, New York and Melbourne: Cambridge University Press, pp. 117–50. Rogoff, K. (1985) ‘The Optimal Degree of Commitment to an Intermediate Monetary Target’, Quarterly Journal of Economics 100 (4):1169–88. Romer, C. (1990) ‘The Great Crash and the Onset of the Great Depression’, Quarterly Journal of Economics 105 (3):597–624. Romer, D. (1996) Advanced Macroeconomics, New York: McGraw-Hill. Romer, P. (1986) ‘Increasing Returns and Long-Run Growth’, Journal of Political Economy 94:1002–37. Rose, D. (1988) ‘The NAIRU in Canada: Concepts, Determinants, and Estimates’, Bank of Canada Technical Report no. 50, Ottawa. Roubini, N. (2000) ‘Material on Currency Crises’, at http://www.stern.nyu.edu/ globalmacro/ 13. Roubini, N. and Sala-i-Martin, X. (1991) ‘The Relation between Trade Regime, Financial Development and Economic Growth’, Mimeo, New Haven, CT: Yale University. Sachs, J. and Warner, A. (1995) ‘Natural Resource Abundance and Economic Growth’, NBER Working Paper 5398. Sah, R. and Stiglitz, J. (1989) ‘Sources of Technological Divergence between Developed and Less Developed Economies’, in G.Calvo, C.Diaz-Alejandro and C.Federico (eds), Debt, Stabilization and Development, Oxford: Basil Blackwell.

482

Bibliography

Salter, W. (1959) ‘Internal and External Balance: The Role of Price and Expenditure Effects’, Economic Record 35 (1):226–38. Santaella, J. (1996) ‘Stylized Facts before IMF-Supported Macroeconomic Adjustment’, IMF Staff Papers 43 (3):502–44. Sarel, M. (1996) ‘Nonlinear Effects of Inflation on Economic Growth’, IMF Staff Papers 43 (1):199–215. Sargent, T. (1999) ‘A Primer on Monetary and Fiscal Policy’, Journal of Banking and Finance 23 (8):1463–82. Scarth, W. (1990) ‘Fighting Inflation: Are the Costs of Getting to Zero too High?’, in R.C.York (ed.), Taking Aim: The Debate on Zero Inflation, Cleveland, Ohio: Federal Reserve Bank of Cleveland, pp. 81–103. Schineller, L. (1993) Hysteresis, Political Risk and International Capital Movements: Essays on Capital Flight, PhD dissertation, New Haven, CT: Yale University. Schneider, M. (2000) ‘When Financial Intermediaries are Corporate Owners: An Agency Model of Institutional Ownership’, Journal of Management and Governance 4 (3):207–37. Schuknecht, L. (1999) ‘Fiscal Policy Cycles and the Exchange Rate Regime in Developing Countries’, European Journal of Political Economy 15 (3):569–80. Selody, J. (1990a) ‘The Goal of Price Stability: A Review of the Issues’, Bank of Canada Technical Report no. 54, Ottawa. Selody, J.G. (1990b) ‘The Benefits and Costs of Price Stability: An Empirical Assessment’, Mimeo, Bank of Canada. Sen, A. (1964) Economic Growth, Harmondsworth: Penguin. Sherden, W. (1998) The Fortune Sellers, New York: John Wiley and Sons. Sidrauski, M. (1967a) ‘Inflation and Economic Growth’, Journal of Political Economy 75 (6):796–810. Sidrauski, M. (1967b) ‘Rational Choice and Patterns of Growth in a Monetary Economy’, American Economic Review 62 (2):534–44. Singh, D. (1997) South East Asian Affairs, Singapore: Institute of Southeast Asian Studies. Solow, R. (1956) ‘A Contribution to the Theory of Economic Growth’, Quarterly Journal of Economics 70 (1):26–41. Solow, R. (1988) ‘Growth Theory and After’, American Economic Review 78 (3): 307–17. Solow, R. (1999) ‘Neoclassical Growth Theory’, in J.Taylor and M.Woodford (eds) Handbook of Macroeconomics, Amsterdam: North Holland, pp. 637–67. Spiegel, M. (1995) ‘Sterilization of Capital Inflows through the Banking Sector: Evidence from Asia’, Federal Reserve Bank of San Francisco Economic Review 3: 17–34. Stein, J. (1995) ‘An Adverse Selection Model of Bank Asset and Liability Management with Implications for the Transmission of Monetary Policy’, NBER Working Paper 5217. Stiglitz, J. (2000) ‘Capital Market Liberalization, Economic Growth, and Instability’, World Development 28 (6):1075–86. Summers, L. (1983) ‘The Non-Adjustment of Nominal Interest Rates: A Study of the Fisher Effect’, in J.Tobin (ed.), Macroeconomics, Prices, and Quantities, Washington, DC: The Brookings Institute. Summers, L. (1999) ‘The Right Kind of IMF for a Stable Global Financial System’, Speech to the London School of Business, London, December 14. Summers, L. (2000) Testimony before the International Monetary and Finance Committee, Washington DC, April 16. Summers, R. and Heston, A. (1995) Penn World Tables, Philadelphia, PA: University of Pennsylvania.

Bibliography

483

Swan, T. (1956) ‘Economic Growth and Capital Accumulation’, Economic Record 32(1):334–61. Tanzi, V. (1993) ‘Fiscal Deficit Measurement: Basic Issues’, in M.Blejer and A.Cheasty (eds), How to Measure the Fiscal Deficit, Washington, DC: IMF, pp. 13–21. Tarr, D. (1999) ‘Design of Tariff Policies for Russia’, in H.Broadman (ed.), Russian Trade Policy Reform for WTO Accession, Discussion Paper no. 401, Washington, DC: World Bank, pp. 7–29. Tatom, J.A. (1976) ‘The Welfare Cost of Inflation’, Federal Reserve Bank of St Louis, Review (November) pp. 9–22. Tatom, J.A. and Turley, J.E. (1978) ‘Inflation and Taxes: Disincentives for Capital Formation’, Federal Reserve Bank of St Louis, Review (January) pp. 2–8. Taussig, F. (1928) International Trade, New York: Macmillan. Taylor, J. (1979) ‘Staggered Wage Setting in a Macro Model’, American Economic Review 69 (1):108–13. Taylor, J. (1980) ‘Aggregate Dynamics and Staggered Contracts’, Journal of Political Economy 88 (1):1–24. Taylor, L. (1979) Macro Models for Developing Countries, New York: McGraw Hill. Taylor, L. (1981) ‘South-South Trade and Southern Growth: Bleak Prospects from the Structuralist Point of View’, Journal of International Economics 11 (2): 589– 602. Taylor, L. (1983) Structuralist Macroeconomics, New York: Basic Books. Taylor, L. (1985) ‘A Stagnationist Model of Economic Growth’, Cambridge Journal of Economics 9 (2):383–403. Taylor, L. (1987) ‘Macro Policy in the Tropics: How Sensible People Stand’, World Development 15 (12):1407–35. Taylor, L. (1988) Varieties of Stabilization Experience: Towards Sensible Macroeconomics in the Third World, New York: Oxford University Press. Taylor, L. (1990) Socially Relevant Policy Analysis, Cambridge, MA: MIT Press. Taylor, L. (1991) Income Distribution, Inflation and Growth, Cambridge, MA: MIT Press. Taylor, L. (1993) The Rocky Road to Reform, Cambridge, MA and London: MIT Press. Taylor, L. (1994) ‘Gap Models’, Journal of Development Economics 45 (1):17–34. Theil, H. (1961) Economic Forecasts and Policy, 2nd edn, Amsterdam: North Holland. Tinbergen, J. (1952) On the Theory of Economic Policy, Amsterdam: North Holland. Tobin, J. (1958) ‘Liquidity Preference as Behavior towards Risk’, Review of Economic Studies 36 (2):312–34. Tobin, J. (1965) ‘Money and Economic Growth’, Econometrica 33 (4):671–84. Tornell, A. and Lane, P. (1999) ‘The Voracity Effect’, American Economic Review 89 (1):22–46. Tower, E. (1971) ‘More on the Welfare Cost of Inflationary Finance’, Journal of Money, Credit, and Banking 3 (4):850–60. Trehan, B. and Walsh, C. (1991) ‘Testing Intertemporal Budget Constraints: Theory and Applications to the US Federal Budget and Current Account Deficits’, Journal of Money, Credit and Banking 23 (2):206–23. Uctum, M. and Wickens, M. (1997) ‘Debt and Deficit Ceilings, and Sustainability of Fiscal Policies: An Intertemporal Analysis’, Discussion Paper no. 1612, Centre for Economic Policy Research. UNCTAD (1999) UNCTAD Series on International Investment Agreements, New York and Geneva. UNCTAD (2000) The Least Developed Countries 2000 Report, Geneva.

484

Bibliography

United Nations (1999) World Investment Report 1999, New York. United Nations (2001) World Investment Report 2001, New York. US Treasury (2000) ‘Response to the Report of the International Financial Institution Advisory Commission’, Washington DC: US Treasury, June 8. Uzan, M. (1998) ‘The International Monetary System of the Twenty First Century: Creating a Financial Architecture as Modern as the Financial Markets’, mimeo, Reinventing Bretton Woods Committee. Van Wijnbergen, S. (1982) ‘Stagflationary Effects of Monetary Stabilization Policies’, Journal of Development Economics 10 (1):133–69. Van Wijnbergen, S. (1983) ‘Credit Policy, Inflation and Growth in a Financially Repressed Economy’, Journal of Development Economics 13 (1):45–65. Van Wijnbergen, S. (1985) ‘Macro-Economic Effects of Changes in Bank Interest Rates: Simulation Results for South Korea’, Journal of Development Economics 18 (3):541–54. Van Wijnbergen, S. (1986) ‘Exchange Rate Management and Stabilization Policies in Developing Countries’, Journal of Development Economics 23 (2):227–47. Van Wijnbergen, S. (1988) ‘Monopolistic Competition, Credibility and the Output Costs of Disinflationary Programmes’, Journal of Development Economics 29 (2): 375–98. Van Wijnbergen, S. (1991) ‘Fiscal Deficits, Exchange Rate Crises, and Inflation’, Review of Economic Studies 58 (1):81–92. Varman-Schendier, B. (1991) Capital Flight from Developing Countries, Boulder, CO and Oxford: Westview Press. Veidyanathan, G. (1993) ‘Consumption, Liquidity Constraints and Economic Development’, Journal of Macroeconomics 15 (Summer):591–610. Ventura, G. (1997) ‘Flat Tax Reform: A Quantitative Exploration’, University of Western Ontario Department of Economics Research Report 9706. Verrecchia, R. (1982) ‘The Use of Mathematical Models in Financial Accounting’, Journal of Accounting Research 20 (1):1–42. Walsh, C. (1995) ‘Optimal Contracts for Central Bankers’, American Economic Review 85 (1):150–67. Watson, M. (1988) ‘International Capital Markets: Developments and Prospects’, IMF World Economic and Financial Surveys Series, Washington, DC. Whalley, J. (1999) ‘Developing Countries in the Global Economy: A Forward-Looking View’, CSGR Working Paper 36/99, University of Warwick. Wilcox, D. (1989) ‘The Sustainability of Government Deficits: Implications of the Present Value Constraint’, Journal of Money, Credit and Banking 21 (2):291–306. Willett, T. (2001) ‘Restructuring IMF Facilities to Separate Lender of Last Resort and Conditionality Programmes: A Political Economy Perspective on the Meltzer Commission Recommendations’, Mimeo. Williamson, J. (1983) The Open Economy and the World Economy, New York: Basic Books. Williamson, J. (1985) ‘On the Question of Debt Relief’, Statement of the Roundtable on Money and Finance, New York, 13–14 December. Williamson, J. (1988) Voluntary Approaches to Debt Relief, Washington, DC: Institute for International Economics. Williamson, J. (1995) ‘Policy Reform in Latin America in the 1980s’, in D. Schydlowsky (ed.), Structural Adjustment: Retrospect and Prospect, Westport, CT and London: Greenwood, pp. 57–63. Williamson, J. (2000) ‘The Role of the IMF: A Guide to the Reports’, International Economics Policy Briefs, no. 00–5, Washington DC: Institute for International Economics, May.

Bibliography

485

Woolmer, R. (1977) ‘Monetary Policy in Nigeria’, Mimeo, University of Nigeria, Lagos. World Bank (1979) World Debt Tables, Washington, DC: World Bank. World Bank (1981) The Berg Report on LDC Debt Relief, Washington, DC: World Bank. World Bank (1984) World Bank Annual Report, Washington, DC: World Bank. World Bank (1986) World Bank Annual Report, Washington, DC: World Bank. World Bank (1991) World Bank Annual Report, Washington, DC: World Bank. World Bank (1998) World Bank Country Studies 1998, various issues. World Bank (1999) World Bank Country Studies 1999, various issues. World Bank (2000) Global Economic Prospects and the Developing Countries, Washington, DC: World Bank. World Bank (2001) World Development Indicators, Washington, DC: World Bank. Yashiv, E. (1989) ‘Optimal Inflation and the Government Revenue Mix’, Economics Letters 31 (3):151–4. Young, A. (1995) ‘The Tyranny of Numbers: Confronting the Statistical Reality of the East Asia Growth Experience’, Quarterly Journal of Economics 110 (3): 641– 80, August. Zarnowitz, V. and Lambros, L.A. (1987) ‘Consensus and Uncertainty in Economic Prediction’, Journal of Political Economy 95 (3):591–621. Zhang, A. (1999) ‘Testing for Moral Hazard in Emerging Market Lending’, IIF Discussion Paper, no. 99–1, Washington, DC: Institute for International Finance.

Name index

Agosin, M. 297 Ajayi, I. 444 Alexander, Sidney 145–6 Alfonsin, Raul 447 Aquino, Corazan 448 Arida, P. 457 Babangida, Ibrahim 447 Barro, R. 250–1, 254–8, 310, 327, 330–4 Baumol, William 86–7 Bayoumi, T. 365 Beck, T. 258 Benhabib, J. 392 Beseler, J. 405 Bhagwati, J. 393, 403 Bordo, M. 265, 267 Bosworth, B. 421, 425 Boyce, K. 445 Buffie, E. 301 Buiter, W. 288 Burnside, C. 355 Cagan, P. 80, 316–18 Calvo, G. 381 Camdessus, Michel 383 Cardoso, E. 306–8 Chalk, N. 297 Chipalkatti, N. 445 Collins, S. 372, 421, 425 Cooper, Richard 429 Corbo, V. 311 Corden, Max 296 De Haan, J. 323 Diamond, D. 271 Dicks-Mireaux, L. 266 Dollar, D. 258, 355 Dooley, M. 445

Dornbusch, R. 161–2, 439–40 Durlauf, S. 251 Dybvig, P. 271 Easterly, W. 251–2, 258, 301 Eckstein, Z. 322–3 Edwards, E. 381 Edwards, S. 364–5 Eichengreen, B. 449 Elbadawi, I. 365 Feldman, –233, 393–4, 403–4 Fischer, S. 258, 322 Fisher, I. 85 Flavin, M. 292 Foster, E. 322 Frankel, J. 374, 376 Garfinkel, M. 322 Gelb, A. 399 Gordon, D. 330–4 Graff, M. 388, 392–3 Granger, C. 77 Gray, J. 67 Greenwood, J. 390 Gregorio, J. 381 Griffith-Jones, A. 442–3 Gundlach, E. 255 Gupta, K.L. 77 Gylfason, T. 265 Hakkio, G. 294, 297 Hamilton, J. 292 Hansen, B. 27–8, 115 Haque, N. 183–9, 193–4, 207, 311 Harberger, A. 286, 296–7 Harrod, R. 233–4, 237 Hemming, N. 297

488

Name index

Hemphill, W. 346–7 Hicks, J. 27–8, 389 Hitler, Adolf 451 Hjertholm, P. 346–7 Hutchinson, M. 266, 276, 280 Jha, R. 358, 429, 434, 457 Johnson, Harry 145, 147 Jovanovic, B. 390 Kaminsky, G. 296 Kaplan, E. 378 Keefer, P. 258 Kendrick, D. 248 Kenny, C. 259 Keynes, J.M. 20–1, 29, 31, 85, 233, 451–2 Khan, M. 177, 267, 387, 391–2 Killick, T. 182 King, R. 258, 389 Knack, S. 258 Kouri, P. 157 Kremer, M. 258 Krueger, Anne 393, 403 Krugman, Paul 252, 265, 382 Kydland, F. 324 Lahiri, K. 183–9, 193–4, 207 Lane, P. 258 Leach, J. 322 Leff, N. 200 Leiderman, L. 322–3 Levine, R. 258, 387, 389 Lewis, Arthur 259 Loayza, N. 258 Londregan, J. 447 Lucas, R.E. 256, 260, 322, 324 McCallum, B. 322 McKinnon, R. 97, 195, 297, 388, 390 McPherson, M. 339–40 Mahalanobis, –229, 233, 393–4, 403–4 Mahathir, Mohammed 382 Mankiw, N. 251, 254–5 Marcos, Ferdinand 448 Mecagni, M. 266 Menem, Carlos 447 Montiel, P. 183–9, 193–4, 207 Moran, C. 347 Muhleisen, M. 311 Mundell, R. 106, 108 Mussa, M. 449

Nam, C. 405 Obstfeldt, M. 421 Palley, T. 90 Papadopoulos, A. 293 Pil, H. 297 Pollak, J. 177 Poole, K. 447 Porter, R. 195 Prachowny, M.J. 115 Prescott, E. 324 Pritchett, L. 258–9 Quah, D. 251 Ranis, G. 211 Ranney, S. 195 Rath, D. 90, 381 Rattso, J. 212, 219 Razin, O. 372 Reinhart, C. 296, 381 Ricardo, David 301 Rishi, M. 445 Rodrik, D. 258–9, 378 Rogoff, K. 326–7 Romer, D. 251, 317 Rush, M. 294, 297 Sachs, J. 258, 354–5 Saggar, M. 429, 434 Sala-I-Martin, X. 250–1, 254–5 Salter, W. 115 Sato, K. 200 Schadler, S. 266 Schatan, C. 358 Schmidt-Hebbel, K. 301, 311 Schwartz, A. 265, 267 Sen, A. 234 Senhadji, A. 387, 391–2 Sherden, W. 259 Sidiropoulos, M. 293 Solow, R. 229, 234, 237–40, 247–54 passim Soros, George 382 Spiegel, M. 392 Spratt, S. 442–3 Stein, J. 365 Stiglitz, Joseph 5, 189, 265–6, 276 Summers, L. 258 Swan, T. 234

Name index

Tanzi, V. 287 Taylor, L. 3, 177, 182–3, 195, 212, 219, 457 Tinbergen, Jan 106, 108 Tobin, J. 89 Tornell, A. 258 Trehan, B. 292 Uctum, M. 292 Valdes, R. 381 van Wijnbergen, Sweder 3 Ventura, G. 256–7 Videla, Jorge Rafael 447 Viola, Roberto 447

Walsh, C. 292, 330, 335, 337 Warner, A. 258 Well, D. 251 White, Harry Dexter 452–3 Wickens, M. 292 Wilcox, D. 292 Williams, A. 405 Williams, D. 259 Williamson, J. 157, 264 Woolmer, R. 77 Young, A. 256 Zervos, S. 258

489

Subject index

absorption see domestic absorption accelerationist hypothesis 72 Africa Reports of the World Bank (1981, 1984, 1986, 1991 and 1994) 351, 355 African Development Bank 356 agricultural sector 211–12, 400 aid flows 421–3 Angola 289, 356 anticipatory buying 149 Argentina 296, 373, 375, 447 asset effects 216–18 asset market balance (AMB) approach to exchange rates 162–71 assignment of policy to target: in developing countries 115–20; with fixed exchange rates 106–9; with flexible exchange rates 109–10 Augmented Physical Quality of Life Index 345 Australia 441 balance, internal and external, model of 137–8 balance of payments 13–15, 94–5, 108–10, 150–1; effect of IMF policies on 265–7; monetary approach to 4, 151–62 balance sheets 17–18 balance of trade 20 Bangladesh 356 Bank for International Settlements 441 banking system 76, 80–3; see also central banks barter 75, 84 Basel Capital Adequacy Accords 442–3 ‘basket peg’ regime (for exchange rates) 374

Bayer (company) 459 Berg Report (1981) 351 Bolivia 190 bond prices 30 Botswana 190, 286 Brandt Commission 459 Brazil 308 Bretton Woods conference 450, 453 budget deficits 167, 219, 402; forms of financing for 302–3 Burkina Faso 286, 346 Cambodia 350 capital account (of the balance of payments) 15, 150–1, 403 capital adequacy norms 442–3 capital controls 377–9, 382–3, 440, 448–9 capital flight 440–9; definition of 443–4; examples of 447–8; measurement of 444–5; prevention of 448–9 capital flows 248, 251, 255, 267, 304–6, 372, 377, 421–5; and monetary policy 379–82 capital market liberalisation 189 capital mobility 97–101, 104–5 capital stock 16, 239 central banks 323–6; credibility and reputation of 63, 327–34; and delegated responsibility for monetary policy 326–7; and Walsh contracts 335–8 Chile 296, 378, 449 China 250, 259, 378, 426, 445 Cobb-Douglas production function 42, 185, 235–6, 239–40, 244–5, 250–1 ‘cold shower’ effect 398

492

Subject index

Colombia 378 Comoros Islands 350 comparative static analysis 216, 219, 228–30 conditionality 269–70, 273, 281–4, 435; evolution of 273–5; structural 273–4 consumption 12 consumption function 21, 150 contracts of employment 66 convergence: ‘absolute’ and ‘conditional’ 252–3; of growth rates and of levels 251; in open economies 254–5; speed of 254–5;of technology and output per capita 249–51 convertibility 283 corruption 355, 395, 397 cost-of-living adjustments 67 Cote d’Ivoire 448 ‘coup trap’ 447 ‘crawling peg’ regime (for exchange rates) 374, 398 credit 158, 179–83 ‘crowding out’ of private investment 35, 62, 300 currency board policy 273, 373–5, 440 currency crises 438–40 Czechoslovakia 283 data quality 193 debt, external 5, 272, 343–4; relief from 346–53, 458; rescheduling of 288; sustainability of 295–8, 353–4 debt, internal 358, 398 debt, multilateral 437–8 debt-export ratio 353–4 debt-GDP ratio 292 deficit, external (in relation to fiscal) 298–303; see also budget deficits; fiscal deficit deflators 13 demand, aggregate 20; schedule of 40–2; stochastic version of 53; in structuralist model 195–8, 201; with variable prices 35–40 devaluation 4, 139, 398, 439, 458; absorption approach to 145–50; for developing countries 201, 205–7; elasticity approach to 140–5 discrete adjustment 110–15 disposable income 21 Doha meeting (2001) 459

dollarisation 373–4, 440 domestic absorption 20, 109, 145–50 dualistic models: of less developed economies 211–19, 416–17; medium-term 219–28; of output determination 230; short-term 212–19 dumping 404–5 dynamic inconsistency of policy 326 East Asian financial crisis 5, 272–4, 377, 419, 425, 427, 443, 446 ‘East Asian miracle’ 256–7 Economic Diversification Index 345 economic repression 82–3, 195, 403, 413; consequences of 393–5; costs of 397–9; facets of 395–7 Economist Intelligence Unit 445 Ecuador 190, 299 Egypt 299, 439 elasticities of supply and demand for exports and imports 144 ‘enclave’ economies 416 Engel effect 222 environmental issues 11, 358, 460 Equatorial Guinea 345, 350 equilibrium: long-run 45–9; multiple 269; simultaneous 31–3; in structuralist model 198–201 euro currency 375, 377 exchange controls 132–4 exchange rate risk 373, 403 exchange rates 94; adjustment policies for 139; asset market analysis of 162–73; choice of regime for 373–7, 420; for developing countries 372–7; ‘effective’ 367–8; equilibrium and misalignment in 363–6, 371–2; fixed or floating 374–6, 398, 451; modeling the dynamics of 367–72; policy for 118–20, 304; and random shocks 135–7; real 363–7; volatility in 161–2; see also macroeconomic policy expectations: adaptive 64–5, 71, 318; information and processes used in formation of 63–4; see also inflation: expectations of; rational expectations expenditure switching policies 139, 147 exports see growth, export-led;net exports

Subject index

factor price equalisation 257 financial assets 16–19; net 18 financial development in relation to economic growth 76–7, 83, 258, 387–93, 399 financial institutions, role of 75–6, 387–91 financial repression see economic repression ‘financially open’ and ‘financially insular’ economies 97 fiscal deficit: in developing countries 285–9; effects of financing of 306–8; sustainability of 289–95 fiscal policy 33–5, 62, 99–100, 103–10 passim, 203–5, 304, 398 Fisher equation and Fisher effect 37, 49 fixed coupon bonds 339 foreign balance (FB) constraint 93–7, 126–7 foreign direct investment (FDI) 286, 378, 421, 425–6 foreign exchange reserves 305, 439 ‘free rider’ behavior 268 free trade 190 G-7 summits 351, 357 Gambia 345 General Agreement of Tariffs and Trade (GATT) 405, 420, 435 General Data Dissemination Standards 9 Ghana 299 globalisation 251, 254 gold standard system 450–1 government expenditure 12, 187, 402; see also budget deficits gross domestic product and gross national product 11, 255 growth, economic: determinants of 252, 257–60; empirical research on 259; and exchange rate overvaluation or undervaluation 372; export-led 257, 414; forecasts of 259; in relation to financial development 76–7, 83, 258, 387–93, 399 growth accounting 238 growth experience, the, nature of 248–60 growth models 233; with capital and labor investment 240; convergence in 238–40; with human capital

493

244–8; neoclassical 234–8; for open economies 254, 257; technical progress in 254 heavily-indebted poor countries (HIPCs) 344–51 hedging against inflation 83 HIPC initiatives (1996 and 1999) 353; problems with 354–8 HIV/AIDS programs 357, 437, 459 Hong Kong 375, 460 ‘hot money’ cycle 189 human capital 392; accumulation of 255; in growth models 244–8 hyperinflation 289, 316–19 idle resources effect 146 import compression 347 import-substitution policies 416–17 ‘impossible trinity’ of exchange rate policy 376 Inada conditions 235 income: per capita 250, 254; world distribution of 256 increasing returns to scale 241 indexation of wages 67–9 India 229, 233, 286, 289, 296, 299, 311, 371, 377–82, 395, 425, 445, 457, 459 Indonesia 190 industrial sector 211–12 industrialisation 230, 394, 416–17 infant industries 405, 416 inflation 4, 301, 304–5; costs of 320–3; for developing countries 397–8; expectations of 45–7, 54–9, 67, 70–2, 83, 148–9, 315, 324–33; minimisation of 323; optimal rate of 322, 324, 326; optimal sustainable rate of 334; policy for control of 62–3, 398; trade-off with unemployment 69–72, 323–31; unanticipated 321, 397; zero rate of 331–4; see also hyperinflation inflation-adjusted deficit 309 inflation targeting 323, 338–9 inflation tax 319–21, 396 inflationary tendencies and inflationary bias 397, 402 information, costs of acquisition of 389–90 infrastructure development 258

494

Subject index

interest rates 27–30, 82, 300; real 62 International Bank for Reconstruction and Development 438, 449–50; see also World Bank international financial architecture 419–21 International Monetary Fund 4–5, 9, 158, 177–83 passim, 188–9, 193, 203, 206, 208, 252, 343–5, 350–8, 368, 378, 383, 413, 420, 437–40, 444–53, 458, 461; agenda for the reform of 429–35; and currency crises 276–80; criticisms of 182–3, 189, 265–6, 276, 434; decision-making procedures of 427–9; effects of programmes on output 265–6, 276, 280; executive directors and voting power 430–3; longer-term programmes 273; objectives of 427–8; overview of recent programmes 271–3; quota structures 428–9, 434; rationale for financing by 267–71 intertemporal budget constraints 290–3, 364 investment 12; see also foreign direct investment Iran 299 IS-LM analysis 27–33; applied to policy-making 33–5, 47; uncertainty in 51; variable prices in 36–7 J-curve 144–5, 368, 371 Japan 156–7 knowledge flows 255–7 labour market equilibrium 23–4, 65, 72, 122–3 labour unions see trade unions law of one price 153, 156–9, 182 Lawson doctrine 296 least-developed countries (LeDCs) 344–57 liberalisation programmes 398, 403, 413, 436, 458, 460; problems with 400–1 liquidity 389 liquidity preference 101 liquidity trap 34 LM schedule 29–34; for an open economy 121–2; with variable prices 37–40

long-run equilibrium 45–9 long-term funds 76 Luxembourg 375 macroeconomic policy: with fixed exchange rates 97–101, 105–6, 124; with flexible exchange rates 101–6, 125 Malawi 345 Malaysia 280, 378, 382–3, 440, 449 market imperfections 268–70 Marshall-Lerner condition 143, 147, 201 maturity transformation 76 Mauritius 286 Mexico 272–3, 296, 308, 377, 427, 446–7, 453 mixed portfolio-loan model 90 monetary contraction 203 monetary policy 33, 60–3, 87, 98–9, 103–10, 168, 195–6, 287, 305; delegated to the central bankers 326–7; instruments of 126–30; reputation models for design of 327–34; taking account of capital flows 379–82; time inconsistency in 323–6 monetary unions 374, 376 money 15–16; demand for 29–30, 83–90, 186, 315–16; models of creation of 77–82; purposes of 320 money illusion 148, 214 money multipliers 81, 90 money supply 4, 62; in developing countries 82–3; with fixed exchange rates 94–5 moral hazard 270–1, 434 Morgan Guaranty Trust Company 445 Mozambique 288 multinational corporations 194 multiplier analysis 22–4, 34, 81, 90, 99–101, 104 Naples terms 352–3 national accounts 9–13, 20 national debt 18 national income, equilibrium in 21–3, 27–8 natural rate: of output growth 72, 233; of unemployment 59–60, 72–3 neoclassical models 225–30, 234; of growth 234–8; for open economies 257

Subject index

Nepal 299 net exports 12, 146, 371 new classical model 62–3, 66 new development model 304 new Keynesian model 63–6 New Zealand 338 Nicaragua 340 Niger 346, 355 Nigeria 288, 345 non-performing assets 305 non-tariff barriers to trade 404–5, 413, 435 non-tradable goods 115–18 oil prices 3–4, 144, 193, 271, 343–4, 404 oligopoly 194, 230 open economies 251, 254, 257, 270; generalised model of 120–3; neoclassical analysis of 257 Opium Wars 190 optimum currency areas 375 Organisation of Petroleum Exporting Countries (OPEC) 144 output per capita 239–40, 249 overshooting 162, 171 ‘ownership’ of development programmes 275 Panama 375 Paraguay 295, 299 Paris Club 344–5, 352–3, 357 patents 459 Philippines, the 448 Phillips curve 69–72, 324 policy: in the generalised model 51–2; ineffectiveness of 59–60; reversals of 286; short-run effect of changes in 60–2 Ponzi games 289, 291–2 population, worldwide distribution of 249–50 portfolio choice model 90 portfolio investment 15, 286, 372, 421, 425 poverty and poverty-reduction programmes 352, 357–8, 457 price indices 13 privatisation 189 production function 42, 185, 234–6, 239–40, 244–5, 250–1 productivity of labor 65–6, 240

495

protectionism 194, 404–6, 413, 416 public goods, international 270–1, 442 purchasing power parity 364 rational expectations 55–9, 71, 333, 336 reform programs 395; international context of 403–6; sequencing of 401–3 rent-seeking 258 reputation models of monetary policy design 327–34 reserve assets 15 reserve requirements 82–3, 396 residence, definition of 15 ‘revolving door’ finance 445 Ricardian equivalence 296, 310–11 risk 76, 388–9; see also exchange rate risk Russia 189, 259, 296, 445–6 Samoa 350 savings 421, 425; mobilisation of 390–1 second best, theory of 400 seignorage revenue 315–19, 373; in highly-indebted countries 339–40 September 11th 2001, events of 459 services, transactions in 14 shocks 65–7, 117–20, 125–37, 258, 286, 372; and exchange rates 135–7; monetary 130–3, 161–2; to supply 49–53 Sidrauski monetary models 322–3 South Africa 286, 295, 299, 437 South Korea 378, 447, 460 Special Data Dissemination Standards 9, 441 Special Drawing Rights 428 speculative bubbles 448–9 speculative demand for money 30, 87–90 Sri Lanka 294–5, 299 stabilisation policies 4–5, 177–82, 188–9, 193, 203, 206–8, 457–8, 461 stagnating economies 256 statistical standards 441–2 steady-state budget constraint 307 steady-state equilibrium 242–3 steady-state growth path 252 sterilisation of capital flows 305–6, 381 structuralism 3, 5, 150, 182, 414–17, 458; applied to modeling of developing economies 193–207

496

Subject index

supply, aggregate: for an open economy 122–3; schedule of 42–5; in structuralist model 194–5, 202–3 Taiwan 460 target zone for an exchange rate 374 tariffs 406, 413, 416; on developing countries’ exports 410–12; distribution of peaks in 407–9 tax evasion 402 tax and inflation 321; see also inflation tax technical progress 234, 238, 389; exogenous 249; incorporated in growth models 254; and knowledge flows 255–7 terms of trade 3–4, 118, 123, 134, 147, 344, 372, 417, 436 Thailand 193, 299, 378, 427 time deposits 80–1 Toronto terms 351 Trade and Intellectual Property Rights (TRIPS) agreement 435 trade restrictions 254; see also tariffs; non-tariff barriers trade unions 194, 230 transactions demand for money 83–7 transition economies 274 Trinidad terms 352 Tunisia 299 Uganda 352, 357 uncovered interest parity 159, 161, 186

unemployment 13, 66–9, 194;natural rate of 59–60, 72–3;trade-off with inflation 69–72, 323–31 unions see trade unions United Kingdom 296, 451–2 United Nations Conference on Trade and Development (UNCTAD) 351–2, 406 United Nations System of National Accounts 9 United States 171, 415, 445, 451–2 Uruguay Round 406, 413, 420, 435, 459 vacancies 73 value added 10–11 ‘variable peg’ regime (for exchange rates) 374 wages: indexation of 67–9; nominal and real changes in 65–6; restraint in increase of 207 Walras’ law 29, 33, 116, 163–5 Walsh contracts 335–8 ‘warranted’ rate of growth 233 Washington consensus 252 World Bank 189, 252, 272, 343–5, 350–8, 420, 437–8, 441, 444, 449, 453; criticisms of 189–90 World Trade Organisation 190, 270, 405–6, 413, 420, 435, 437, 458–9 World War II 450–1 Zaire 340 Zambia 286, 294–5, 299, 355

Macroeconomics for Developing Countries

Economics of Developing Countries

Transport and Developing Countries

Regionalism Among Developing Countries

Wireless Internet Opportunity for Developing Countries

Fueling Development Energy Technologies for Developing Countries

Public Sector Economics For Developing Countries

Innovation Policy: A Guide for Developing Countries

SMEs in Asian Developing Countries

WTO Law and Developing Countries

Child Welfare in Developing Countries

Capacity Building in Developing Countries

Foreign Investments in Developing Countries

Medical Tourism in Developing Countries

Developing Countries in the WTO

SMEs in Asian Developing Countries

Public Transport in Developing Countries

Trade Policy in Developing Countries

Sustainable Cities in Developing Countries

Financial Reform in Developing Countries

Trade Policy in Developing Countries

Globalization and the Developing Countries

Marine biotechnology and developing countries

Electronic Business in Developing Countries

Krugman's Macroeconomics for AP*

Compact Cities: Sustainable Urban Forms for Developing Countries

Electronic Health Records: A Manual for Developing Countries

Macroeconomics for Today

Macroeconomics for Managers

Macroeconomics for managers

Macroeconomics for Developing Countries

Economics of Developing Countries

Transport and Developing Countries

Regionalism Among Developing Countries

Wireless Internet Opportunity for Developing Countries

Fueling Development Energy Technologies for Developing Countries

Public Sector Economics For Developing Countries

Innovation Policy: A Guide for Developing Countries

SMEs in Asian Developing Countries

WTO Law and Developing Countries

Child Welfare in Developing Countries

Capacity Building in Developing Countries

Foreign Investments in Developing Countries

Medical Tourism in Developing Countries

Developing Countries in the WTO

SMEs in Asian Developing Countries

Public Transport in Developing Countries

Trade Policy in Developing Countries

Sustainable Cities in Developing Countries

Financial Reform in Developing Countries

Trade Policy in Developing Countries

Globalization and the Developing Countries

Marine biotechnology and developing countries

Electronic Business in Developing Countries

Krugman's Macroeconomics for AP*

Compact Cities: Sustainable Urban Forms for Developing Countries

Electronic Health Records: A Manual for Developing Countries

Macroeconomics for Today

Macroeconomics for Managers

Macroeconomics for managers

Recommend Documents