Beyond Inflation Targeting

Beyond Infl ation Targeting

Beyond Infl ation TargetingAssessing the Impacts and Policy Alternatives

Edited by

Gerald A. EpsteinPolitical Economy Research Institute (PERI), University of Massachusetts Amherst, USA

A. Erinç YeldanBilkent University, Ankara, Turkey and International Development Economics Associates (IDEAs), India

Edward ElgarCheltenham, UK • Northampton, MA, USA

© Gerald A. Epstein and A. Erinç Yeldan 2009

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher.

Published byEdward Elgar Publishing LimitedThe Lypiatts15 Lansdown RoadCheltenhamGlos GL50 2JAUK

Edward Elgar Publishing, Inc.William Pratt House9 Dewey CourtNorthamptonMassachusetts 01060USA

A catalogue record for this bookis available from the British Library

Library of Congress Control Number: 2009933402

ISBN 978 1 84720 938 2

Printed and bound by MPG Books Group, UK

v

Contents

List of contributors viiList of abbreviations xiiPreface xiii

PART I INTRODUCTION AND THEORETICAL FRAMEWORKS

1 Beyond infl ation targeting: assessing the impacts and policy alternatives 3

Gerald Epstein and A. Erinç Yeldan 2 Real exchange rate, monetary policy and employment:

economic development in a garden of forking paths 28 Roberto Frenkel and Lance Taylor 3 Infl ation targeting and the real exchange rate in a small

economy: a structuralist approach 44 Jose Antonio Cordero

PART II THEMATIC ISSUES: CLASS RELATIONS AND GENDER IMPACTS OF INFLATION TARGETING

4 Income, class and preferences towards anti-infl ation and anti-unemployment policies 71

Arjun Jayadev 5 The gendered political economy of infl ation targeting:

assessing its impacts on employment 93 Elissa Braunstein and James Heintz 6 Infl ation and economic growth: a cross-country non-linear

analysis 116 Robert Pollin and Andong Zhu

PART III INFLATION TARGETING: CRITIQUES AND COUNTRY-SPECIFIC ALTERNATIVES

7 Infl ation targeting in Brazil: 1999–2006 139 Nelson H. Barbosa-Filho

vi Beyond infl ation targeting

8 Alternatives to infl ation targeting in Mexico 158 Luis Miguel Galindo and Jaime Ros 9 Five years of competitive and stable real exchange rate in

Argentina, 2002–07 179 Roberto Frenkel and Martín Rapetti10 A general equilibrium assessment of twin-targeting in Turkey 203 Cagatay Telli, Ebru Voyvoda and A. Erinç Yeldan11 Employment targeting central bank policy: a policy proposal

for South Africa 227 Gerald Epstein12 Infl ation targeting and the design of monetary policy in India 248 Raghbendra Jha13 Towards an alternative monetary policy in the Philippines 271 Joseph Anthony Lim14 Monetary policy in Vietnam: alternatives to infl ation targeting 299 Le Anh Tu Packard

Index 315

vii

Contributors

Nelson H. Barbosa-Filho is Professor of Macroeconomics and Public Finance at the Institute of Economics of the Federal University of Rio de Janeiro, Brazil and holds a PhD in economics from the New School for Social Research. Since 2003, Barbosa has been working in President Lula’s Administration and he is currently the Secretary of Economic Policy at the Brazilian Ministry of Finance.

Elissa Braunstein is Assistant Professor of Economics at Colorado State University, Fort Collins, Colorado. She received her PhD in economics from the University of Massachusetts Amherst and a Master of Pacifi c International Aff airs from the School of International Relations and Pacifi c Studies at the University of California San Diego. Prior to joining CSU, Braunstein was an assistant research professor at the Political Economy Research Institute (PERI) at the University of Massachusetts Amherst. Her research focuses on feminist analyses of international eco-nomics, including the role of foreign direct investment in development. Recent work involves analysing the impact of foreign investment on gen-der-based wage inequality in mainland China, the gender-diff erentiated employment impacts of defl ationary monetary policy in developing coun-tries, the linkages between trade liberalization and reproductive health and rights, and the impact of US state tax policy on women’s employment. Among her recent publications are Trading Women’s Health & Rights? Trade Liberalization and Reproductive Health in Developing Countries (Zed Books, 2006, co-edited with Caren Grown and Anju Malhotra).

Jose Antonio Cordero is Professor of Economics at the Escuela de Economia, Universidad de Costa Rica, San Jose, Costa Rica, and Center for Economic and Policy Research (CEPR). Portions of his chapter were written during an academic year as visiting professor at Mount Holyoke College. Cordero received his PhD from the University of Notre Dame in 1995. His research focuses on the macroeconomics of growth and distribution in open-mone-tary economies, and on the eff ects of FDI on economic development.

Gerald Epstein is Professor of Economics and founding Co-Director of the Political Economy Research Institute (PERI) at the University of Massachusetts Amherst. His major research and teaching areas are the

viii Beyond infl ation targeting

political economy of central banking, the political economy of inter-national fi nance and egalitarian macroeconomic policy reform. He has published widely in these areas. He has also served as a consultant to the United Nations Development Programme (UNDP), International Labor Organization (ILO), United Nations Conference on Trade and Development (UNCTAD) and is a staff economist with the Center for Popular Economics (CPE) in Amherst, Massachusetts.

Roberto Frenkel is Principal Research Associate at CEDES (since 1977) and Professor at the University of Buenos Aires, Argentina (since 1984). Presently he is also Director of the Graduate Program on Capital Markets, University of Buenos Aires and teaches graduate courses at the Di Tella and FLACSO-San Andrés Universities in Argentina and the University of Pavia in Italy. He is a member of the UNDP Advisors Group, member of the Board of the World Institute for Development Economic Research (WIDER), United Nations University and member of the Academic Council of CEFID-AR. He has published numerous books and articles in academic journals on macroeconomic theory and policy, money and fi nance, infl ation and stabilization policies, and labor market and income distribution, with special focus on Argentina and Latin America.

Luis Miguel Galindo is Professor of Economics at the Faculty of Economics at the National Autonomous University of Mexico in Mexico City. He received his PhD in economics from Newcastle University in Newcastle upon Tyne in the UK. His main research interests are monetary economics and environmental economics. His recent research is concentrated in the transmission mechanism of monetary policy in emerging economics and the economics of climate change. He has worked as an advisor to several fi nancial institutions and international organizations.

James Heintz, an Associate Research Professor and Associate Director of the Political Economy Research Institute (PERI at the University of Massachusetts, Amherst), holds a PhD from the University of Massachusetts and a Master’s degree from the University of Minnesota. He has written on a wide range of economic policy issues, including job creation, global labor standards, egalitarian macroeconomic strategies and investment behavior. He has worked as an international consultant on projects in Africa sponsored by the International Labor Organization (ILO) and the United Nations Development Program (UNDP), that focus on employment-oriented development policy. From 1996 to 1998 he worked as an economist at the National Labour and Economic Development Institute in Johannesburg , a policy think tank affi liated with the South African labor movement.

Contributors ix

Arjun Jayadev is Assistant Professor of Economics at the University of Massachusetts Boston. He has also been a research fellow at Columbia University’s Committee on Global Thought. His work focuses on development and inequality. He received his PhD from the University of Massachusetts Amherst in 2005.

Raghbendra Jha is Professor of Economics and Executive Director of the Australia South Asia Research Centre, Arndt-Corden Division of Economics, Australian National University, Canberra. He obtained his PhD in economics in 1978 from Columbia University, New York, and has previously taught at Columbia University and Williams College in the US, Queen’s University in Canada, the University of Warwick in the UK, and the Delhi School of Economics and Indira Gandhi Institute of Development Research in India. His work focuses on macroeconomics and public economics.

Joseph Anthony Lim is currently a Professor in the Economics Department of the Ateneo de Manila University, the Philippines. He was also pro-fessor in economics at the University of the Philippines from 1978 to 2004. He was a Rockefeller scholar and his PhD and Master of Science graduate degrees were obtained from the University of Pennsylvania and Massachusetts Institute of Technology, respectively. His fi elds of expertise are macroeconomics and development economics. His research works include analyses of the Asian crises, growth analyses and diagnostics of the Philippine macroeconomy, the debt burden and its adverse impact on social and economic spending of developing countries, and the gender aspects of employment generation. He was a macroeconomics, fi nance and debt advisor to the Bureau for Development Policy (BDP) division of the United Nations Development Programme (UNDP) New York head-quarters from 2002 to 2004. He is currently a board member of the Journal of the Asia Pacifi c Economy published by Francis and Taylor.

Le Anh Tu Packard is Senior Economist at Moody’s Economy.com, a divi-sion of Moody’s Analytics. She is also a Research Fellow and Convener of the Research and Study Group on Vietnamese Social and Economic Reform, Center for Vietnamese Philosophy, Culture and Society, Temple University, Philadelphia, PA, US, and has spent many years helping the Vietnamese government on various aspects of economic reform. Her current research focus is on the international business cycle, trade and capital fl ows, and fi nancial crises.

Robert Pollin is Professor of Economics and founding Co-Director of the Political Economy Research Institute at the University of Massachusetts Amherst. He received a PhD in economics from the New School for Social

x Beyond infl ation targeting

Research in New York City in 1982. His research centers on macroeco-nomics, conditions for low-wage workers in the US and globally, and the analysis of fi nancial markets. Has also written and consulted in the area of labor market policies, including the viability of ‘living wage’ policies in the US. He has directed United Nations Development Programme sponsored projects on employment targeted macroeconomic policies for South Africa and Kenya, and most recently has focused his research on the analysis of job creation in the US in the transition to a renewable energy based economy.

Martín Rapetti is Assistant Researcher at CEDES and a PhD student at the University of Massachusetts Amherst. His fi eld of specialization is macroeconomics for developing countries and his current research focuses on the eff ects of competitive real exchange rates on economic development.

Jaime Ros is Professor of Economics at the University of Notre Dame, Notre Dame, Indiana and Fellow of the Kellogg Institute for International Studies. Ros specializes in development economics with special reference to Latin America. His most recent book is Development Theory and the Economics of Growth (University of Michigan Press, 2000). His articles have appeared in the Cambridge Journal of Economics, World Development, Journal of Development Studies, The Manchester School of Economics and Social Studies, El Trimestre Economico, Desarrollo Economico and other scholarly journals and edited books. Current projects include an edited handbook with Amitava Dutt on development economics as well as a book on the history of Mexico’s economic development with Juan Carlos Moreno Brid.

Lance Taylor is the Arnhold Professor of International Cooperation and Development at the New School University, New York. He received a PhD in economics from Harvard University in 1968. He has been Professor in the economics departments of Harvard and the Massachusetts Institute of Technology, as well as a visiting professor at the University of Minnesota, the Universidade da Brasilia, Delhi University and the Stockholm School of Economics. He moved to the New School for Social Research in 1993. Taylor has published widely in the areas of macroeconomics, develop-ment economics and economic theory. He has served as a visiting scholar or policy advisor in over 25 countries, including Chile, Brazil, Mexico, Nicaragua, Cuba, Russia, Egypt, Tanzania, Zimbabwe, South Africa, Pakistan, India and Thailand.

Cagatay Telli is a Researcher at the State Planning Organization, Ankara, Turkey. He is a graduate of Bilkent University and his areas of expertise are on general equilibrium modelling and computational methods.

Contributors xi

Ebru Voyvoda is Assistant Professor of Economics at the Department of Economics, Middle East Technical University, Ankara, Turkey. She holds a PhD from Bilkent University. Her areas of specialization include general equilibrium modeling, economics of productivity growth and technical change, and environmental economics.

A. Erinç Yeldan is Professor of Economics at Bilkent University, Ankara, Turkey. He was a Fulbright Scholar and Visiting Professor at the University of Massachusetts Amherst and Amherst College while this book was being written. He is also one of the directors of the International Development Economics Associates (IDEAs), New Delhi. Yeldan received his PhD from the University of Minnesota in 1988. His recent work focuses on development macroeconomics and on applied general equilibrium models with emphasis on the Turkish economy.

Andong Zhu is Assistant Professor of Economics at Tsinghua University, Beijing, China. He is a Research Associate of PERI at the University of Massachusetts Amherst. Zhu received his PhD from the University of Massachussetts Amherst in 2005. His recent work focuses on political economy and the imbalances in the world economy and China.

xii

Abbreviations

BCB Brazilian Central BankCBRT Central Bank of TurkeyCGE computable general equilibriumCMN National Monetary CouncilCPE Center for Popular EconomicsCPI consumer price indexET employment targetingFX foreign exchangeGEAR growth, employment and redistributionILO International Labour OrganizationIMF International Monetary FundISSP International Social Survey ProgramIT infl ation targetingMDG Millennium Development GoalsMLE medium to large enterprisePERI Political Economy Research InstitutePPP purchasing power parityRER real exchange rateRBI Reserve Bank of IndiaSAPs structural adjustment programsSBV State Bank of VietnamSCRER stable and competitive real exchange rateSME small- to medium-sized enterpriseSOCB state-owned commercial bankSOE state-owned enterpriseUNCTAD United Nations Conference on Trade and DevelopmentUNDP United Nations Development ProgrammeVAR vector auto-regressionWTO World Trade Organization

xiii

Preface

The chapters in this edited volume report on the results of a three year international research project designed to evaluate the impacts of the infl a-tion targeting approach to central banking which has swept the global eco-nomic landscape in the last decade or so. Moving beyond critique, though, our main motivation for this project has been to develop socially useful alternatives to infl ation targeting. This task has become only more crucial over the years since we began the project, because, as we describe more fully in the following pages, we believe infl ation targeting has largely had negative consequences for economic development, equality and poverty reduction, especially in the developing world, even as more and more countries adopt this fl awed monetary policy regime. This has become especially apparent as a global fi nancial crisis engulfs much of the world while many central banks have been slow to respond, partly due to their obsession with keeping commodity infl ation in the low single digits.

In structuring the project, we have been anxious to avoid the one size fi ts all approach that underlies not only infl ation targeting itself but also much of the policy approach of the neoliberal ‘Washington Consensus’ of which infl ation targeting has become a key part. To that end, the authors of the country study chapters were tasked with designing concrete, country- specifi c alternatives to infl ation targeting. In addition, other authors have written studies of long neglected themes in central banking, including the gender impacts and class aspects of monetary policy. While most of the chapters are historical and empirical in nature, a few lay out a basic mac-roeconomic modeling framework to help us understand the key issues at stake.

Initiated and directed by Gerald Epstein of the Political Economy Research Institute (PERI) and Economics Department of the University of Massachusetts and Erinç Yeldan of Bilkent University in Turkey, this ‘Alternatives to Infl ation Targeting project’ has benefi ted from the numerous presentations and discussions we have had over these several years in various parts of the world with many economists, activists and practitioners. More generally, a multi-author, multi-country project such as ours requires many other things to be successful, including, fi rst and foremost, creative, skilled and dedicated researchers, which we are fortu-nate to have found. The work of many of them appears in these pages.1 In

xiv Beyond infl ation targeting

addition it takes fi nancial resources to mount a project such as this, and we thank the Ford Foundation, the Rockefeller Brothers Fund, the United Nations Department of Economic and Social Aff airs (UN-DESA) and PERI for fi nancial support. The project has also benefi ted from the help and encouragement of many people and we especially thank K.S. Jomo of UN-DESA, Manuel Montes and Michael Conroy previously of the Ford Foundation and Rockefeller Brothers Fund respectively, and Jo-Marie Greisgraber and Jamie Baker of New Rules for Global Finance. We are also grateful to Roberto Frenkel and CEDES for hosting an important project conference in Buenos Aires. Thanks are also due to our graduate students, Hasan Comert and Luis Rosero, who have contributed fi rst rate research assistance, and to Judy Fogg of PERI who has provided untold, key support along the way. Finally, we are indebted to Alan Sturmer, Acquisitions Editor at Edward Elgar Publishing, for his help, support and encouragement along the way.

NOTE

1. Many of these chapters appeared in shorter form in a Special Issue of the International Review of Applied Economics (IRAE), March 2008: ‘Infl ation targeting, employment creation and economic development: assessing the impacts and policy alternatives’. Thanks very much to Malcolm Sawyer, editor of the IRAE, for his help with that issue and his cooperation with this book project as well.

PART I

Introduction and theoretical frameworks

3

1. Beyond infl ation targeting: assessing the impacts and policy alternativesGerald Epstein and A. Erinç Yeldan1

1.1 INTRODUCTION

Infl ation targeting (IT) is the new orthodoxy of mainstream macroeco-nomic thought. The approach has now been adopted by 24 central banks, and many more, including those in developing countries, are expressing serious interest in following suit. Initially adopted by New Zealand in 1990, the norms surrounding the IT regime have been so powerful that the central banks of both the industrialized and the developing economies alike have declared that maintaining price stability at the lowest possible rate of infl ation is their only mandate. It was generally believed that price stability is a pre-condition for sustained growth and employment, and that ‘high’ infl ation is damaging the economy in the long run.

In broad terms, the IT policy framework involves ‘the public announce-ment of infl ation targets, coupled with a credible and accountable com-mitment on the part of government policy authorities to the achievement of these targets’ (Setterfi eld, 2006, p. 653). As advocated, ‘full fl edged’ infl ation targeting consists of fi ve components: absence of other nominal anchors, such as exchange rates or nominal GDP; an institutional com-mitment to price stability; absence of fi scal dominance; policy (instrument) independence; and policy transparency and accountability (Bernanke et al., 1999; Mishkin and Schmidt-Hebbel, 2001, p. 3). In practice, while few central banks reach the ‘ideal’ of being ‘full fl edged’ infl ation targeters, many others still focus on fi ghting infl ation to the virtual exclusion of other goals.

For its proponents, the appropriate infl ation target is typically pre-scribed as maintaining price stability, though there is less agreement on the meaning of this term and on its precise measurement. Many prac-titioners simply adopt the widely cited defi nition of Alan Greenspan, the former Governor of the US Fed, issued at the July 1996 meeting of the Federal Open Market Committee, as ‘a rate of infl ation that is suffi ciently low that households and businesses do not have to take it into account in

4 Beyond infl ation targeting

making every day decisions’. For Feldstein (1997), however, price stability meant a long-run infl ation rate of zero. In addition, IT is usually associ-ated with appropriate changes in the central bank law that enhances the independence of the institution (Bernanke et al., 1999, p. 102; Mishkin and Schmidt-Hebbel, 2001, p. 8; see also Buiter, 2006 for an evaluation). Note that this promotion of central bank independence often is inconsistent with the above mentioned commitment to accountability, if by account-ability, one means democratic accountability.

Ironically, employment creation has dropped off the direct agenda of most central banks just as the problems of global unemployment, under-employment and poverty are taking center stage as critical world issues (Heintz, 2006). The International Labour Organization (ILO) estimates that in 2003 approximately 186 million people were jobless, the highest level ever recorded (ILO, 2004a). The employment to population ratio – a measure of unemployment – has fallen in the last decade, from 63.3 percent to 62.5 percent (ILO, 2004). And as the quantity of jobs relative to need has fallen, there is also a signifi cant global problem with respect to the quality of jobs. The ILO estimates that 22 percent of the developing world’s workers earn less than $1 a day and 1.4 billion (or 57 percent of the devel-oping world’s workers) earn less than $2 a day. To reach the Millennium Development Goal of halving the share of working poor by 2015, sus-tained, robust economic growth will be required. The ILO estimates that on average, real GDP growth has to be maintained at 4.7 percent per year to reduce the share of $1 a day poverty by half by 2015, and signifi cantly more than that to reduce the share of $2 a day poverty by half.

Moreover, China’s and India’s opening up to the global markets and the collapse of the Soviet system together have added 1.5 billion new workers to the world’s economically active population (Freeman, 2004, 2005; Akyuz, 2006). This means almost a doubling of the global labor force and a reduction of the global capital-labor ratio by half. Concomitant with the emergence of the developing countries in the global manufacturing trade, about 90 percent of the labor employed in world merchandise trade is low skilled and unskilled, suff ering from marginalization and all too frequent violation of basic worker rights in informalized markets (see, for example, Akyuz, 2003, 2006; Akyuz et al., 2006).

Under these conditions, a large number of developing countries have suff ered de-industrialization, serious informalization and consequent worsening of the position of wage-labor, resulting in a deterioration of income distribution and increased poverty. Many of these phenomena have occurred in tandem with the onset of neoliberal conditionalities2 imposing rapid liberalization of trade and premature deregulation of the indigenous fi nancial markets.

Impacts and policy alternatives 5

The key problem is that the ongoing ‘fi nancial globalization’ appears primarily to redistribute shrinking investment funds and limited jobs across countries, rather than to accelerate capital accumulation across a global scale (Adelman and Yeldan, 2000; Akyuz, 2006). Simply put, the world economy is growing too slowly to generate suffi cient jobs and it is allocating a smaller proportion of its income to fi xed capital formation. In addition, asset price bubbles and crashes, with their attendant fi nancial fall-out are plaguing the system. Under these conditions, it ought to be clear that price stability, on its own, will not suffi ce to maintain macro-economic stability, as it cannot suffi ce to secure fi nancial stability and employment growth. In the words of Akyuz (2006, p. 46), ‘the source of macroeconomic instability now is not instability in product markets but asset markets, and the main challenge for policy makers is not infl ation, but unemployment and fi nancial instability’ (emphasis added).

Yet, surprisingly, despite a disappointing record, this almost single minded focus on commodity infl ation is gaining a more secure foothold in monetary policy circles and the circles are widening to include an increasing number of developing countries. According to a recent report by the International Monetary Fund (IMF), an increasing number of central banks in emerging markets are planning to adopt IT as their operating framework (Batini et al., 2006; Table 1.1). An IMF staff survey of 88 non-industrial countries found that more than half expressed a desire to move to explicit or implicit quantitative infl ation targets. More relevant to our concerns, nearly three-quarters of these countries expressed an interest in moving to ‘full-fl edged’ IT by 2010. To support and encourage this movement, the IMF is provid-ing technical assistance to many of these countries and is willing to provide more (Table 1.1 and further discussion below). In addition, the IMF is considering altering its conditionality and monitoring structures to include infl ation targets. In short, despite little evidence concerning the success of IT in its promotion of economic growth, employment creation and poverty reduction, and mixed evidence at best that it actually reduces infl ation itself, a substantial momentum is building up for full-fl edged IT in developing countries. Promotion eff orts by the IMF and Western-trained economists are at least partly responsible for this increasing popularity.

While it might seem obvious that stabilization focused monetary policy represents the only proper role for central banks, in fact looking at history casts serious doubt on this claim. Far from being the historical norm, in many of the successful currently developed countries, as well as in many developing countries in the post-Second World War period, pursuing development objectives was seen as a crucial part of the central banks’ tasks (Epstein, 2007). Now, by contrast, development has dropped off the policy agenda of central banks in most developing countries.


Table 1.1 Infl ation targeting countries: initial conditions and modalities

Developing countries(in order of adoption)

IT adoption

date

Infl ation rate at start

(% per annum)

Current infl ation target(% per annum)

Offi cially declared policy instrument

Israel 1997Q2 8.5 1–3 headline O/N rate

Czech Rep. 1998Q1 13.1 3 (1/–1) 2 week repoPoland 1998Q4 9.9 2.5 (1/–1) 28 day

interventionBrazil 1999Q2 3.3 4.5 (1/–2) selic O/N rateChile 1999Q3 2.9 2–4 O/N rateColombia 1999Q3 9.3 5 (1/–0.5) repoSouth Africa 2000Q1 2.3 3–6Thailand 2000Q2 1.7 0–3.5 14 day repoKorea 2001Q1 3.2 2.5–3.5 O/N call rateMexico 2001Q1 8.1 3 (1/–1) 91-day CetesHungary 2001Q2 10.5 3.5 (1/–1) 2 week depositPeru 2002Q1 –0.8 2.5 (1/–1)The Philippines 2002Q1 3.8 5–6 reverse repoSlovak Republic 2005Q1 3.2 3.5 (1/–1)Indonesia 2005Q3 7.8 5.5 (1/–1) 1-month SBIRomania 2005Q3 8.8 7.5 (1/–1)Turkeya 2006Q1 7.8 5 (1/–2) CB O/N rateTurkeyb 2001Q2 82.0 n.a CB net

domestic assets

Industrial CountriesNew Zealand 1990Q1 7.0 1–3 cash rateCanada 1991Q1 6.2 1–3 O/N funding

rateUnited Kingdom 1992Q4 3.6 2 repoSweden 1993Q1 4.8 2 (1/–1) repoAustralia 1993Q2 1.9 2–3 cash rateIceland 2001Q1 3.9 2.5Norway 2001Q1 3.7 2.5

Candidate CountriesCosta Rica, Egypt, Ukraine

Near Term (1–2 years)

Albania, Armenia, Botswana, Dominican

Medium Term (3–5

years)


The theme of this book is that modern central banking ought to have more policy space in balancing out various objectives and instruments. In particular, employment creation, poverty reduction and more rapid economic growth should join infl ation stabilization and stabilization more generally as key goals of central bank policy. In introducing this volume, this chapters outlines why a shift away from IT, the increasingly fashion-able, but problematic approach to central bank policies, and a move toward a more balanced approach is both desirable and feasible.

The rest of the chapter is organized as follows. In the next section, we briefl y survey the macroeconomic record of IT and its current structure. Section 1.3 focuses on the role of the exchange rate as one of the key macro prices, and discusses alternative theories of its determination. We also make remarks on the issue of IT in the context of the so-called ‘trilemma’ of monetary policy. In Section 1.4 we discuss various alternatives to infl a-tion focused central banks, concentrating on the results of a multi-country research project undertaken with the support of UN-DESA, among other organizations. This section shows that there are viable, socially productive

Table 1.1 (continued)

Rep., Gutemala, Mauritius, Uganda, Angola, Azerbaijan, Georgia, Moldova, Serbia, Sri Lanka, Vietnam, Zambia

Medium Term

(3–5 years)

Belarus, China, Kenya, Kyrgyz Rep., Moldova, Serbia, Sri Lanka, Vietnam, Zambia Bolivia, Honduras, Nigeria, Papua New Guinea, Sudan, Tunisia, Uruguay, Venezuela

Long Term (. 5 years)

Notes:a. Offi cial adoption date for Turkey.b. Turkish central bank declared ‘disguised infl ation targeting’ in the aftermath of the 2001

February crisis.

Source: Batini et al. (2006).


alternatives to IT, including those that focus on employment generation, and makes the case that these alternatives should be further developed. Section 1.5 concludes the chapter.

1.2 MACROECONOMIC RECORD OF INFLATION TARGETING

Much of the existing literature on the record of IT has focused mostly on whether systemic risks and accompanying volatility has been reduced in the IT economies, and whether infl ation has come down actually in response to adoption of the framework itself or due to a set of ‘exogenously welcome’ factors. On the one side, there is a fair amount of agreement that IT has been associated with reductions in infl ation. Furthermore, exchange rate pass-through eff ects were reportedly reduced and consumer prices have become less prone to shocks (Edwards, 2005; Mishkin and Schmidt-Hebbel, 2001). Yet existing evidence also suggests that IT has not yielded infl ation below the levels attained by the industrial non-targeters that have adopted other monetary regimes (Ball and Sheridan, 2003; Mishkin and Scmidt-Hebbel, 2001; Roger and Stone, 2005). Moreover, even if domestic monetary policy has reduced infl ation, the hoped for gains in economic growth and employment have, generally, not materialized.

On the ‘qualitative’ policy front, it is generally argued that with the onset of central bank independence, communication and transparency have improved and that the central banks have become more ‘account-able’. Yet in practice, ‘central bank independence’ means that central banks have become less accountable to their governments, and, arguably, more accountable to fi nancial elites and international organizations such as the IMF.

Moreover, little is known about the true costs of IT on potential output growth, employment, and on incidence of poverty and income distri-bution. Bernanke et al. (1999) and Epstein (2007), for instance, report evidence that IT central banks do not reduce infl ation at any lower cost than other countries’ central banks in terms of foregone output. That is, IT does not appear to increase the credibility of central bank policy and therefore does not appear to reduce the sacrifi ce ratio. Per contra, based on an econometric study of a large sample of infl ation targeters and non-targeters, Corbo et al. (2001) concluded that sacrifi ce ratios have declined in the emerging market economies after adoption of IT. They also report that output volatility has fallen in both emerging and industrialized economies after adopting IT. This position is recently complemented by a study of the IMF economists, who, using a complex econometric model


and policy simulations, report fi ndings that infl ation targeting economies experience reductions in the volatility in infl ation, without experiencing increased volatility in real variables such as real GDP (Batini et al., 2006). According to these estimates, IT central banks do enhance economic ‘sta-bility’ relative to other monetary rules, such as pegged exchange rates and monetary rules.

However, in the assessments of ‘stability’, these papers do not consider the issue of the stability of asset prices, including exchange rates, stock prices and other fi nancial asset prices. As we discuss further below, asset price stability may need to be included in a full analysis of the impact of IT on overall economic stability. Asset price stability aside, while intriguing, these results are only as strong as the simulation model on which they are based and are only as relevant as the relevance of the questions they pose. Moreover, they are only as broad as the alternatives they explore. On all these scores, these results are problematic. First, they do not simulate the impact of IT relative to other possible policy regimes, such as targeting the real exchange rate as discussed below. Second, the model is based on esti-mates of potential output that are themselves aff ected by monetary policy (see, for example, Michl, 2007; Tobin, 1980). Hence, if monetary policy slows economic growth, it also lowers the rate of growth of potential output and, therefore, reduces the gap between the two, thereby appearing to stabilize the economy.

Equally, if not more important, is the practical problematique of setting the targeted rate of infl ation itself. Even if the advocated requisites of the IT regime are taken for granted, it is not yet clear what the practically targeted rate of infl ation should be. Even though there appears to be a consensus among the advocates of the IT regime that the infl ation target has to be ‘as low as possible’, there is no theoretical justifi cation of this assertion; and as such, it sounds more of a recommendation than a careful calculation. Most disturbing is the common belief that what is good for the industrialized/developed market economies should simply be repli-cated by the developing countries as well. That this may not be the case is forcefully argued in Pollin and Zhu (2006). Based on their non-linear regression estimates of the relationship between infl ation and economic growth for 80 countries over the period 1961–2000, Pollin and Zhu report that higher infl ation is associated with moderate gains in GDP growth up to a roughly 15–18 percent infl ation threshold. Furthermore they report that there is no justifi cation for IT policies as they are currently being practiced throughout the middle- and low-income countries, that is, to maintain infl ation with a 2–4 percent band.

Moreover, we have other evidence on the negative consequences of monetary policy designed to produce extremely low levels of infl ation in


developing countries. Braunstein and Heintz show that contractionary monetary policy used to fi ght infl ation often has a diff erentially negative impact on the employment rates of women relative to men (Braunstein and Heintz, Chapter 5, this volume). Given the possible negative costs of IT on output and employment, there should be some direct survey evidence indicating people’s preferences with respect to infl ation and unemployment. While some studies have indicated that people have an absolute preference for low infl ation, Arjun Jayadev (Chapter 4, this volume) reports on survey results asking people in diff erent countries and income levels what is their bigger concern, high infl ation or high unemployment. His main result is that, in his sample, poorer people are concerned more about high unemployment than high infl ation, while richer respondents have the opposite preferences. Hence, concerns over employment and infl ation probably have an important class dimension to them, something that economists and historians have suspected for many years.

Finally, is the issue of the role of IT in the context of supply shocks, which periodically aff ect individual economies and, as we have seen recently, the world economy as a whole. Rigid IT rules can prove highly problematic in the context of supply shocks, where the main problem facing countries is too little supply, not too much demand. The solution is to help the economy absorb the loss of real income associated with the supply shock, without incurring collateral damage associated with greater income loss than absolutely necessary, while at the same time making the investments necessary to increase the supply of the key commodities or fi nd substitutes for them over the medium term. Contractionary monetary policy is a decidedly highly ineffi cient tool for carrying out this complex task. This problem is exacerbated by rigid infl ation targets and is only slightly ameliorated by the use of ‘core infl ation’ indices which usually ostensibly exclude the fi rst round costs of volatile energy and food prices. This is only a partial solution because in the medium term the increased costs of key commodities must fi lter their way through the input output structure and, even without real wage resistance on the part of workers which the central banks are presumably concerned with, will raise the core infl ation rate temporarily. Hence, rigid adherence to IT in this context will lead to possibly large unnecessary costs in slower economic growth and income.

An overall picture on the selected macroeconomic indicators of the infl ation targeters can be obtained from Tables 1.2 and 1.3. In Table 1.2 we provide information on the observed behavior of selected macro aggregates as the annual average of fi ve years before the adoption of the IT versus the annual average after the adoption date to the current period.


Table 1.3 keeps the same calendar frames and reports data on key macro prices, viz. the exchange rate and the interest rates.

Evidence on the growth performance of the IT countries is mixed. Taking the numbers of Table 1.2 at face value, we see that seven of the 21 countries report a decline in the average annual rate of real growth, while three countries (Canada, Hungary and Thailand) have not experienced much of a shift in their rates of growth. Yet clearly it is quite hard to dis-entangle the eff ects of the IT regime from other direct and indirect eff ects on growth. One such factor is the recent fi nancial glut in the global asset markets and the associated surge of the household defi cit spending bubble, which is now bursting. The Institute of International Finance data reveal, for instance, that the net capital infl ows to the developing economies as a whole has increased from $47 billion in 1998, to almost $400 billion in 2006, surpassing their peak before the Asian crisis of 1997. As the exces-sive capital accumulation in telecommunications and the dot.com high tech industries phased out in late 1990s, the global fi nancial markets seem to have entered another phase of expansion, and external eff ects such as these make it hard to isolate the growth impacts of the IT regimes.

Despite the inconclusive verdict on the growth front, the fi gures on unem-ployment indicate a signifi cant increase in the post-IT era. Only three coun-tries on our list (Chile, Mexico and Switzerland) report a modest decline in their rates of unemployment in comparison to the pre-IT averages. The deterioration of employment performance is especially pronounced (and puzzling) in countries such as the Philippines, Peru and Turkey where rapid growth rates were attained. The increased severity of unemployment at the global scale seems to have aff ected the IT countries equally strongly, and the theoretical expectation that ‘price stability would bring growth and employment in the long run’ seems quite far from materializing yet.

The adjustment patterns on the balance of foreign trade have been equally diverse. Ten of the 21 countries in Table 1.2 achieved higher (improved) trade surpluses (balances). While there have been large defi cit countries, such as Turkey, Mexico, the Philippines and Australia, there were also sizable surplus generators such as Brazil, Korea, Thailand, Canada and Sweden. Not surprisingly much of the behavior of the trade balance could be explained by the extent of over-valuation of the exchange rates. This information is tabulated in Table 1.3.

Table 1.3, like Table 1.2, calculates the annual averages of the fi ve-year period before the IT versus annual averages after IT to date. Focusing on the infl ation-adjusted real exchange rate movements, we fi nd a general tendency towards appreciated currencies in the aftermath of adoption of the IT regimes. Mexico, Indonesia, Korea and Turkey are the most signifi cant currency appreciating countries, while Brazil, and to some


Table 1.2 Selected macroeconomic aggregates in the infl ation targeting countries

Before : Annual average of 5 years prior to adoption of IT; After : Annual average of adoption of IT to current

Growth rate

Unemploy-ment rate

Trade Balance (external

balance on goods and services)

/GDP (%)

Central bank foreign reserves (US$ Million)

Year IT started

Before After Before After Before After Before After

New Zealand 1990 2.7 3.0 4.2 6.9 0.4 1.3 2 897.9 4 623.2Canada 1991 2.9 2.8 8.4 8.7 0.5 2.7 11 964.0 24 256.0United Kingdom 1992 2.2 2.7 7.4 5.2 –2.5 –1.6 39 666.5 37 408.5Sweden 1993 0.8 2.7 2.8 6.1 1.3 6.2 15 399.0 18 521.8Australia 1994 2.2 3.9 8.6 7.3 –0.6 –1.3 13 777.9 20 337.1Israel 1997 5.8 3.1 8.5 9.4 –14.6 –7.2 7 567.3 24 421.1Czech Rep.a 1998 4.5 3.2 4.0 8.9 –3.4 –1.8 9 172.5 21 686.5Poland 1998 7.9 3.7 14.3 16.7 0.0 –4.1 12 591.8 31 581.8Brazilb 1999 3.2 2.3 7.0 9.8 –1.7 1.0 47 701.3 42 304.5Colombia 1999 3.3 2.3 11.1 15.8 –6.0 –0.5 7 567.3 24 421.1Mexico 1999 1.7 4.8 2.7 1.9 –0.5 –1.9 20 630.9 51 396.6South Africac 2000 2.6 3.8 n.a 27.7 0.0 0.0 15 860.0 9 580.0Switzerland 2000 1.4 1.7 4.1 3.1 0.1 0.1 38 277.1 40 646.5Thailand 2000 1.5 1.7 1.9 2.4 0.0 0.1 32 556.1 40 474.8Korea 2001 4.6 4.5 4.4 3.7 0.0 0.0 55 299.5 157 739.2Hungary 2001 4.2 4.2 8.0 6.1 –1.3 –2.9 9 918.1 13 652.1Perud 2002 2.0 5.2 7.8 10.2 –3.2 –0.4 9 264.8 11 222.9Philippines 2002 3.1 5.1 10.2 11.5 –3.6 –0.7 11 281.6 14 006.6Indonesia 2005 4.6 5.6 6.5 10.3 7.3 –4.6 31 326.7 32 989.2Turkeye 2006 4.5 7.8 9.9 10.4 –9.8 –11.0 33 237.4 56 990.4Turkeye 2001Q2 4.0 4.5 6.6 10.0 –7.5 –9.8 20 083.4 33 237.4

Notes:a. The period before the infl ation targeting refers to the period of 1994–97 for ‘Growth’ and

‘CPI infl ation’ for the Czech Republic.b. The period before the infl ation targeting refers to the period of 1996–98 for reserves in

Brazil.c. The period before the infl ation targeting refers to the period of 1994–97 and after infl a-

tion targeting refers to the period of 1999–2004 for unemployment rate in South Africa. Note that due to change in methodology and data coverage, unemployment fi gures are not directly comparable before and after apartheid.

d. The period after the infl ation targeting refers to the period of 2003–04 for unemployment rate in Peru.

e. Offi cial adoption date for Turkey is 2006. However, Turkish central bank declared ‘dis-guised infl ation targeting’ in the aftermath of the 2001 February crisis.

Source: IMF (2008).


Table 1.3 Macroeconomic prices in the infl ation targeting countries

Before: Annual average of 5 years prior to adoption of IT; After: Annual average of adoption of IT to current

Infl ation rate (variations

in CPI)

Exchange rate real

depreciationg,h

CB real interest rateh,i

Public assets real

interest ratei,j

Year IT started

Before After Before After Before After Before After

New Zealand 1990 11.6 2.2 –7.6 –0.6 7.0 5.5 2.1 5.1Chilea 1991 19.7 7.2 –6.0 –4.0 .. 0.0 –16.0 –4.6Canada 1991 4.5 2.1 –7.5 –1.7 6.0 2.6 5.8 2.5United Kingdom 1992 6.4 2.6 –2.4 –2.2 5.4 3.0 5.0 2.8Sweden 1993 6.9 1.5 –8.5 1.2 2.8 1.7 5.0 2.9Australiab 1994 4.2 2.5 –6.9 –1.1 7.1 3.2 6.3 4.0Israel 1997 11.3 3.1 –4.2 0.9 2.0 5.0 1.5 5.0Czech Republicc 1998 9.1 3.1 –6.6 –6.2 1.9 0.7 0.0 0.9Polandd 1998 24.1 4.7 –4.5 –4.6 1.6 6.2 1.8 11.6Brazil 1999 819.2 7.9 –428.0 5.5 –782.6 15.7 –786.9 12.4Colombia 1999 20.4 7.5 –9.5 0.5 18.4 6.6 1.5 2.0Mexico 1999 24.5 7.2 2.8 –4.6 7.5 5.0 3.2 3.8Thailand 2000 5.1 2.2 4.5 –1.0 4.9 1.6 4.7 3.1South Africae 2000 7.3 5.1 4.3 –2.5 8.6 4.4 7.3 4.2Switzerland 2000 0.8 1.0 1.6 –3.7 0.2 0.1 0.9 0.3Korea 2001 4.0 3.3 6.0 –5.0 –0.2 –1.0 6.5 2.1Hungary 2001 15.2 5.9 2.5 –12.4 2.0 3.4 2.3 3.4Peru 2002 5.0 1.9 –1.6 1.4 9.3 2.0 3.8 –0.5Philippines 2002 6.3 5.0 8.7 –3.0 5.1 0.9 5.2 1.2Indonesia 2005 8.0 10.5 –6.2 –1.9 4.2 2.3 4.1 –2.4Turkeyf 2006 28.3 10.5 –6.3 –1.2 11.7 7.5 14.8 10.5Turkeyf 2001Q2 74.1 28.3 –9.9 –6.3 –13.3 12.7 23.9 15.5

Notes:a. The period after the infl ation targeting period refers to the period of 1993–2005; the

period before the infl ation targeting refers to the period of 1987–90.b. Treasury Bill: the period after the infl ation targeting refers to the period of 1994–2000;

CB Rate: the period after the infl ation targeting refers to the period of 1994–95.c. The period before the infl ation targeting refers to the period of 1994–97.d. Treasury Bill rates; the period after the infl ation targeting refers to the period of 1998–2000.e. Treasury Bill: the period before the infl ation targeting refers to the period of 1994–2000.f. Offi cial adoption date for Turkey is 2006. However, Turkish central bank declared ‘dis-

guised infl ation targeting’ in the aftermath of the 2001 February crisis.g. A rise in value indicates depreciation. Annual average market rate is used for: UK,

Canada, Turkey, Australia, New Zealand, Brazil, Peru, Israel, Indonesia, Korea and Philippines. Annual average Offi cial Rate is used for: Colombia, Thailand, Hungary, Poland and Switzerland. Principle rate is used for: South Africa, Mexico and Czech Republic.


extent Columbia, have pursued active export promotion strategies and maintained real depreciation. The Czech Republic, Switzerland and Hungary are observed to have experienced nominal currency apprecia-tion, and Poland seems to have maintained an appreciating path for its real exchange rate.

Clearly much of this generalized trend towards appreciation can be explained by reference to the increased expansion of foreign capital infl ows due to the global fi nancial glut mentioned above. With the IT central bankers announcing a ‘no-action’ stance against exchange rate movements led by the ‘markets’, a period of expansion in the global asset markets has generated strong tendencies for currency appreciation. What is puzzling, however, is the rapid and very signifi cant expansion in the foreign exchange reserves reported by the IT central banks. As reported in the last two columns of Table 1.2, foreign exchange reserves held at the central banks rose signifi cantly in the aftermath of the IT regimes. The rise of reserves was especially pronounced in Korea, the Philippines and Israel where almost a fi ve-fold increase had been witnessed. Of all the countries surveyed in Table 1.3, the UK and Brazil are the only two countries that had experienced a fall in their aggregate reserves.3

This phenomenon is puzzling because the well-celebrated ‘fl exibility’ of the exchange rate regimes were advocated precisely with the argument that, under the IT framework, the central banks would gain freedom in their monetary policies and would no longer need to hold reserves to defend a targeted rate of exchange. In the absence of any offi cially stated exchange rate target, the need for holding such sums of foreign reserves at the central banks should have been minimal. The proponents of the IT regimes argue that the central banks need to hold reserves to ‘maintain price stability against possible shocks’. Yet, the acclaimed ‘defense of price stability’ at the expense of such large and costly funds that are virtually kept idle at the IT central banks’ reserves is questionable at best in an era of prolonged unemployment and slow investment growth, and needs to be justifi ed economically as well as socially.

We now turn to the issue of exchange rate policy more formally.


h. Nominal values are defl ated by the corresponding infl ation averages (CPI column).i. Sweden, New Zealand, Canada: Bank Rate; Mexico: Banker’s Acceptance.j. Colombia: Interbankaria TBS; Peru and Chile: Saving Rate; New Zealand Newly

issued three months Treasury bill rates; Indonesia: three Months Deposit Rate; Korea: National Housing Bond Rate; Thailand: Government Bond Yield Rate.

Source: IMF (2008).


1.3 THE ROLE OF THE EXCHANGE RATE UNDER INFLATION TARGETING

As stated above, part of the broader requirements surrounding the IT system is often argued to be the implementation of a ‘fl oating/fl exible’ exchange rate system in the context of free mobility of capital. Thus, ‘exchange rate fl exibility and fl oating exchange rate system’ became the new motto, and to many advocates, central bank ‘policy’ has typically been reduced to mean merely ‘setting the policy interest rate’. The exchange rate and macro prices are, in theory at least, thereby left to the unfettered workings of the global fi nance markets. The role of the exchange rate as an adjustment variable has clearly increased over the last decade since the adoption of the fl oating exchange rate systems. In the meantime, however, the role of the interest rates and reserve movements has declined substan-tially as counter-cyclical instruments available to be used against shocks4 (see Table 1.2).

Against this background a number of practical and conceptual ques-tions are inevitable: what is the role of the exchange rate in the overall macroeconomic policy when an explicit IT regime is adopted? Under what conditions should the central bank, or any other authority, react to shocks in the foreign exchange market? And perhaps more importantly, if an intervention in the foreign exchange market is regarded necessary against, say, the disruptive eff ects of an external shock, what are the proper instruments?

To the proponents of IT, the answer to these questions is simple and straightforward: the central bank should not have any objective in mind with regards to the level of the exchange rate, yet it might interfere against the volatility of the exchange rate in so far as it aff ects the stability of prices. However nuances remain. To what may be grouped under ‘strict conformists’, the central bank should be concerned with the exchange rate only if it aff ects its ability to forecast and target price infl ation. Any other response to the foreign exchange market represents a departure from the IT system. Advocated in the seminal works by Bernanke et al. (1999) and Fischer (2001), the approach argues that attending to IT and reacting to the exchange rate are mutually exclusive. Beyond this assertion, the con-formist view also holds that intervention in the foreign exchange market could confuse the public regarding the ultimate objective of the central bank with respect to its priorities, distorting expectations. In a world of credibility game, such signals would be detrimental to the central bank’s authority.

Yet, while maintaining the IT objective, one can also distill a more active role for the exchange rate in the literature. As outlined by Debelle


(2001), this ‘fl exible IT’ view proposes that the exchange rate can also be a legitimate policy objective alongside the infl ation target. More formally, an operational framework for the ‘fl exible IT’ view was envisaged within an expanded Taylor rule. Taylor (2000) argued, for instance, that an exchange rate policy rule can legitimately be embedded in a Taylor rule that is consistent with the broad objectives of targeted infl ation rate and the output gap.

In contrast to all this, the structuralist tradition asserts that irrespective of the conditionalities of foreign capital and boundaries of IT, it is very import-ant for the developing economies to maintain a stable and competitive real exchange rate (SCRER) (see, for example, Cordero, Chapter 3, this volume; Frenkel and Taylor, Chapter 2, this volume; Galindo and Ros, Chapter 8, this volume; Frenkel and Ros, 2006; Frenkel and Rapetti, Chapter 9, this volume). They argue that the real exchange rate can aff ect employment, and the economy more generally, through a number of channels: (1) by aff ecting the level of aggregate demand (the macroeconomic channel); (2) by aff ecting the cost of labor relative to other goods and thereby aff ecting the amount of labor hired per unit of output (the labor intensity channel); and (3) by aff ecting employment through its impact on investment and eco-nomic growth (the development channel). While the size and even direction of these channel eff ects might diff er from country to country, maintaining a competitive and stable real exchange rate is likely to have a positive employ-ment impact through some combination of these eff ects.

The gist of the structuralist case for SCRER rests on a recent (and unfortunately not well understood or appreciated) paper by Taylor (2004). Resting his arguments on the system of social accounting ident-ities, Taylor argues that the exchange rate cannot be regarded as a simple ‘price’ determined by temporary macro equilibrium conditions. The main-stream case for exchange rate determination rests on the well-celebrated Mundell (1963) and Fleming (1962) model where the model rests on an assumed duality between reserves (fi xed exchange rate system) versus fl exible exchange rate adjustments. The orthodox mainstream model, according to Taylor, presupposes that a balance of payments exists with a potential disequilibrium that has to be cleared. This, however, is a false presumption. The exchange rate is not an ‘independent’ price and has no fundamentals such as a given real rate of return (or a trade defi cit) that can make it self-stabilizing. In Taylor’s words, ‘the balance of payments is at most an accumulation rule for net foreign assets and has no independent status as an equilibrium condition. The Mundell-Fleming duality is irrel-evant, and in temporary equilibrium, the exchange rate does not depend on how a country operates its monetary (especially international reserve) policy’ (Taylor, 2004, p. 212).


The preceding discussions clearly underscore that the real world behav-ior of exchange rates is quite complex and the focus of the infl ation target-ing regime for fl oating exchange rates (in expectation of dropping it from the policy agenda altogether) is a mirage. This view of exchange rates helps to explain why many believe that there are no viable alternatives to IT as a mode of central bank policy. However, as this book tries to demonstrate, and as we briefl y discuss in the next section, this view of no viable alterna-tives to infl ation targeting is not correct.

1.4 SOCIALLY RESPONSIBLE ALTERNATIVES TO INFLATION TARGETING CENTRAL BANK POLICIES

One reason that ‘infl ation-focused monetary policy’ has gained so many adherents is the common perception that there is no viable alternative monetary policy that can improve growth and employment prospects. There are two main factors accounting for this perception. First, as we dis-cussed in the previous section, in an internationally fi nancially integrated economy with high levels of international capital fl ows, monetary policy can be extremely challenging. In particular it might be very diffi cult to gear monetary policy by targeting monetary aggregates, or by pegging an exchange rate along with trying to promote employment growth. This is often seen as the so-called ‘trilemma’ which commands that central banks can only have two out of three of the following: open capital markets, a fi xed exchange rate system and an autonomous monetary policy geared toward domestic goals. While this so-called ‘trilemma’ is not strictly true as a theoretical matter, in practice it does raise serious issues of mon-etary management (see the above arguments cited from Taylor, 2004 and Frenkel and Taylor, 2006). From our perspective, the real crux of the problem turns out to be the very narrow interpretation of the constraints of the trilemma: central banks are often thought to be restricted to choose two ‘points’ out of three. Yet, the constraints of the trilemma could as well be regarded as the boundaries of a continuous set of policies, as would emerge out of a bounded, yet continuous depiction of a ‘policy triangle’. Thus even within the boundaries of the trilemma a menu of choices does exist, ranging from administered exchange rate regimes to capital manage-ment/control techniques. In fact, many successful developing countries have used a variety of capital management techniques to manage these fl ows in order, among other things, to help them escape the rigid con-straints of the so-called ‘tri-lemma’ (Epstein et al., 2005; Ocampo, 2002).

In this section we report on a series of country studies undertaken by a


team of researchers working on a Political Economy Research Institute (PERI) (University of Massachusetts, Amherst)/Bilkent project on alter-natives to infl ation targeting, as well as a United Nations Development Programme (UNDP) sponsored study of employment targeting economic policy for South Africa. A range of alternatives were developed by the researchers, all the way from modest changes in the IT framework to allow for more focus on exchange rates and a change in the index of infl a-tion used, to a much broader change in the overall mandate of the central bank to a focus on employment targeting, rather than IT. Some of the alternative policies focus exclusively on changes in central bank policy, while for other countries changes in the broad policy framework and in the interactions of monetary, fi nancial and fi scal policy are proposed. Some incorporate explicit goals and targets, while others prefer more fl exibility and somewhat less transparency.

It has to be noted at the outset that ‘infl ation control’ is revealed among the ultimate objectives in all country studies summarized below. Thus there is a clear consensus among the country authors that controlling infl ation is important and desirable. However all agree that the current prescription insisting on ‘very low’ rates of infl ation at the 2–4 percent band is not warranted, and that responsibilities of the central banks, par-ticularly in developing countries, must be broader than that. Accordingly, the policy matrix of the central banks should include other crucial ‘real’ variables that have a direct impact on employment, poverty and economic growth, such as the real exchange rate and/or investment allocation. They also agree that in many cases, central banks must broaden their available policy tools to allow them to reach multiple goals, including, if necessary, the implementation of capital management techniques.

Table 1.4 presents a summary of the alternatives proposed in the PERI/Bilkent project and is discussed further in what follows.

1.4.1 Modest but Socially Responsible Adjustments to the Infl ation Targeting Regime

Some of the country studies in the PERI/Bilkent project proposed only modest changes to the IT regime. In the case of Mexico, for example, the authors argue that the IT regime has allowed for more fl exible monetary policy than had occurred under regimes with strict monetary targets or strict exchange rate targets (Galindo and Ros, Chapter 8, this volume). They suggest modifying the IT framework to make it somewhat more employment friendly. In the case of Mexico, Galindo and Ros fi nd that monetary policy was asymmetric with respect to exchange rate movements – tightening when exchange rates depreciated, but not loosening when

19

Tab

le 1

.4

PE

RI/

Bilk

ent a

ltern

ativ

es to

infl a

tion

targ

etin

g pr

ojec

t, su

mm

ary

of p

olic

y re

com

men

datio

ns

Cou

ntry

Ulti

mat

e ta

rget

sIn

term

edia

te

targ

ets

Stri

ct ta

rget

or

dis

cret

ion

Too

ls/in

stru

men

tsC

entr

al b

ank:

in

depe

nden

t, in

tegr

ated

or

coor

dina

ted?

Arg

entin

aIn

fl atio

n co

ntro

l, ac

tivity

leve

l an

d em

ploy

men

t ex

pans

ion,

ext

erna

l su

stai

nabi

lity

SCR

ER

, int

eres

t ra

teD

iscre

tion

Ster

iliza

tion,

rese

rve

requ

irem

ents

(o

ther

pru

dent

ial

requ

irem

ents

), ca

pita

l m

anag

emen

t te

chni

ques

Coo

rdin

ated

Bra

zil

Infl a

tion

cont

rol,

expo

rt p

rom

otio

n,

inve

stm

ent e

xpan

sion

SCR

ER

Disc

retio

nIn

tere

st ra

te,

asym

met

ric

man

aged

fl oa

t (m

ovin

g fl o

or o

n ex

chan

ge ra

te),

bank

rese

rves

, ba

nk c

apita

l re

quire

men

ts, b

ank

capi

tal

requ

irem

ents

Inte

grat

ed/

coor

dina

ted

with

th

e fi s

cal a

nd a

nti-

pove

rty

obje

ctiv

es

Indi

aG

DP

grow

th,

infl a

tion

cont

rol,

expo

rt p

rom

otio

n

Slig

htly

und

erva

lued

(c

ompe

titiv

e)

exch

ange

rate

Disc

retio

nIn

tere

st ra

te,

capi

tal

man

agem

ent

tech

niqu

es, i

f ne

cess

ary

Inte

grat

ed

20

Tab

le 1

.4

(con

tinue

d)

Cou

ntry

Ulti

mat

e ta

rget

sIn

term

edia

te

targ

ets

Stri

ct ta

rget

or

dis

cret

ion

Too

ls/in

stru

men

tsC

entr

al b

ank:

in

depe

nden

t, in

tegr

ated

or

coor

dina

ted?

Mex

ico

Infl a

tion,

SC

RE

RIm

plem

enta

tion

of

‘dom

estic

’ infl

atio

n m

easu

re, S

CR

ER

, ‘sl

idin

g fl o

or’ o

n ex

chan

ge ra

te

Disc

retio

nC

apita

l man

agem

ent

tech

niqu

esIn

tegr

ated

Sout

h A

fric

aE

mpl

oym

ent

gene

ratio

n, in

fl atio

n co

ntro

l

Rea

l GD

P gr

owth

Stric

t em

ploy

men

t ta

rget

(c

oord

inat

ed

with

oth

er

inst

itutio

ns),

loos

er

infl a

tion

cons

trai

nt

Inte

rest

rate

; cr

edit

allo

catio

n te

chni

ques

(e.g

. as

set-

base

d re

serv

e re

quire

men

ts, l

oan

guar

ante

es e

tc.),

ca

pita

l man

agem

ent

tech

niqu

es

Inte

grat

ed

Tur

key

Infl a

tion

cont

rol;

empl

oym

ent

gene

ratio

n; so

lven

cy

of p

ublic

deb

t; co

nsol

idat

e an

d ex

pand

soci

al

infr

astr

uctu

re

Rea

l int

eres

t rat

e,

non-

appr

ecia

ted

exch

ange

rate

Disc

retio

nC

apita

l man

agem

ent

tech

niqu

es if

ne

cess

ary,

labo

r-ta

x re

form

, inc

reas

ed

publ

ic in

vest

men

ts

in so

cial

cap

ital

Inte

grat

ed/

coor

dina

ted

with

the

fi sca

l an

d em

ploy

men

t ob

ject

ives

21

Phili

ppin

esIn

fl atio

n co

ntro

l; so

lven

cy o

f pub

lic

debt

SCR

ER

Disc

retio

nIn

tere

st ra

te, c

apita

l m

anag

emen

t te

chni

ques

, pr

uden

tial

supe

rvisi

on o

f ba

nks,

targ

eted

cr

edit;

inco

mes

po

licie

s

Inte

grat

ed

Vie

tnam

Inco

mes

gro

wth

, ex

port

pro

mot

ion,

in

fl atio

n co

ntro

l

SCR

ER

Disc

retio

nIn

tere

st ra

te, c

apita

l m

anag

emen

t te

chni

ques

, pr

uden

tial

supe

rvisi

on o

f ba

nks,

targ

eted

cr

edit,

inco

mes

po

licie

s

Inte

grat

ed

Not

e:

SCR

ER

: Sta

ble a

nd C

ompe

titiv

e Rea

l Exc

hang

e Rat

e; C

entr

al B

anks

: int

egra

ted

mea

ns in

tegr

ated

into

gov

ernm

enta

l mac

roec

onom

ic p

olic

y m

akin

g fr

amew

ork;

coo

rdin

ated

mea

ns in

depe

nden

t but

com

mitt

ed to

clo

se c

oord

inat

ion

with

oth

er m

acro

econ

omic

pol

icy

mak

ing

inst

itutio

ns.

Sour

ce:

IMF

(200

8).


exchange rates appreciated. This lent a bias in favor of an over-valued exchange rate in Mexico. So they propose a ‘neutral’ monetary policy so that the central bank of Mexico responds symmetrically to exchange rate movements and thereby avoids the bias toward over-valuation without fundamentally changing the IT framework. In their own words, ‘the central bank would promote a competitive exchange rate by establishing a sliding fl oor to the exchange rate in order to prevent excessive appreciation (an “asymmetric band”. . .). This would imply intervening in the foreign exchange market at times when the exchange rate hits the fl oor (i.e. an appreciated exchange rate) but allows the exchange rate to fl oat freely otherwise’. They point out that such a fl oor would work against excessive capital infl ows by speculators because they would know the central bank will intervene to stop excessive appreciation. If need be, Galindo and Ros also propose temporary capital controls, as do some of the other authors from the PERI/Bilkent project.

In his study of Brazil Nelson Barbosa-Filho 2008 Chapter 7, (this volume) also proposed extending the IT framework, but in a more dra-matic way. According to Barbosa-Filho: ‘because of Brazil’s past experi-ence with high infl ation, the best policy is to continue to target infl ation while the economy moves to a more stable macroeconomic situation. However, the crucial question is not to eliminate infl ation targeting, but actually make it compatible with fast income growth and a stable public and foreign fi nance’.

Given Brazil’s large public debt, Barbosa-Filho proposes that the tar-geted reduction in the real interest rate would reduce the Brazilian debt service burdens and help increase productive investment. In terms of the familiar targets and instruments framework, he proposes that the Brazilian central bank choose exports, infl ation and investment as ultimate targets, and focus on the infl ation rate, a competitive and stable real exchange rate and the real interest rate as intermediate targets. Furthermore, in order to achieve these goals, the central bank can use direct manipulation of the policy interest rate, bank reserve requirements and bank capital requirements.

Brazil is not the only highly indebted country in our project sample. Turkey is another case with that problem. Here too the authors raise con-cerns to the conformist straightjacket of IT, and develop an alternative macroeconomic framework. Using a fi nancial-linked computable general equilibrium model (CGE) for the case of Turkey, Telli et al. (Chapter 10, this volume) illustrate the real and fi nancial sectorial adjustments of the Turkish economy under the conditionalities of the twin targets: on primary surplus to GNP ratio and on the infl ation rate. They utilize their model to study the impact of a shift in policy from a strict IT regime, to


one that calls for revisions of the primary fi scal surplus targets in favor of a more relaxed fi scal stance on public investments on social capital, together with a direct focus on the competitiveness of the real exchange rate. They further study the macroeconomics of a labor tax reform imple-mented through reduction of the payroll tax burden on the producers, and an active monetary policy stance via reduction of the central bank’s interest rates. They report signifi cant employment gains due to a policy of lower employment taxes. They also fi nd that the economy’s response to the reduction of the central bank’s interest rate is positive in general; yet very much dependent on the path of the real exchange rate, thus they also call for maintaining a stable real exchange rate path à la Frenkel, Ros and Taylor.

Frenkel and Rapetti (Chapter 9, this volume), in the case of Argentina, show that targeting a stable and competitive real exchange rate has been very successful in helping to maintain more rapid economic growth and employment generation. In the case of India, Jha (Chapter 12, this volume) also argues against an IT regime, and in favor of one that ‘errs on the side of undervaluation of the exchange rate’ with possible help from temporary resort to capital controls. Jha argues that, to some extent, such a policy would be a simple continuation of policies undertaken in India in the past. In Vietnam Packard concludes: ‘a strict infl ation targeting regime is not appropriate for Vietnam. Infl ation targeting’s rigid rules constrain policy makers to operate in a framework that requires infl ation to take priority over more pressing development objectives. Thus, a stable and competitive real exchange rate is a superior alternative, precisely because it sets as a target a key macroeconomic relative price that is realistic, sustain-able and growth enhancing’ (Packard, Chapter 14, this volume).

1.4.2 More Comprehensive Alternatives to Infl ation Targeting

Other country case studies propose more comprehensive policy alterna-tives to simple infl ation-focused monetary policy, including IT. Joseph Lim (Chapter 13, this volume) proposes a comprehensive alternative to IT for the case of the Philippines. He argues that the Philippines’ govern-ment has been seeking to achieve a record of dramatically higher economic growth, but that its monetary policy has been inadequate to achieving that goal. He therefore proposes an ‘alternative’ that ‘clearly dictates much more than just a move from monetary targeting to infl ation targeting’ with the following proposals: (1) Maintenance of a competitive real exchange rate, either by pegging the exchange rate or intensively managing it as in South Korea. (2) Implementation of capital management techniques, as in China and Malaysia, to help manage the exchange rates. This should


include strong fi nancial supervision to prevent excessive undertaking of short-term foreign debt, and tax based capital controls on short-term capital fl ows, as was used, for example in Chile. (3) An explicit statement of output and employment goals, as the central bank transits from a purely IT regime. (4) Incomes and anti-monopoly policies to limit infl a-tion to moderate levels. (5) Targeted credit programs, especially for export oriented and small and medium sized enterprises that can contribute to productivity growth and employment.

These policy proposals in broad outlines are similar to those proposed by Epstein (Chapter 11, this volume) for the case of South Africa, which, in turn, have been developed in a much broader framework and in more detail by Pollin et al. (2006). Pollin et al. developed an ‘employment-targeted economic program’ designed to accomplish this goal, with a focus on monetary policy, credit policy, capital management techniques, fi scal policy and industrial policy. The purpose of the program is to reduce unemployment rate by half in line with the government’s pledge to reduce the offi cial unemployment rate to 13 percent by 2014.5 Here, ‘employment targeting’ replaces IT as the proposed operating principle behind central bank policy, and moderate infl ation becomes an additional formal con-straint which the central bank must take into account when formulating its policies.

1.5 CONCLUDING COMMENTS

In this introductory chapter we have argued that the current day ortho-doxy of central banking – namely, that the top priority goal for central banks is to keep infl ation in the low single digits – is, in general, neither optimal nor desirable. This orthodoxy is based on several false premises: fi rst, that infl ation, in any magnitude, has high costs; second, that in a low infl ation environment economies will naturally perform best, and in particular, will generate high levels of economic growth and employment; and third, that there are no viable alternatives to this ‘infl ation-focused’ monetary policy.

In fact, moderate rates of infl ation episodes reveal to have very low or no costs; and whether countries where central banks have adopted formal or informal IT have not performed better in terms of economic growth or employment generation is a matter of dispute. Per contra, there are viable alternatives to IT, historically, currently, and looking forward.

Historically, countries both in the currently developed and develop-ing worlds had central banks with multiple goals and tools, and pursued broad developmental as well as stabilization goals. Currently, very


successful economies such as Argentina, China and India have central banks that are using a broad array of tools to manage their economies for developmental purposes. And looking forward, the PERI/Bilkent project on alternatives to IT and PERI’s UNDP work on South Africa have developed an array of ‘real targeting’ approaches to central banking which we believe are viable alternatives to IT and, in particular, do a better job than mere IT in balancing the developmental and stabilization functions of central banks.

NOTES

1. We are indebted to Hasan Comert, Luis Rosero and Lynda Pickbourn for their diligent research assistance, and to Roberto Frenkel, Jose Antonio Ocampo, K.S. Jomo, Geoff rey Woglom, Refet Gürkaynak, James Heintz, Leonce Ndikumana, Arjun Jayadev and Robert Pollin for their valuable comments and suggestions on previous versions of the chapter. Research for this chapter was completed when Yeldan was a visiting Fulbright scholar at the University of Massachusetts, Amherst for which he acknowledges the generous support of the J. William Fulbright Foreign Scholarship Board and the hospitality of the Political Economy Research Institute at the University of Massachusetts, Amherst. We are also grateful to the funders of the PERI/Bilkent ‘Alternatives to Infl ation Targeting’ project, including UN-DESA, Ford Foundation, Rockefeller Brothers Fund and PERI for their support. Needless to mention, the views expressed in the chapter are solely those of the authors and do not implicate in any way the institutions mentioned above.

2. Note that with the use of the term ‘conditionality’ here we refer not to the IMF’s stand-by rules in the narrow sense of balance of payments stabilization, but to the broader set of reforms and structural adjustment agenda as advocated by the international fi nance community and the transnational corporations. Often dubbed as the (post-) Washington consensus, the warranted set of policies range from IT central banks and fl exible foreign exchange markets to broader institutional reforms such as fl exible labor markets, priva-tization and increased governance. See Williamson (1993) for the original deployment of the term, and Rodrik (2003) for further discussion.

3. Brazil’s case is actually explained in part by the recent decision (late 2005) of the Lula government to close its debt arrears with the IMF with early payments out of its reserves.

4. Though note the one sided ever increase in the aggregate reserves of the central banks. The social desirability and economic optimality of this phenomenon in the aftermath of the adoption of fl oating exchange rate systems is another issue that warrants further research.

5. As of March 2005, South Africa had an unemployment rate of anywhere from 26 percent to 40 percent, depending on exactly how it is counted.

REFERENCES

Adelman, I. and E. Yeldan (2000), ‘The minimal conditions for a fi nancial crisis: a multi-regional inter-temporal CGE model of the Asian crisis’, World Development, 28 (6), 1087–100.


Akyuz, Y. (2003), The Developing Countries and World Trade. Performance and Prospects, London: ZED Books.

Akyuz, Y. (2006), ‘From liberalization to investment and jobs: lost in transla-tion’, paper presented at the Carnegie Endowment for International Peace Conference, 14-15 April 2005, Washington, DC, pp. 45–6.

Akyuz, Y., H. Flassback and R. Kozul-Wright (2006), ‘Globalization, inequal-ity and the labour market’, in A. Kose, F. Senses and E. Yeldan (eds), Neoliberal Globalization as New Imperialism: Case Studies on Reconstruction of the Periphery, New York: NOVA Science.

Ball, L. and N. Sheridan (2003), ‘Does infl ation targeting matter?’, International Monetary Fund working paper no. 03/129.

Batini, N., P. Breuer, K. Kochhar and S. Roger (2006), ‘Infl ation targeting and the IMF’, International Monetary Fund staff paper, Washington, DC, March.

Bernanke, B.S., T. Laubach, A.S. Posen and F.S. Mishkin (1999), Infl ation Targeting: Lessons from the International Experience, Princeton, NJ: Princeton University Press.

Buiter, W. (2006), ‘Rethinking infl ation targeting and central bank independence’, paper presented at the Turkish Economic Association Conference, Ankara, September.

Corbo, V., O. Landerretche and K. Schmidt-Hebbel (2001), ‘Does infl ation target-ing make a diff erence? Central Bank of Chile working paper no. 106, accessed 15 January 2007 at www.bcentral.cl/Estudios/DTBC/doctrab.htm.

Debelle, G. (2001), ‘The case for infl ation targeting in East Asian countries’, in Future Directions for Monetary Policies in East Asia, Sydney: Reserve Bank of Australia, pp. 65–87.

Edwards, S. (2005), ‘The relationship between exchange rates and infl ation tar-geting revisited’, National Bureau of Economic Research, working paper no. W12163.

Epstein, G., I. Grabel and K.S. Jomo (2005), ‘Capital management techniques in developing countries: an assessment of experiences from the 1990s and lessons for the future’, in Gerald Epstein (ed.), Capital Flight and Capital Controls in Developing Countries. Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 301–33.

Feldstein, M. (1997), ‘Capital income taxes and the benefi ts of price stability’, National Bureau of Economic Research working paper no. 6200, September.

Fischer, S. (2001), ‘Distinguished lecture on economics in government’, Journal of Economic Perspectives, 15 (2), 3–24.

Fleming, J. Marcus (1962), ‘Domestic fi nancial policies under fi xed and under fl ex-ible exchange rates’, International Monetary Fund staff papers 9, p. 369–79.

Freeman, R. (2004), ‘Doubling the global workforce: the challenges of integrat-ing China, India, and the former Soviet Bloc into the world economy’, paper presented at the conference on ‘Doubling the Global Work Force’, Institute of International Economics, 8 November, Washington, DC.

Freeman, Richard B. (2007), ‘The great doubling: is your job going to Bombay or Beijing?’, in America Works; Critical Thoughts an the Exceptional US Labor Market, New York: Russell Sage Foundation, pp. 128–40.

Frenkel, R. and J. Ros (2006), ‘Unemployment and the real exchange rate in Latin America’, World Development, 34 (4), 631–46.

Frenkel, R. and L. Taylor (2008), ‘Real exchange rate, monetary policy, and


employment’, United Nations Department of Economic and Social Aff airs (DESA) working paper no 19, February, New York.

Heintz, J. (2006), ‘Globalization, economic policy and employment: poverty and gender implications’, Employment Strategy Department employment strategy paper 2006/3, International Labour Organization, Geneva.

International Labour Organization (ILO) (2004), Global Employment Trends 2004, Geneva: ILO.

International Monetary Fund (IMF) (2008) International Financial Statistics, CD Rom, Washington, DC: IMF.

Michl, Thomas (2007), ‘Tinbergen rules the Taylor rule’, Levy Economics Institute working paper no. 444.

Mishkin, F.S. and K. Schmidt-Hebbel (2001), ‘One decade of infl ation targeting in the world: what do we know and what do we need to know?’, National Bureau of Economic Research working paper no 8397, July accessed at www.nber.org/papers/w8397.

Mundell, R.A. (1963), ‘Capital mobility and stabilization policy under fi xed and fl exible exchange rates’, Canadian Journal of Economics and Political Science, 29, 475–85.

Ocampo, J.A. (2002), ‘Capital-account and counter-cyclical prudential regulations in developing countries’, World Institute for Development Economics Research discussion paper, August.

Pollin, R. and A. Zhu (2006), ‘Infl ation and economic growth: a cross-country nonlinear analysis’, Journal of Post Keynesian Economics, 28 (4) (Summer), 593–614.

Pollin, R., G. Epstein, J. Heintz and L. Ndikumana (2006), An Employment-Targeted Economic Program for South Africa, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

Rodrick, D. (2003), ‘Growth strategies’, National Bureau of Economic Research working paper no. 10050.

Roger, S. and M. Stone (2005), ‘On target? The international experience with achieving infl ation targets’, International Monetary Fund working paper WP/05/163, Washington, DC.

Setterfi eld, M. (2006), ‘Is infl ation targeting compatible with post keynesian eco-nomics?’, Journal of Post Keynesian Economics, 28 (4), 653–71.

Taylor, J.B. (2000), ‘Low infl ation, pass-through and the pricing power of fi rms’, European Economic Review, 7 (44), 1389–408.

Taylor, L. (2004), ‘Exchange rate indeterminacy in portfolio balance, Mundell-Fleming and uncovered interest parity models’, Cambridge Journal of Economics, 28, 205–27.

Tobin, J. (1980), ‘Stabilization policy ten years after’, Brookings Papers on Economic Activity, 1, 19–71.

Williamson, J. (1993), ‘Democracy and the Washington Consensus’, World Development, 21 (8), 1329–36.

28

2. Real exchange rate, monetary policy and employment: economic development in a garden of forking pathsRoberto Frenkel and Lance Taylor

Dejo a los varios porvenires (no a todos) mi jardín de senderos que se bifurcan.[I am leaving my garden of diverse paths to some (but not all)]

Jorge Luis Borges, ‘El jardín de los senderos que se bifurcan’

2.1 INTRODUCTION

The exchange rate aff ects any economy through many channels. It scales the national price system to the world’s, infl uences key macro price ratios such as those between tradable and non-tradable goods, capital goods and labor, and even exports and imports (via the costs of intermediate inputs and capital goods, for example). The exchange rate is an asset price, par-tially determines infl ation rates through the cost side and as a monetary transmission vector, and can have signifi cant eff ects (both short and long run) on eff ective demand.

Correspondingly the exchange rate can be targeted toward many policy objectives. In developing and transition economies fi ve have been of primary importance in recent decades:

1. Resource allocation: through its eff ects on the price ratios just men-tioned, the exchange rate can signifi cantly infl uence resource alloca-tion, especially if it stays stable in real terms for an extended period of time. Through eff ects on both resource allocation and aggregate demand, a relatively weak rate can help boost employment, a point of concern in light of stagnant job creation in many developing econ-omies over the past 10–15 years.

2. Economic development: often in conjunction with commercial and industrial policies, the exchange rate can be deployed to

Real exchange rate, monetary policy and employment 29

enhance overall competitiveness and thereby boost productivity and growth.

3. Finance: the rate shapes and can be used to control expectations and behavior in fi nancial markets. Exchange rate policy ‘mistakes’ can easily lead to highly destabilizing consequences.

4. External balance: the trade and other components of the current account usually respond to the exchange rate, directly via ‘substitu-tion’ responses and (at times more importantly) to shifts it can cause in eff ective demand.

5. Infl ation: the exchange rate can serve as a nominal anchor, holding down price increases via real appreciation and/or maintenance by the authorities of a consistently strong rate. As will be seen below, it can also serve as an important transmission mechanism for the eff ects of monetary policy.

All these objectives have fi gured in recent policy experience. Use of the exchange rate to try to improve external balance has been central to countless stabilization packages over the decades, especially in small poor economies. The infl ation objective became crucial in middle-income countries in the last quarter of the twentieth century (and is notably less urgent as of 2005). Along with capital market liberalization, fi xed rates were signifi cant contributors to the wave of fi nancial crises in the 1990s.

But in many ways the resource allocation and developmental objec-tives can be the most important in the long run – the central point of this chapter. We trace the reasons why in the following section on channels of infl uence. We then take up the policy implications, contrasting the use of the exchange rate as a development tool in conjunction with its other uses (often in coordination with monetary policy) to maintain external balance, contain infl ation and stabilize asset markets,

2.2 RESOURCE ALLOCATION, LABOR INTENSITY, MACROECONOMICS AND DEVELOPMENT

Following Frenkel (2004), in this section we trace out three ways in which the exchange rate can have medium- to long-term impacts on develop-ment. We begin with overall resource allocation, and go on to the labor market and macroeconomics.


2.2.1 Resource Allocation

The traditional 2 3 2 trade theory model is a useful starting point. It does focus on the key role of relative prices. It does not take into consideration important non-price components of industrial and commercial polices. Both themes are woven into the following discussion.

The Lerner Symmetry Theorem (1936) is a key early result. Its basic insight is that if only the import/export price ratio is relevant to resource allocation, then it can be manipulated by either an import or an export tax-cum-subsidy. There is ‘symmetry’ between the two instruments, so that ‘under appropriate conditions’ (at hand in the textbooks) only one need be employed.

A now obvious extension is to bring three goods into the discussion: exportable, importable and non-tradable in a ‘Ricardo-Viner’ model. Two price ratios – say importable/non-tradable and exportable/non-tradable – in principle guide allocation. The real exchange rate (RER or r) natu-rally comes into play as the relative price between the non-tradable and a Hicksian aggregate of the two tradable goods.1 These observations lead to two important policy puzzles.

The fi rst has to do with ‘level playing fi elds’. As applied in East Asia and elsewhere, industrial policy often involved both protection of domestic industry against imports by the use of tariff s and quotas, and promotion of exports through subsidies or cheap credits. In the case of a tariff on imports, the domestic price Pm becomes

Pm 5 e(1 1 t)P*m (2.1)

with e as the nominal exchange rate (defi ned as units of local currency per unit of foreign), t the tariff , and P*m the world price. Similarly if the internal price Pe for exports is set from abroad we have

Pe 5 eP*e/ (1 2 s) (2.2a)

with P*e as the world price and s as the subsidy rate.The level playing fi eld rests on the trade theorists’ notion that inter-

nal and external relative prices of tradable goods should be equal, Pm/Pe 5 P*m/P*e . This situation can be arranged if t 5 s 5 0 or more generally (1 1 t) 5 1/ (1 2 s) . The mainstream argument asserts that if all that industrial policy does is give more or less equal protection to both imports and exports, then its costs, administrative complications, and risks of rent-seeking and corruption are unjustifi able. You might as well set t 5 s 5 0 and go to a free trade equilibrium.


In a Ricardo-Viner set-up, with Pn as a price index for non-tradables the price ratios Pe/Pn and Pm/Pn become of interest. Positive values of t and s move domestic relative prices in favor of tradable goods. From a more or less mainstream perspective (Woo, 2005) this outcome can be interpreted as a justifi cation for industrial policy.

The world, however, is a bit more complicated. If the home country is exporting a diff erentiated product, for example, a more appropriate version of Equation (2.2a) is

P*e 5 Pe (1 2 s) /e (2.2b)

so that the foreign price of home exports is set by the subsidy and exchange rate. Presumably, a lower value of P*e stimulates sales abroad. Moreover, if the economic bureaucracy has the requisite motivation and organiz-ation, it can tie export subsidies to the attainment of export, productivity and other targets and so pursue a proactive industrial policy. In such a context, import protection and export promotion serve diff erent purposes: the former allows domestic production to get started along traditional infant industry lines, while the latter enables national fi rms to break into international markets.2

Now focus on the exchange rate. An increase in the nominal rate e would also switch incentives toward production of tradables, without the need for extravagant values of s and t. This simple observation is in fact a strong argument in support of the use of a depreciated RER as a develop-mental tool. If we defi ne r as

r 5 [mPm 1 (1 2 m)Pe ] /Pn (2.3)

with m as the weight in a tradable goods price index, then a high value of e means that the RER r will also be weak or depreciated.

Of course, a weak RER may not be a suffi cient condition for long-term development. For example, it may usefully be supplemented by an export subsidy or tariff protection to infant industries with their additional potential benefi ts as mentioned above. Even without an eff ective bureauc-racy, generalizing Lerner symmetry to a Ricardo-Viner world suggests that more than one policy instrument may be helpful because there are two relative price ratios that can be manipulated (Lerner, 1936). The rub is that a strong exchange rate implies that commercial/industry policy interventions also have to be strong, with correspondingly high interven-tion costs. A weak RER may be only a necessary condition for benefi cial resource reallocation to occur, but a highly appreciated RER is likely to be a suffi cient condition for ‘excessive intervention’ in a situation in which


development cannot happen. It is hard to fi nd examples of economies with strong exchange rates that kept up growth for extended periods of time.

2.2.2 Labor Intensity

Continuing with the allocational theme, it is clear that the exchange rate will aff ect relative prices of imported intermediates and capital goods on the one hand, and labor on the other. Moreover, the RER largely deter-mines the economy’s unit labor costs in terms of foreign currency.

To explore the implications, we can consider the eff ects of sustained real appreciation on diff erent sectors. Producers of importables will face tougher foreign competition. To stay in business they will have to cut costs, often by shedding labor. If they fail and close down, more jobs will be destroyed. If home’s export prices P*e are determined by a relationship like Equation (2.2b), similar logic applies to that sector. In non-tradables, which will have to absorb labor displaced from the tradable sectors, jobs are less likely to open up insofar as cheaper foreign imports in the form of intermediates and capital goods substitute for domestic labor. On the whole, real appreciation is not likely to induce sustained job creation and could well provoke a big decrease in tradable sector employ-ment. Reasoning in the other direction, RER depreciation may prove employment-friendly.

In both cases it is important to recognize that a new set of relative prices must be expected to stay in place for a relatively long period if these eff ects are going to work through. Changes in employment/output ratios will not happen swiftly because they involve restructuring fi rms and sectoral labor market behavior. This must take place via changes in the pattern of output among fi rms and sectors, by shifts in the production basket of each fi rm and sector, and adjustments in the technology and organization of production. These eff ects arise from a restructuring process in which indi-vidual fi rms and the organization of economic activity adapt to a new set of relative prices. Gradual adjustment processes are necessarily involved.

Finally, in the long run if per capita income is to increase there will have to be sustained labor productivity growth with employment creation sup-ported by even more rapid growth in eff ective demand. Macroeconomics comes into play.

2.2.3 Macroeconomics

The question is how a weak exchange rate (possibly in combination with other policies aimed at infl uencing resource allocation among traded goods) fi ts into the macroeconomic system. Much depends on labor


market behavior in the non-traded sector. Following Rada (2005) we work through one scenario here, to illustrate possible outcomes.

Assume that output in the tradable sector is driven by eff ective demand, responding to investment, exports and import substitution as well as fi scal and monetary policy. The level of imports depends on economic activity and the exchange rate (along with commercial/industrial policies). A worker not utilized in tradable sectors must fi nd employment in non-tradables, become under- or unemployed, or leave the labor force.

For concreteness, we assume that almost all labor not employed in tradables fi nds something to do in non-tradable production as a means of survival. Typical activities would be providing labor services in urban areas or engaging in labor-intensive agriculture. If Lt is tradable sector employment and L is the economically active population, then employ-ment in non-tradables is Ln 5 L 2 Lt. With wn as the non-tradable wage, the value of labor services provided is Yn 5 wnLn. The tradable sector wage rate wt is determined institutionally, at a level substantially higher than wn.

The non-tradable sector’s demand-supply balance thus takes the form

Yn 2 wnLn 5 Yn 2 wn (L 2 Lt) 5 0 (2.4)

Demand for Yn is generated from the value of tradable sector output PtXt. At the same time, real output Xt determines Lt and thereby Ln. Suppose that Pt is set by mark-up pricing on variable costs including labor and imports. Then from both the demand and supply sides an increase in Xt leads to a tighter non-traded labor market which should result in an increase in wn. Equation (2.4) becomes the upward-sloping ‘Non-tradable equilibrium’ schedule in Figure 2.1. Non-tradable labor services become more valuable when economic activity rises. In national accounting terms this signals a productivity increase in the sector because each worker pro-ducers a higher value of output in terms of tradable goods, or a general price index. In other words, an endogenous productivity level is built into the specifi cation.

If workers in both sectors don’t save, then their behavior does not infl u-ence overall macroeconomic balance. Leaving aside a formal treatment of fi scal and monetary instruments for simplicity, the equation takes the form

It 1 Et 2 spXt 2 eP*m (1 1 t)aXt/Pt 5 0 (2.5)

Demand injections come from investment It, exports Et and changes in the magnitude of the import coeffi cient a via import substitution. Saving leakages come from profi ts with p as the tradable sector profi t share and


s the saving rate as well as from ‘foreign saving’ in the form of imports. Equation (2.5) is the vertical ‘Macroeconomic equilibrium’ line in Figure 2.1. Together, the two schedules determine Xt and wn. In the lower quad-rant the trade defi cit is assumed to be an increasing function of tradable sector output in the short run.

Now consider the outcomes of a devaluation. It will have impacts all over the economy, including a loss in national purchasing power if imports initially exceed exports, redistribution of purchasing power away from low-saving workers whose real wages decrease, a decline in the real value of the money stock and capital losses on the part of net debtors in international currency terms. Presumably exports will respond positively to an RER depreciation but that may take time if ‘J-curve’ and similar eff ects matter. Another positive impact on the demand for tradables will

Trade deficit

Non-tradablesector wage

wn

Tradable sectoroutput Xt

Macroeconomic equilibrium

Non-tradableequilbrium

Figure 2.1 Equilibrium between tradable and non-tradable sectors


come from import substitution, reducing the magnitude of the coeffi cient a.

One implication is that for a given level of output, the trade defi cit should fall with devaluation, or the corresponding schedule should shift toward the horizontal axis in the lower quadrant. If devaluation is con-tractionary, the macroeconomic equilibrium schedule will shift leftward in the upper quadrant, reducing Xt, wn, and the trade defi cit further still. In this case real devaluation should presumably be implemented together with expansionary fi scal and monetary policies. As discussed in detail below, exchange rate strategies must be coordinated with other policy moves.

If export demand and production of import substitutes are stimulated immediately or over time by a sustained weak RER, the macroeconomic equilibrium curve should drift to the right, driving up economic activity and employment in the medium to long run.

So far, the analysis has taken labor productivity as a constant. Medium- and long-run considerations have to take into account the evolution of productivity. For the tradable sector, this question can be analysed in terms of Figure 2.2, sketched verbally but not actually drawn by Kaldor in his 1966 Inaugural Lecture (published in Kaldor, 1978). To the tra-ditional diagram we follow Rada and Taylor (2004) by adding dashed

Output growth rate

X t

Labor productivitygrowth rate

ξ Lt

Kaldor-Verdoorn

Output growth

Employmentgrowth

contours

Figure 2.2 Output, labor productivity and employment growth in the tradable sector


‘Employment growth contours’ with slopes of 45 degrees. Each one shows combinations of the output growth rate (Xt 5 (dXt/dt) /Xt 5 X

#t/Xt) and

labor productivity growth rate (xLt) that hold the employment growth rate (Lt 5 Xt 2 xLt) constant. Employment growth is more rapid along contours further to the south east.

Movements across contours show the eff ects on employment growth of shifts in the diagram’s two solid curves. The ‘Kaldor-Verdoorn’ sched-ule represents a ‘technical progress’ function of the form proposed by Verdoorn (1949) and Okun (1962),

xLt 5 xLt 1 gXt (2.5)

in which the productivity trend term xLt could be aff ected by human capital growth, industrial policy, international openness, population growth and other factors.

The ‘Output growth’ curve refl ects the assumption that more rapid pro-ductivity growth can make output expand faster, for example, by reducing the unit cost of exports. The diagram presupposes that this eff ect is rather strong because the slope of the ‘Output growth’ line is less than 45 degrees, implying that 0Xt/0xLt . 1.

If a depreciated RER stimulates net export growth, the ‘Output growth’ curve will shift to the right, causing xLt, X

$t and Lt all to increase. One might

also imagine that the trend rate xLt of productivity growth could rise in the new regime. The ‘Kaldor-Verdoorn’ schedule would shift upward, and with a relatively fl at output growth curve, all three growth rates would rise.

However, if the slope of the ‘Output growth’ curve were to exceed 45 degrees, eff ective demand would not increase as rapidly as productivity so that Lt would have to fall.

What happens to wages and productivity in the non-tradable sector? Let l 5 Lt/L be the share of tradable sector employment in the total. Then lLt 1 (1 2 l)Ln 5 L where L is overall employment growth. Non-tradable employment expansion becomes

Ln 51

1 2 l[L 2 l (Xt 2 xLt) ]. (2.6)

Let the elasticity of demand for non-tradables with respect to Xt be u. Diff erentiating Equation (2.4) then gives

wn 5 uXt 11

1 2 l[l (Xt 2 xLt) 2 L ] (2.7)

Even taking into account the favorable eff ects on employment of a weak exchange rate that were mentioned above, a low demand elasticity u and


fast labor force growth L could mean that a strong export performance translates into weak or even negative wage and productivity growth in the non-traded sector. A case like this calls for fi scal and social policies intended to foster demand for non-tradables and compensate for the nega-tive eff ects on income distribution and employment.

2.3 MACROECONOMIC POLICY REGIMES FOR A STABLE COMPETITIVE REAL EXCHANGE RATE

For the reasons just indicated, a competitive and stable RER can make a substantial contribution to economic growth and employment crea-tion. Programming the RER, however, is no easy task. It is most directly impacted by the nominal exchange rate, itself infl uenced by many factors, but also depends on the overall infl ation rate and shifting relative prices. Nor can the RER be the only macro policy objective. In any economy there are bound to be multiple and partially confl icting objectives. And all policies – exchange rate, fi scal, monetary and commercial/industrial – are interconnected and have to be coherently designed and implemented.

The following discussion focuses on these exchange rate coordination issues in the context of middle income economies with at least sporadic access to private international capital markets. Although they are not addressed in detail here, somewhat similar questions can easily arise in low income countries receiving offi cial capital infl ows, especially if they jump to levels of 10–20 percent of GDP as suggested in the discussion of the Millenium Development Goals (MDG).

So how can policy makers target the RER while at the same time con-trolling infl ation, reducing fi nancial fragility and risk, and aiming toward full employment of available resources? Our focus necessarily has to shift from the ‘real economy’ to encompass monetary and expectational consid-erations. The principal emphasis is on the degrees of freedom available to the monetary authorities, if only because they have been at center stage in recent policy debates.

2.3.1 What Determines the Nominal Exchange Rate?

To set the stage, a few observations about how the nominal exchange rate fi ts into the macroeconomic system make sense.

Theories that can reliably predict the level of the rate and its changes over time when it is not strictly pegged do not exist. (The fact that pegs not infrequently break down means they do not have 100 percent predictive


power either.) In present circumstances in developing and transition economies (especially those at middle income levels) it is not unreason-able to assume that a more or less fl oating rate is determined in spot and future asset markets; in eff ect the spot rate fl oats against its ‘expected’ future values. The quotation marks mean that we view expectations along Keynesian lines as emerging from diverse opinions on the part of market participants about how the rate may move. ‘Beauty contests’ which magnify small shifts in average market opinion and other sources of seem-ingly capricious market behavior can easily come into play (Eatwell and Taylor, 2000).

With regard to the level of the rate, it is useful to think about a simple bond market equilibrium condition such as

i 5 f(e, e# exp, M) (2.8)

with i as the local interest rate, e the spot exchange rate, e# exp the expected (as an aggregate of by market perceptions) change in the rate over time and M an index of monetary relaxation. A high or depreciated value of e means that national liabilities are cheap as seen from abroad. It should be associated with high local bond prices or low interest rates. If expected depreciation e# exp rises, on the other hand, foreign wealth-holders will want to shift away from local liabilities and i will increase. Open market bond purchases will increase M and be associated with a reduction in i.

Over the past couple of decades under conditions of external liberaliz-ation, most developing economies have been affl icted by high local interest rates and appreciated currencies. This unfavorable constellation of ‘macro prices’ is consistent with Equation (2.8).

The dynamics of the exchange rate will be infl uenced by interest rates, because it is an asset price. One crucial question is whether lower domes-tic rates will tend to make the nominal rate depreciate or appreciate. If it tends to rise (or depreciate) over time, then exchange rate dynamics can be a powerful mechanism for transmitting the eff ects of expansionary mon-etary policy into infl ation by driving up local production costs.

Standard arbitrage arguments as built into interest rate parity theorems imply that the expected change in the spot rate e# exp should be an increasing function of the diff erence between domestic and foreign rates. If myopic perfect foresight applies, the expected change will be equal to the observed change (up to a ‘small’ error term). Hence a lower local interest rate should cause appreciation over time. On Wall Street, such an analysis of exchange rate movements is called an ‘operational’ view.

A ‘speculative’ view is that the exchange rate will depreciate when the local interest rate decreases.3 This view makes intuitive sense insofar as low


interest rates should make national liabilities less attractive. It was perhaps fi rst advanced macroeconomically by Minsky (1983) and can be made consistent with the parity theorems if it is assumed that there is a relatively strong positive feedback of expected exchange rate increases into the domes-tic interest rate via the bond market equilibrium condition Equation (2.6).

Recent macroeconomic history (Frenkel, 2004) suggests that the specu-lative view is the more accurate description of exchange rate behavior in middle income economies.

2.3.2 Avoiding Catastrophes

The most fundamental justifi cation for avoiding a persistently strong exchange rate is that it is an invitation to disaster. Exchange appreciation is always welcome politically because it may be expansionary (at least in the short run), is anti-infl ationary and reduces import costs (including foreign junkets for those who can aff ord them). However, for the reasons discussed above, it can have devastating eff ects on resource allocation and prospects for development. Moreover, fi xed or quasi-fi xed strong real rates can easily provoke destabilizing capital fl ow cycles as perhaps fi rst described analytically by Frenkel (1983) and re-enacted many times since.

The existence and severity of these cycles is in practice a powerful argu-ment for a stable exchange rate regime built around some sort of managed fl oat (details below). A fl oating rate does appear to moderate destabiliz-ing capital movements in the short run, and is therefore a useful tool to deploy. At the same time, the central bank has to prevent the formation of expectations that there will be RER appreciation, which can easily become self-fulfi lling along beauty contest lines. A commitment to a stable rate, backed up by forceful intervention if necessary, is one way the bank can orient expectations around a competitive RER.

2.3.3 Trilemmas

Possibilities for central bank intervention are often said to be constrained by a ‘trilemma’ among (1) full capital mobility, (2) a controlled exchange rate and (3) independent monetary policy. Supposedly, only two of these policy lines can be consistently maintained. If the authorities try to pursue all three, they will sooner or later be punished by destabilizing capital fl ows, as in the run-up to the Great Depression around 1930 and Britain and Italy’s diffi culties during the European Exchange Rate Mechanism (ERM) crisis more than 60 years later.

The trilemma as just stated is a textbook theorem which is, in fact, invalid. Even with free capital mobility, a central bank can undertake


transactions in both foreign and domestic bonds (not to mention other monetary control maneuvers) to regulate the money supply, regardless of whatever forces determine the exchange rate (Taylor, 2004).

Nevertheless, something like a trilemma can exist in the eye of a beholder. There are practical limits to the volume of interventions that a central bank can practice, along with complicated feedbacks. Possibilities for sterilizing capital infl ows or outfl ows are bounded by available asset holdings. Volumes of fl ows depend on exchange rate expectations which in turn can be infl uenced by central bank behavior and signaling.

So how does the market decide when a perceived trilemma is ripe to be pricked? The fact that no single form of transaction or arbitrage opera-tion determines the exchange rate means that monetary authorities have some leeway in setting both the scaling factor between their country’s price system and the rest of the world’s and the rules by which it changes. However, their sailing room is not unlimited. A fi xed rate is always in danger of violating what average market opinion regards as a fundamental. Even a fl oating rate amply supported by forward markets can be an invita-tion to extreme volatility. Volatility can lead to disaster if asset preferences shift markedly away from the home country’s liabilities in response to shifting perceptions about fundamentals or adverse ‘news’. Unregulated international capital markets are at the root of any perceived trilemma. It is a practical problem that must be evaluated in each case, taking into account the context and circumstances of policy implementation.

2.3.4 Monetary and Exchange Rate Policies and Capital Flows

The implication is that if it wishes to target the RER, the central bank has to maintain tolerable control over the macroeconomic impacts of cross-border fi nancial fl ows in a world with relatively open foreign capital markets. For the sake of clarity, it makes sense to analyse situations of excess supply and excess demand for foreign capital separately.

Large capital infl ows can easily imperil macro stability. Indeed, central bank attempts to sterilize them by selling domestic liabilities from its portfolio may even bid up local interest rates and draw more hot money. Preservation of monetary independence in this case may well require capital market regulation. Measures are available for this task.4 They do not work perfectly, but can certainly moderate infl ows during a boom. Booms never last forever; the point is that the authorities can use capital market inter-ventions to slow one down to avoid an otherwise inevitable crash.

If there are capital outfl ows too large to manage with normal exchange rate and monetary policies, the authorities certainly do not want to engage in recession-triggering monetary contraction. If the exchange rate has been


maintained at a relatively weak level, the external defi cit is not setting off fi nancial alarm bells, and infl ation is under control, then there are no ‘fun-damental’ reasons for market participants to expect a maxi-devaluation. Under such circumstances the way for the authorities to maintain a policy regime consistent with a targeted RER is to impose exchange controls and restrictions of capital outfl ows.

Contrary to IMF-style opinion that all runs against a currency must be triggered by poor fundamentals (even if they momentarily escape the notice of the authorities and IMF offi cials), it is perfectly clear that they can arise for reasons extraneous to economic policy – think of a politi-cal crisis, the fallout from mismanagement of an important bank or the impacts of fi nancial contagion from a regional neighbor. In all such cases outfl ow controls can be used to maintain an existing policy package in place. They may not have to be utilized for very long.5

2.3.5 Monetary Policy

In a developmental policy regime, monetary policy must be designed in view of its likely eff ects on the RER, infl ation control and the level of eco-nomic activity. There is nothing very surprising here – in practice central banks always have multiple objectives. In the USA, despite lip service to controlling price infl ation, the Federal Reserve certainly responds to the level of economic activity and fi nancial turmoil (witness the 1990s stock market bubble and the long-term capital management (LTCM near-crisis). In many developing countries central banks intervene more or less system-atically in the exchange markets. The proposal here is that these interven-tions should help support a developmentally oriented RER for the reasons presented above. That is, the nominal rate should move to hold the RER in the vicinity of a stable competitive level for an extended period of time.

Infl ation targeting, on the other hand, is the current orthodox buzzword. The nominal exchange rate and other policies should be programmed to ensure a low, stable rate of infl ation. A trilemma-like argument is involved. If exchange market interventions target the RER as opposed to the nominal exchange rate and the central bank cannot manage the money supply, there is no nominal anchor on infl ationary expectations. The infl a-tion rate cannot be controlled.

As we have seen, in practical terms the trilemma can be circumvented, allowing the monetary authorities to bring developmental objectives into their remit. But they have to take at least fi ve important considerations into account in monetary management.

First, many developing countries now have low to moderate infl ation rates, demoting infl ation control in the hierarchy of policy objectives.


Second, will low interest rates tend to set off infl ationary nominal depre-ciation (under ‘speculative’ exchange rate dynamics as discussed above)? RER targeting can help the central bank steer away from this problem.

Third, shifts in aggregate demand likely to result from changes in the exchange rate and monetary policy must be taken into account, and appropriate off setting policies deployed.

Fourth, also as mentioned above, some mix of temporary capital infl ow or outfl ow controls may be needed to allow the central bank to regulate monetary aggregates and interest rates rather than be overwhelmed by attempts at sterilization.

Finally, unstable money demand and other unpredictable factors mean that the monetary authorities have to be alert and fl exible. Indeed, ‘infl a-tion targeting’ is a codeword for orthodox recognition that quantitative monetary and even interest rate targets are impractical. It is a means for giving more discretion in trying to attain a single target.

The point being made here is that discretion can and should serve other ends. A stable competitive RER in coordination with sensible industrial and commercial policies can substantially improve prospects for economic development. Surely that should be the overriding goal of the mon-etary and all other economic authorities in any developing or transition economy.

NOTES

1. Just to be clear, we will treat the RER as the ratio of tradable to non-tradable price indexes. Real devaluation or weakening the RER means that r increases.

2. Again, these arguments are old. Ocampo and Taylor (1998) provide a recent summary.3. To be more precise, the change over time in the spot rate e# 5 de/dt will turn negative

when i decreases if the operational view applies and positive when the speculative view is true.

4. For an ample menu, see papers by Deepak Nayyer, Eric Helleiner and Gabriel Palma in Eatwell and Taylor (2002) and Epstein et al. (2003). Salih Neftci and Randall Dodd assess the possibilities of using fi nancial engineering to circumvent controls.

5. Argentina, for example, successfully managed exchange controls and capital outfl ow restrictions in mid 2002. The measures were transitory. They were gradually softened as buying pressure in the exchange market diminished.

REFERENCES

Eatwell, J. and L. Taylor (2000), Global Finance at Risk: The Case for International Regulation, New York: The New Press

Eatwell, J. and L. Taylor (eds) (2002), International Capital Markets: Systems in Transition, New York: Oxford University Press.


Epstein, Gerald, Ilene Grabel and K.S. Jomo (2003), ‘Capital management tech-niques in developing countries’, in Ariel Buira (ed.), Challenges to the World Bank and IMF; Developing Country Perspectives, London: Anthem Press, reprinted in Gerald Epstein (ed.) (2005), Capital Flight and Capital Controls in Developing Countries, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

Frenkel, R. (1983), ‘Mercado Financiero, expectativas cambiales, y movimientos de capital’, El Trimestre Economico, 50, 2041–76.

Frenkel, R. (2004), ‘Real exchange rate and employment in Argentina, Brazil, Chile, and Mexico’, paper prepared for the Group of 24, Washington, DC: September.

Kaldor, N. (1978), ‘Causes of the slow rate of growth of the United Kingdom’, in Further Essays on Economic Theory, London: Duckworth.

Lerner, A.P. (1936), ‘The symmetry between import and export taxes’, Economica, 3, 306–13.

Minsky, H.P. (1983), ‘Monetary policies and the international fi nancial environ-ment’, Washington University Department of Economics mimeo, St Louis, MO.

Ocampo, J.A. and L. Taylor (1998), ‘Trade liberalization in developing econ omies: modest benefi ts but problems with productivity growth, macro prices, and income distribution’, Economic Journal, 108, 1523–46.

Okun, A.M. (1962), ‘Potential GNP: its measurement and signifi cance’, reprinted in Joseph Pechman (ed.) (1983), Economics for Policy-Making, Cambridge, MA: MIT Press.

Rada, C. (2005), ‘A growth model for a two-sector open economy with endog-enous employment in the subsistence sector’, Schwartz Center for Economic Policy Analysis, New School University, New York.

Rada, C. and L. Taylor (2004), ‘Empty sources of growth accounting, and empiri-cal replacements à la Kaldor with some beef’, Schwartz Center for Economic Policy Analysis, New School University, New York.

Taylor, L. (2004), ‘Exchange rate indeterminacy in portfolio balance, Mundell-Fleming, and uncovered interest rate parity models’, Cambridge Journal of Economics, 28, 205–27.

Verdoorn, P.J. (1949), ‘Fattori che Regolano lo Sviluppo della Produttivita del Lavoro’, L’Industria, 1, 3–10.

Woo, W.T. (2005), ‘Some fundamental inadequacies in the Washington Consensus: misunderstanding the poor by the brightest’, in Jan Joost Teunissen (ed.), Stability, Growth, and the Search for a New Development Agenda: Reconsidering the Washington Consensus, The Hague: Forum on Debt and Development, (FONDAD).

44

3. Infl ation targeting and the real exchange rate in a small economy: a structuralist approachJose Antonio Cordero

3.1 INTRODUCTION

In the past few years several nations, especially encouraged by the International Monetary Fund (IMF), have decided to move into a mon-etary regime based on explicit infl ation targets. This regime has been strongly recommended to economies that were struggling to bring their infl ation rates down to the level of the more advanced countries. The application of this regime has been very eff ective in reducing infl ation but, as argued in Epstein and Yeldan (2008), the fi nal eff ects on employment and economic growth are not so clear. A rigorous analysis of the conse-quences of infl ation targeting in open economies requires a formal model exploring the interaction among money, the foreign sector, output and employment.

Within the structuralist tradition several models have been developed in order to examine growth and distribution in open economies. These works, however, do not always include the monetary sector and, when they do, they rarely establish the links between the latter and the balance of payments; they are not designed to examine the association between the monetary regime and the foreign exchange rate system. This chapter attempts to fi ll that void by means of a framework that allows comparing two alternative monetary regimes: one resulting from infl ation targeting, and another resulting from real exchange rate targeting.

The model allows formalizing the adjustment process in three periods: fi rst, output and employment vary with given prices, second, infl ation clears an explicit monetary sector and, fi nally, capital accumulation, income distribution and the real exchange rate reach a stable steady state. The growth of labor productivity, through learning eff ects, leads to endog-enous growth.

Infl ation is explained here as a monetary phenomenon, but eff ective

Infl ation targeting and the real exchange rate in a small economy 45

demand remains a critical determinant of output, employment and eco-nomic growth. The chapter thus also contributes to a better understanding of the interaction between the real and the monetary sector.

In Section 3.2 the central bank adjusts the nominal exchange rate to hit a real exchange rate target, with infl ation control playing an important but secondary role. In this case the growth and employment performance are favorable but the economy suff ers the diffi culties arising from the ‘tri-lemma’. It is shown that these complications may be mitigated by reducing the emphasis on open market operations, and utilizing the possibilities provided by the legal reserve requirement or capital controls.

The model in Section 3.3 presents a monetary system in which the nominal interest rate is chosen to hit an infl ation target, and the nominal exchange rate fl oats freely to equilibrate the balance of payments. Under this infl ation targeting scheme, price stability becomes the only goal of monetary policy. We show that this arrangement is very eff ective in bring-ing infl ation down, but overdetermination hurts both the employment and growth potential of the economy.

For each of the models developed in Sections 3.2 and 3.3, the dynamic properties of the long-run equilibrium are also analysed. Section 3.4 presents some concluding remarks.

3.2 A SMALL OPEN ECONOMY WITH A REAL EXCHANGE RATE TARGET

Here a real exchange rate target is set by the central bank at a level that is just the one required to maintain the competitiveness of exports. Maintaining exports will help employment so that, in a way, a real exchange rate target is analogous to an employment target.

The economy operates in the following way. The foreign exchange rate system is better defi ned as one in which the nominal rate is adjusted (crawl-ing peg) by the central bank to bring the actual real rate to the specifi ed target. This framework analyses the diffi culties the central bank faces when attempting to control both the money supply and the foreign exchange rate under an open capital account. This is the well known problem of the ‘tri-lemma’ of monetary policy: with a rigid nominal exchange rate, attempts to reduce the money supply (and therefore infl ation) by raising the interest rate cause an increase in international monetary reserves which, in turn, off sets the reduction in money supply (and thus infl ation) that authorities were originally looking for.

This situation leads to infl ation rates that are higher than internat-ionally accepted levels. But it also allows the central bank to make the


adjustments that are needed to maintain the competitiveness of exports and the rate of economic growth.

3.2.1 Assumptions of the Model

Some assumptions below are rather general, but others reveal the structur-alist approach that is followed in building the model.1

1. The country produces a consumption good ‘Q’, which could be exported or consumed domestically.

2. The country imports a composite good that may be used for consump-tion or investment purposes. All capital used in the country is imported.

3. The small economy determines domestically the price of local produc-tion PQ, but it cannot aff ect the international price of imports PM, assumed exogenous. Local prices PQ, are assumed fi xed in the short run, but may change in the medium run. This assumption, in relation to the price of local production, allows analysing the eff ect that infl a-tion may have on the competitiveness of exports.

4. The price of imports, in terms of foreign currency is PM, assumed exog-enous as seen above. In terms of local currency, the price of imports is ePM, where ‘e’ represents the nominal exchange rate (number of units of local currency that have to be paid for one unit of foreign currency).

5. There are two social classes: producers and workers. Workers spend all their wage income on consumption of both the local good and the imported good, while producers save a fi xed portion ‘s’ of their profi t income and the rest they spend on consumption of both the domestic and imported goods.

6. Good Q is produced by means of a fi xed-coeffi cient production func-tion: Q 5 Min [L

a, uK ] where ‘L’ represents the labor input, ‘a’ is a technical coeffi cient, ‘K’ is capital and ‘u’ represents the output/capital ratio. Capital does not depreciate, and fi rms are assumed to hold excess capacity so that ‘u’ is a variable that moves up (down) when capacity utilization increases (decreases).

7. The interest rate is a policy variable fi xed by the central bank for stabilization purposes.

8. The capital stock K and the labor force N (and of course their ratio, denoted here as k) are given in the short and medium run, but they may change in the long run.

9. The nominal exchange rate ‘e’ is a policy variable that remains fi xed in the short and medium run, but is adjusted in the long run by the central bank. The adjustment process follows a rule that allows hitting a real exchange target.


10. The real exchange rate, which determines the competitiveness of exports, is defi ned as h 5

ePM

PQ .11. The wage share is assumed given in the short and medium run, and is

defi ned as A 5WaPQ . In the short run W and the technical coeffi cient a

are also fi xed, but they adjust in the long run.12. The analysis is conducted in three stages: in the short run output

adjusts while prices, the real exchange rate and the distribution of income remain given; in the medium run infl ation will move to clear the monetary market; and in the long run the wage share, the real exchange rate and the capital/labor force ratio are allowed to vary.

3.2.2 General Overview

In this section we present the goods market, and defi ne the conditions for macroeconomic equilibrium in the short run (equality between the invest-ment and savings rates). With the previous assumptions we may fi nd equi-librium values for the rate of growth of the capital stock, the rate of profi t, the output/capital ratio and the employment rate; infl ation only adjusts in the medium run, as seen below.

Within the foreign sector, in the short run, the given levels of the nominal interest rate and real exchange rate help us determine the increase in international monetary reserves; this then enters into the determination of money supply growth. At this point we have already determined the short run variables, and the monetary sector may then clear in the medium run through adjustments in the infl ation rate.

Once the whole model is solved we examine the impact that open market operations have on infl ation and economic growth. It is at this point that the ‘trilemma’ arises and the real exchange rate emerges as a leading char-acter in the story. The relevance of this variable stems also from the fact that it is, along with the interest rate, the only one capable of stimulating the level of economic activity.2

3.2.3 Equations of the Model

Goods market

gs 5 sr 1 B (3.1)

gd 5 b0 1 b1r 2 b2i (3.2)

r 5uh

(1 2 A) (3.3)


PQ 5 PQ (3.4)

gd 5 gs (3.5)

LN

5LQ

QK

KN

5 l 5 auk (3.6)

Monetary market

Md 5 h1p 1 h2u 2 h3i 2 h4pE (3.7)

MS 5 R 1 t0 2 t1i (3.8)

MS 5 Md (3.9)

i 5 i0 (3.10)

Foreign sector

B 5 F 2 R (3.11)

B 5 B(h) ; Bh , 0 (3.12)

F 5 F(i) ; Fi . 0 (3.13)

e 5 W [hT 2 h ] (3.14)

In following the structuralist tradition, several variables are measured as ratios over the value of the capital stock ePMK: desired investment ( gd ) , total saving (gs), the current account defi cit (B), capital infl ows (F), the increase in international monetary reserves (R) and output (u). The nominal interest rate (i), the actual infl ation rate (p), the expected infl ation rate (pE), and the rates of growth of money supply and money demand (Ms, Md) are written in percentage terms. The real exchange rate, the wage share and the capital/labor force ratio are denoted by h, A and k respec-tively, and the symbols h1, h2, h3, h4, t0, t1, b0, b1, b2 all represent positive parameters.

Equation (3.1) defi nes total saving as domestic saving ‘sr’, plus foreign saving ‘B’. In Equation (3.2) desired investment depends on the profi t and interest rates. Then Equation (3.3) shows the rate of profi t in terms of the fl exible output/capital ratio, the real exchange rate and the wage share, and Equation (3.4) indicates that local prices remain fi xed in the short run. Goods market equilibrium is defi ned in Equation (3.5), while Equation


(3.6) shows the employment rate (l) as a function of the output/capital ratio (for given levels of K/N, denoted here as k).

In the monetary market equilibrium is attained when the rates of growth of money supply and money demand (in nominal terms) are equal. In Equation (3.7) money demand is determined by infl ation, the output/capital ratio, the nominal interest rate and an exogenous expected infl a-tion term. Agents face a dual decision regarding price increases: on the one hand, infl ation raises their demand for money as they attempt to maintain the value of their real balances. But, on the other hand, agents will also try to get rid of (that is, spend) their holdings of money if they expect price increases, the public will attempt to win the race against infl ation.

The money supply in Equation (3.8) grows with the accumulation of foreign exchange reserves, but decreases with the exogenous interest rate. This latter component attempts to formalize central bank policy making: in order to reduce money growth, the bank conducts open market opera-tions and drives interest rates up, and the opposite would happen if policy makers wished to increase money growth. This is more elaborate than the ‘Taylor rule’ that appears in the ‘new consensus macroeconomics’ (Arestis and Sawyer, 2003a; Romer, 2000): we argue here that changes in the interest rate will aff ect the money supply and this may or may not aff ect infl ation; while the Taylor rule (Taylor, 1993) is set up as a mechanical relationship between infl ation and the central bank’s reaction. Finally, the constant parameter t0 captures the eff ect of monetary policy that is not related to open market operations (changes in the legal reserve require-ment or the discount rate, for example).

Equation (3.9) presents money market equilibrium, and Equation (3.10) shows the exogenous nominal interest rate. Infl ation is here a monetary phenomenon: excess money supply growth leads to excess spending, and hence to higher infl ation. But also the lack of money supply results from high interest rates, which hurts employment and output growth. Thus, eff ective demand continues as a relevant determinant of economic activity. Here money is not neutral, so, unlike several versions of the ‘new consen-sus’ macroeconomics (Meyer, 2001), money does matter. Our approach also diff ers from the post-Keynesian view of infl ation as resulting from cost-push and social confl ict (as presented, for example, in Arestis and Sawyer (2006)). This model creates a bridge between monetarist notions of the money market, and Keynesian views on the importance of eff ective demand.

Finally, we need to keep in mind that, ceteris paribus, the use of the interest rate to control infl ation, will only lead to a stable adjustment process if money supply is more sensitive to changes in the interest rate than money demand (that is, t1 is bigger than h3 ). Only in this case will the


higher interest rate lead to a reduction in the excess money supply, and to a lower infl ation rate.3

The foreign sector is described in Equations (3.11) through (3.14). In Equation (3.11) the current account defi cit is fi nanced by capital infl ows and depletion of international monetary reserves. The current account defi cit depends negatively on the real exchange rate and net capital infl ows depend positively on the interest rate. In Equation (3.14) we see how, in the long run, the monetary authority devalues the nominal exchange rate to bring the real rate to its target (hT); Ω is a positive parameter.

3.2.4 Formal Solution of the Model

The process starts when the central bank defi nes the nominal interest rate and the target for the real exchange rate. These two policy variables are utilized, respectively, to control the money supply, and to maintain the competitiveness of net exports. The current account defi cit and net capital infl ows can then be determined, and with this we can solve for R in Equation (3.11).

In the goods sector Equations (3.1), (3.2), (3.12) and (3.5) determine the profi t rate. In Equation (3.3), given A and h, we fi nd the output/capital ratio, and Equation (3.6) then leads to the employment rate (l). Again in Keynesian fashion, employment is determined in the goods market (not in the labor market as monetarist and new classical approaches would suggest). The short-run solution for the goods sector is shown in Figure 3.1.

We may now move into the medium run and use Equations (3.7), (3.8), (3.9) and (3.10) in the monetary sector to solve for infl ation, as shown below:

p0 5R0 1 t0 1 (2 t1 1 h3) i0 2 h2u0 1 h4p

E

h1 (3.15)

where u0 represents the level of the output/capital ratio found in the goods market, and i0 denotes the interest rate fi xed by the central bank. The rela-tionship between the interest rate and infl ation is shown in Figure 3.2.

The money market equilibrium schedule (MM line) in Figure 3.2 has a negative slope when the conditions specifi ed in Note 3 hold. This schedule provides the combinations of infl ation and interest rate that are consistent with equilibrium in the monetary market. Here p0 is the infl ation rate that clears the money market when the central bank fi xes the interest rate at i0. In order to bring infl ation down, the central bank raises the interest rate to reduce money supply growth; as the availability of money declines, aggre-gate spending decreases and infl ation goes down (the economy moves to the left along the MM line).


g

g0

gd

gs

l l

u

l = auku

r

r

u0

u = rh1–A

Figure 3.1 Goods market equilibrium in the short run

�0

i0

�

i

MM

Figure 3.2 Money market equilibrium schedule


But, in the foreign sector, a higher ‘i’ will attract more capital fl ows, making international monetary reserves go up. The money supply will then increase (the MM line shifts out), compensating (at least partially) the initial attempt to reduce infl ation pressure. Thus the monetary authorities cannot control both the exchange rate and interest rate (and the money supply) under an open capital account.

Second, in the goods market, as ‘i’ goes up, investment demand goes down; the gd line in Figure 3.1 shifts down and the rate of growth, the rate of profi t, the output/capital ratio and the employment rate decrease. Stabilization is not too successful and hurts economic activity.

Fortunately, the central bank can decide to increase the target level for h, which will cause foreign savings (the current account defi cit) to go down. The gs line in Figure 3.1 will shift to the right (down) leading to faster growth, higher profi t rate, and higher output/capital ratio and employment rate.

We have thus confi rmed the double edged character of the real exchange rate: it is to blame for the diffi culties the central bank faces in controlling infl ation; had it had more fl exibility in the short run, there would had been no accumulation of international reserves and we would not have lost ground in the fi ght against infl ation. But, on the other hand, it is the ability of the central bank to move the real exchange rate to a targeted level, which provides the economy with the possibility to reach faster growth and higher employment rates. Arguments and evidence in favor of target-ing the real exchange rate may be found in Frenkel (2004) and in Frenkel and Ros (2006).

Now a fi nal comment on monetary policy: there is no reason for the central bank to only use open market operations to control the money supply. Other policy instruments, like the legal reserve requirement, could play an important control role without a direct impact on interest rates.4 A higher reserve requirement ratio may be seen in our model as a reduction in the term t0 in Equation (3.15). In Figure 3.2 this will show as a downward shift of the MM schedule, indicating that there will now be a lower level of infl ation at every level of the interest rate. The central bank may thus be able to curve down price growth, but without changes in international monetary reserves, breaking in this way the ‘trilemma’. The alternative presented here is also in line with the arguments in Frenkel (2004), who goes even further to point out that capital controls are feas-ible; a similar argument on the convenience of capital controls is provided by Ffrench Davis (2003).

Many central banks, however, would not like to use the legal reserve requirement, or capital controls as instruments for monetary policy: they prefer open market operations because they are, well, more market


oriented.5 Policy makers are thus opting for more fl exible exchange rate arrangements, combined with infl ation targeting regimes, as will be explained in Section 3.3.

3.2.5 Long-run Dynamics

We follow Cordero (1995) in analysing the interaction between the wage share (A) and the real exchange rate (h), which we assumed given in the short and medium run. But we have an additional state variable here: the capital/labor force ratio k, so this is really a 3 × 3 system. We assume that, in the long run, the variables which adjusted in the short run to clear the goods market (u, g, r, l ), and in the medium run to clear the monetary sector (p) remain at their equilibrium levels.

The motion of A is given by

A 5 W 1 a 2 p (3.16)

where the hats denote rate of growth. The coeffi cient a was defi ned as LQ so we can write that y 5

QL 5

1a where ‘y’ represents average labor

productivity.Following Arrow (1962) it is argued that labor productivity depends

on learning, which is assumed to be positively related to production. We also follow Dutt (1994) in assuming that production is associated to the investment rate ‘g’:

y 5 2a 5 Y( g 2 g) , Yg . 0 (3.17)

with Yg the partial derivative of productivity growth with respect to the investment rate, and g is an arbitrary constant number.

The specifi cation of wage growth assumes that workers try to increase their nominal wage whenever the actual wage share falls below an exog-enous level AW desired by workers as indicated in Equation (3.18):

W 5 q [AW 2 A ] (3.18)

Finally, the infl ation rate that clears the monetary market depends on the real exchange rate and the wage share:

p0 5 p (h, A) ; ph , 0; pA , 0 (3.19)

where ph, pA denote partial derivative of p with respect to h and A, respec-tively. The dependency of p0 on ‘h’ and ‘A’ comes from the presence of the


equilibrium output/capital ratio (u0) in Equation (3.15); and from the fact that this u0 was determined in the goods market (through the interaction between gd and gs for given h and A).

We may now put Equations (3.17) through (3.19) in (3.16) to get a dif-ferential equation in h and A:

A 5 q [AW 2 A ] 2 Y [g(h) 2 g ] 2 p [h, A ] (3.20)

The dynamic behavior of the real exchange rate is described by:

h 5 e 2 p (3.21)

According to Equation (3.14), reproduced below, the nominal exchange rate moves up when the actual level of the real exchange rate falls below the target chosen by the central bank:

e 5 W [hT 2 h ] (3.14)

Next use Equations (3.19) and (3.14) in Equation (3.21) to get:

h 5 W [hT 2 h ] 2 p [h, A ] (3.22)

We should recall now that we had a third state variable (k), but the system with Equations (3.20) and (3.22) may be solved independently of k, so we may proceed as if we had a 2 × 2 system with A and h as the state vari-ables.6 The stability of this system (Equations 3.20 and 3.22) is analysed by means of the Jacobian matrix:

Det(J) 5 ≥ 0A0A 0A0h

0h0A

0h0h

¥ 5 £ (2 q 2 pA) (2Yggh 2 ph)

2pA (2 W 2 pA)§

The determinant of the Jacobian is:

Det(J) 5 (q 1 pA) (W 1 ph) 2 pA(Yggh 2 ph)

5 q [W 1 ph ] 1 pA [W 2 Yggh ]

which will be positive if W . ph and W . Yggh. In other words, Det (J) will be positive if W is large.

The trace is given by Tr(J) 5 2q 2 pA 2 W 2 ph, which will be


negative if q, and Ω are large. So the long-run equilibrium will be stable if wages are fl exible and the central bank is very committed to maintaining the real exchange rate close to the chosen target. Finally, as the short-run equilibrium level of growth depends on h, and h is determined in the system formed by Equations (3.20) and (3.22), we may say that economic growth is endogenous in this model.

3.3 THE SMALL ECONOMY UNDER AN INFLATION TARGETING REGIME

3.3.1 The Infl ation Targeting Regime

In the previous section we saw the problems that may arise if, with a fi xed real exchange rate, the central bank attempts to control infl ation by using open market operations. Although, as we saw, there are several ways in which the ‘trilemma’ could be faced successfully, there is a strong tendency to move to an infl ation targeting regime and turn the exchange rate system into a fl exible one. This bias, according to Epstein (2005), is part of a global change in the practice of central banking, promoted mostly by institutions like the IMF. In this view central banking practices must be based on independence, infl ation as the goal of policy (including application of infl ation targeting regimes), and the use of indirect methods of monetary policy (mostly open market operations). Arestis and Sawyer (2003b) relate this bias to the ‘new consensus’ macroeconomics in which infl ation targets become the focus of monetary policy. In this section we want to emphasize that, once this approach is adopted, everything in the economy will be tied to the infl ation target; the result is great success in bringing infl ation down, but a rather disappointing performance in terms of growth and employment.

According to Galindo and Ros (2008) this monetary regime leads to possible appreciation of the real exchange rate, and increased vulner-ability to monetary shocks coming from the external sector. In the case of Mexico, these authors show that infl ation targeting has led to an appreci-ation of the real exchange rate, which has had a negative impact on output growth.

Epstein (2002) reports that the utilization of this monetary system in South Africa led to high interest rates, accompanied by low employment and investment rates. His recommendation is that South Africa move to a scheme in which employment is targeted, but subject to an infl ation goal. He also recommends the use of capital controls in order to better control the eff ect of the world economy on South Africa. Setterfi eld (2006) also


argues that governments should pursue output and employment targets, in addition to infl ation goals.

Other more general cross-country studies have found that there is no reason to keep infl ation below the 3 to 5 percent range, especially because in middle income economies higher single digit rates of infl ation could stimulate economic growth (Pollin and Zhu, 2005). Another study con-ducted by Ball and Sheridan (2003) concludes that infl ation targeting has generated no major benefi ts in economic performance (other than a decline in infl ation rates). In this section we develop a model to examine the eff ects of infl ation targeting on growth and employment.

3.3.2 Revised Equations

Money market

i 5 G(p) ; Gp , 0; p 5 p(i) , pi , 0 (3.23)

Ms 5 t0 2 t1i (3.24)

Md 5 h1p 1 h2u 2 h3i (3.25)

p 5 pT (3.26)

Ms 5 Md (3.27)

Foreign sector

B 5 F (3.28)

B 5 B(h) ; Bh , 0 (3.29)

F 5 F(i) ; Fi . 0 (3.30)

Goods sector

gd 5 gd (r, i) , gdr . 0, gd

i , 0 (3.31)

G 5 gd (r, i) 2 B(h) (3.32)

S 5 sr (3.33)

G 5 S (3.34)


PQ 5 PQ (3.4)

r 5uh

(1 2 A) (3.3)

l 5 auk (3.6)

In Equation (3.23) the interest rate is the instrument the central bank uses to control infl ation. Once a target is defi ned in Equation (3.26), open market operations allow bringing the interest rate to the level that is exactly needed to hit the chosen infl ation target. As for money demand (Equation 3.25), we use a formulation that is similar to the one in Section 3.2. Finally, and in accordance to this monetary system, we assume that the central bank announces an infl ation target in the understanding that this goal is prior to any other objective the bank might have. The priority assigned to hitting this level will be clear as the model is explained in more detail.

The monetary sector operates in the following manner. First, the central bank recognizes that the interest rate and the infl ation target are not chosen independently of one another: the bank knows that there is an inverse relationship between nominal interest rates and money supply. Thus, in order to fi ght infl ation, the bank conducts open market opera-tions (to reduce the money supply) and pushes interest rates up. In other words, higher interest rates, are associated to lower infl ation rates, and this is exactly what Equation (3.23) tells us.

Again this is clearly diff erent from the ‘Taylor rule’ that appears in the new synthesis models that we referred to in previous sections of this chapter. Our Equation (3.23) is a behavioral relationship indicating how infl ation reacts to changes in the nominal interest rate. The so-called Taylor rule is a more mechanical expression indicating how the central bank moves the interest rate when infl ation goes up. In the new synthesis models this rule replaces the monetary sector. In this chapter, however, we argue that economic authorities in less developed countries still believe that their policy decisions should be based on their perception of the mon-etary sector; hence there is no Taylor rule in this model, but there is a fully structured monetary market.

Next we turn to the foreign sector where we start by getting rid of the accumulation of international monetary reserves. As Equation (3.28) shows, the current account defi cit has to be fi nanced by net capital infl ows. This is made possible by allowing the real exchange rate to fl uctuate as required to satisfy Equation (3.28).7

In the goods sector the specifi cation looks a bit diff erent from Section 3.2, but it still holds the same meaning: Equation (3.32) defi nes demand


injections, while (3.33) defi nes domestic savings. Equations (3.3), (3.4) and (3.6) are borrowed from the previous section to describe prices, the rate of profi t and the employment rate.

3.3.3 Solution of the Model

Price rigidities are an important feature of this model: in the short run prices remain fi xed by Equation (3.4), and will adjust only in the medium run (the growth of prices in the medium run is determined by the infl ation target chosen by the central bank). In the short run the burden of adjust-ment falls on the output/capital ratio and the real exchange rate.

Once the infl ation target pT is announced, Equations (3.26) and (3.23) allow defi ning the nominal interest rate. In the foreign sector the real exchange rate adjusts to equilibrate the balance of payments and we have:

h 5 h [i ], hi , 0 (3.35)

Next with i and h known we move to the monetary sector: Equations (3.24), (3.25) and (3.27) provide an expression for money market equilibrium:

t0 2 t1i 2 h1p(i) 2 h2u 1 h3i 5 0 (3.36)

This equation describes a relationship between the interest rate and the output/capital ratio, and the slope is given by

0u0i

5[2 t1 2 h1pi 1 h3 ]

h2, 0

As before, the sign of this slope results from the conditions in Note 3 (Figure 3.3 presents the corresponding graph) and we write that:

u 5 u [i ], ui , 0 (3.37)

Once we know the infl ation target we can plug the known interest rate i0 in Equation (3.36) to fi nd the output/capital ratio (u) that clears the money market. This implies that, in order to derive u, it is not necessary that savings and investment be equal: the model is overdetermined.

In the next step we go to the goods sector, and examine the behavior of the rate of profi t. From Equation (3.3), and also using Equations (3.35) and (3.37) we may write the following expression:

r 5[u(i) ][h(i) ]

(1 2 A) (3.38)


and the response of the rate of profi t to changes in the interest rate is given by:

0r0i

5 ri 5(1 2 A)[h(i) ]2 [uih 2 uhi ] (3.39)

We do not have information to defi ne the sign of the term [uih 2 uhi ], so we can just examine two possible cases:

r 5 r [i, (1 2 A) ], ri , 0, rA , 0 (3.40a)

or

r 5 r [i, (1 2 A) ], ri . 0, rA , 0 (3.40b)

Signing the derivatives of r with respect to ‘i’ is very important as that will allow to also sign the derivatives of demand injections (G) and domestic saving (S) with respect to ‘i’.

Case 1

ri , 0, rA , 0

i

u

MM

Figure 3.3 Money market equilibrium schedule under infl ation targeting


From Equation (3.32) and from Equation (3.40a) we get:

G 5 gd{r [i, (1 2 A) ], i} 2 B [h(i) ] (3.41)

and the slope of this line is

0G0i

5 gdrri 1 gd

i 2 Bhhi , 0 (3.42)

Next from Equations (3.33) and (3.40a) obtain:

S 5 sr [i, (1 2 A) ] (3.43)

and the slope of the savings function is

0S0i

5 sri , 0 (3.44)

Just as in other Keynesian models, the slope of the S line has to be steeper than that of the G line in order to secure stability of equilibrium in the short run:

P 0G0i

P , P 0S0i

P (3.45)

The graphical solution appears in Figure 3.4. In the fi rst panel choos-ing an infl ation target fi xes the interest rate by means of Equation (3.23). The foreign sector appears in the second panel, where the known interest rate leads to the real exchange rate that clears the balance of payments. The monetary sector in the third panel shows that, given the interest rate, the ‘MM’ schedule helps us determine the output/capital ratio. Then we examine the behavior of demand injections (G) and domestic saving (S), and overdetermination becomes clear. If the interest rate is below the one that brings G and S to equality, then the economy will be at a point on the G line. Any reduction in the infl ation target (and thus increase in the inter-est rate) will lead to real appreciation, higher unemployment and a lower rate of economic growth. A reduction of the infl ation target will reduce the level of activity even if the interest rate is to the right of the one that makes G and S equal. As the nominal exchange rate is no longer a policy variable, the central bank cannot stimulate economic activity and reduce the infl ation rate at the same time.

Case 2

ri . 0, rA , 0


Here the G function keeps the negative slope, but the S function will defi -nitely have a positive slope in the S versus i plane, as shown in Figure 3.5.

With the interest rate at the given level i0 (which is, again, lower than the one that brings G and S to equality), the economy must be at a point on the

l

u l = uak

ii0

�

h

�T

� = p (i)

i

B (h) = F (i)

u

i

MM

G, S

ii0

GS

Figure 3.4 An open economy under infl ation targeting and overdetermination in the goods sector


S schedule (investors cannot satisfy their desires). A reduction in the infl a-tion target will require a rightward movement of the interest rate, with the economy moving along the domestic saving line towards the right and to a higher rate of growth. But notice that investment desires will suff er a down-ward movement along the G schedule. So what we are observing here is that, as infl ation is brought down, the investment drive also declines. Thus in this case we see that within the infl ation targeting regime, overdetermination causes a gap between actual and desired investment. When the given inter-est rate lies to the right of the level that brings domestic savings and demand injections to equality, the economy is positioned at a point on the G sched-ule; in this case stabilization policies lead to a lower rate of growth.

3.3.4 Long-run Dynamics

In the long run we analyse two state variables: A and k, where k repre-sents the capital (K) to labor force (N) ratio. The analysis here is based on Cordero (2002).

The motion of A is described by Equation (3.16) reproduced below:

A 5 W 1 a 2 p (3.16)

G, S

ii0

G

S

Figure 3.5 Overdetermination on the goods sector when savings slopes up


and for the technical coeffi cient ‘a’ we use Equation (3.17) also reproduced below:

y 5 2a 5 Y(g 2 g) , Yg . 0 (3.17)

The growth of nominal wages is again depicted by Equation (3.18)

W 5 q [AW 2 A ] (3.18)

but the desired wage share is now endogenized and made dependent on the employment rate:

AW 5 a0 1 a1aLNb 5 a0 1 a1aL

Q QK

KNb 5 a0 1 a1 (auk) (3.46)

Finally, we recall that the local infl ation rate was set by a target defi ned by the central bank, according to Equation (3.26):

p 5 pT (3.26)

The use of Equations (3.17), (3.18), (3.46), (3.37) and (3.26) in Equation (3.16) leads to a diff erential equation for the evolution of A:

A 5 qa0 1 qa1au [i ]k 2 qA 2 Y [g 2 g ] 2 pT (3.47)

The motion of k is described by:

k 5 K 2 N

where K is the investment rate (g), and N is assumed fi xed at a level denoted by ‘n’:

k 5 g 2 n (3.48)

In order to fi nish writing Equations (3.47) and (3.48) in terms of A and k, we have to replace g with its short-run equilibrium value.8 Before we proceed we must generate expressions for the desired investment rate and total savings. The fi rst of these is provided by Equation (3.31) which, after using Equation (3.40a), becomes:

gd 5 gd{r [i, (1 2 A) ], i}, gdr . 0, gd

i , 0, ri , 0, rA , 0 (3.49)


Total savings, on the other hand, is defi ned as domestic savings (sr) plus foreign savings [B(h)]. From Equations (3.33), (3.40a) and our defi nition of foreign savings, we get

gs 5 s{r [i, (1 2 A) ] } 1 B [h ], s . 0, ri , 0, rA , 0 (3.50)

For the case depicted in Figure 3.4 the g level is determined by the corre-sponding value of gd in Equation (3.49).9

The system formed by Equations (3.47) and (3.48) may be rewritten as:

A 5 qa0 1 qa1au [i ]k 2 qA 2 Y [gd{r [i, (1 2 A) ], i} 2 g ] 2 pT (3.51)

k 5 gd{r [i, (1 2 A) ], i} 2 n (3.52)

The stability of the system is analysed by means of the Jacobian matrix:

Det(J) 5 ≥ 0A0A 0A0k

0k0A

0k0k

¥ 5 £ (2q 2 YggdrrA) qa1au [i ]

gdrrA 0

§We have:

Det(J) 5 2 [gdrrA ] [qa 1au [i ] ] . 0

and

Tr(J) 5 (2q 2 YggdrrA)

Since rA , 0, the long-run equilibrium will be stable when nominal wage fl exibility is important (q large) compared to learning eff ects (Yg). This does not mean that the economy has to stay away from productivity gains; it means instead that wage fl exibility is necessary to accommodate produc-tivity growth.

3.4 CONCLUDING REMARKS

In this chapter we compare the eff ects of diff erent monetary regimes on the growth and infl ation performance of an open economy. With a real exchange rate target the decisions of the central bank are aff ected by the ‘trilemma’ of monetary policy, but the real exchange rate can still be


utilized to push employment and economic growth. Thus, although stabil-ization based on open market operations has a tendency to lower economic activity, the central bank can always target a level of the real exchange rate which compensates, at least partially, the contractionary eff ects of stabil-ization. The use of monetary instruments like the legal reserve requirement may help bring infl ation down without the problems associated with the ‘trilemma’. The long-run equilibrium is stable if nominal wages are fl ex-ible, and the central bank maintains a competitive real exchange rate.

Under an infl ation targeting regime, the problems of the ‘trilemma’ disappear and prices are more easily stabilized. In this setting all eco-nomic results depend on monetary policy: the goods sector is not neces-sary to fi nd equilibrium levels for the relevant variables. This situation causes overdetermination as there is no level of the rate of profi t which is capable (except by accident) of bringing saving and investment to equal-ity. Stabilization policy is harmful to growth and employment and, even if savings increased, investment desires remain unsatisfi ed. The infl ation targeting regime leads to a stable long-run equilibrium if nominal wages are fl exible enough to accommodate the productivity gains resulting from learning.

The most important result in this chapter is the recognition of the criti-cal importance that the real exchange rate has on the determination of the rates of accumulation and employment of the small open economy. Thus countries which have already decided to embark on infl ation targeting regimes should allow themselves to tolerate higher infl ation goals to avoid excessive appreciation of the real exchange rate.

For countries which still fi nd themselves battling the diffi culties of the ‘trilemma’ the recommendation would be to combine the use of open market operations with the legal reserve requirement or even with some control over the capital account of the balance of payments.

NOTES

1. The structuralist approach is described in Taylor (1991) and in Dutt (1992).2. In other structuralist models (Dutt, 1984; Taylor, 1985, for example) income distribution

plays a critical role in the determination of economic growth. Blecker (1989) then extends those models to examine how income distribution aff ects growth in an open economy. In this chapter, however, the investment function (our Equation (2)) looks more like the NeoKeynesian version used by Marglin (1984), and the ouput/capital ratio has been taken out of the desired accumulation function. This modifi cation prevents the wage share from having an eff ect on the rate of growth.

3. The adjustment mechanism may be described by 0i/0t 5 f (Ms 2 Md) , f . 0; and this will be stable only if f (2t1 1 h3) , 0.

4. This point was analysed with a formal model in Cordero (2005).


5. The mainstream view argues that higher levels of the legal reserve requirement cause higher administrative costs, and thus may lead to higher spreads and margins for fi nancial intermediation. But this is not necessarily true: in the case of Costa Rica, for example, the reserve requirement fell from 15 percent in 1999 to 5 percent in 2002, but in the same period the spread between active and passive interest rates declined only half a percentage point. Then, in 2003, when the reserve requirement jumped from 5 percent to 10 percent, the spread went up by only 1.5 percent (Cordero, 2005).

6. The long-run motion of k is given by: k 5 g 2 N. The short-run equilibrium level of g depends on h: g0 5 g (h) , gh . 0. Assume that N 5 n (k) with nk . 0, and we get k 5 g (h) 2 n (k) . The system formed by Equations (3.20) and (3.22) provides a solution for h, which we use to fi nd the level of k required for k 5 0. The equilibrium is stable if nk . 0.

7. With prices fi xed in Equation (3.4), h adjusts as a result of variations in the nominal exchange rate e.

8. Of course we also analysed a second case which here would require that we plug Equation (3.40b) in Equation (3.31) for the gd. expression, and in Equation (3.33) for the gs. equation. The long-run dynamics will be analogous to that of the fi rst case, but will not be presented in the chapter.

9. If the interest rate that is consistent with the infl ation target falls to the right of the level making G equal to S, then the g level is determined by the corresponding value of gs in Equation (3.50). Again the dynamics of this case is analogous to the one resulting from Figure 3.4 and will not be analysed here in detail.

REFERENCES

Arestis, P. and M. Sawyer (2003a), ‘New consensus, new Keynesianism, and the economics of the Third Way’, Levy Economics Institute of Bard College working paper no. 364.

Arestis, P. and M. Sawyer (2003b), ‘Infl ation targeting: a critical appraisal’, Levy Economics Institute of Bard College working paper no. 388.

Arestis, P. and M. Sawyer (2006), ‘The nature and role of monetary policy when money is endogenous’, Cambridge Journal of Economics, 30 (6), 847–60.

Arrow, K. (1962), ‘The economic implications of learning-by-doing’, Review of Economic Studies, 29 (3), 155–73.

Ball, R. and N. Sheridan (2003), ‘Does infl ation targeting matter?’, National Bureau of Economic Research working paper no. 9577, Cambridge, MA.

Blecker, R. (1989), ‘International competition, income distribution and economic growth’, Cambridge Journal of Economics, 13 (3), 395–412.

Cordero, J. (1995), ‘Essays on growth and distribution for open economies’, unpublished PhD dissertation, University of Notre Dame, Indiana.

Cordero, J. (2002), ‘A model of growth and confl ict infl ation for a small open economy’, Metroeconomica, 53 (3), 261–89.

Cordero, J. (2005), ‘Infl acion, politica monetaria y regimen cambiario en Costa Rica’, in J.R. Vargas and Y. Xirinachs (eds), La formación de economistas: Ensayos en honor de Pepita Echandi, San Jose, Costa Rica: Posgrado en Economia, Universidad de Costa Rica.

Dutt, A. (1984), ‘Stagnation, income distribution and monopoly power’, Cambridge Journal of Economics, 8 (1), 25–40.

Dutt, A. (1992), ‘Two issues on the state of development economics’, in A. Dutt and K. Jameson (eds), New Directions in Development Economics, Aldershot, UK and Brookfi eld, VT, USA: Edward Elgar, pp. 1–34.


Dutt, A. (1994), ‘On the long run stability of capitalist economies’, in A. Dutt (ed.), New Directions in Analytical Political Economy, Aldershot, UK and Brookfi eld, VT, USA: Edward Elgar, pp. 93–120.

Epstein, G. (2002), ‘Employment-oriented central bank policy in an integrated world economy: a reform proposal for South Africa’, Political Economy Research Institute working paper series no. 39, University of Massachusetts, Amherst.

Epstein, G. (2005), ‘Central banks as agents of economic development’, Political Economy Research Institute working paper series no. 104, University of Massachusetts, Amherst.

Epstein, G. and E. Yeldan (2008), ‘Infl ation targeting, employment creation and economic development: assessing the impacts and policy alternatives’, International Review of Applied Economics, 22 (2) (March), 129–30.

Ffrench Davis, R. (2003), ‘Macroeconomic balances in emerging economies: the confl ict between purely-fi nancial and real-economy macrobalances’, paper pre-pared for the Task Force on Macroeconomic Policy at the Initiative for Policy Dialogue, Columbia University, New York.

Frenkel, R. (2004), ‘Real exchange rate and employment in Argentina, Brazil, Chile and Mexico’, paper prepared for the G24, Centro de Estudios de Estado y Desarrollo, Argentina.

Frenkel, R. and J. Ros (2006), ‘Unemployment and the real exchange rate in Latin America’, World Development, 34 (4) (April), 631–46.

Galindo, L. and J. Ros (2008), ‘Alternatives to infl ation targeting in Mexico, International Review of Applied Economics, 22 (2), 201–14.

Marglin, S. (1984), Growth, Distribution, and Prices, Cambridge, MA: Harvard University Press.

Meyer, L.H. (2001), ‘Does money matter?’, Federal Reserve Bank of St Louis Review, 83 (5), 1–16.

Pollin, R. and A. Zhu (2005), ‘Infl ation and economic growth: a cross country non-linear analysis’, Political Economy Research Institute working paper no. 109, University of Massachusetts, Amherst.

Romer, D. (2000), ‘Keynesian macroeconomics without the LM curve’, Journal of Economic Perspectives, 14 (2), 149–69.

Setterfi eld, M. (2006), ‘Is infl ation targeting compatible with Post Keynesian eco-nomics’, Journal of Post Keynesian Economics, 28 (4) (Summer), 653–72.

Taylor, J. (1993), ‘Discretion versus policy rules in practice’, Carnegie-Rochester Conference Series on Public Policy, 39, 195–214.

Taylor, L. (1985), ‘A stagnationist model of economic growth’, Cambridge Journal of Economics, 9 (4), 383–403.

Taylor, L. (1991), Income Distribution, Infl ation and Growth: Lectures on Structuralist Macroeconomic Theory, Cambridge, MA: MIT Press.

PART II

Thematic issues: class relations and gender impacts of infl ation targeting

71

4. Income, class and preferences towards anti-infl ation and anti-unemployment policiesArjun Jayadev1

4.1 INTRODUCTION

From one important point of view, indeed, the avoidance of infl ation and the maintenance of full employment can be most usefully regarded as confl icting class interests of the bourgeoisie and the proletariat respectively, the confl ict being resolvable only by the test of relative political power in the society.

Harry Johnson (Johnson, 1968, p. 986)

Among the casualties of the advent of the rational expectations revolu-tion in the 1970s was a rich vein of political economy which analysed the macroeconomic dynamics of unemployment and infl ation as deriving from distributive struggles between capitalists and workers (Bach and Stephenson, 1974; Boddy and Crotty, 1975; Hibbs, 1977; Rosenberg and Weisskopf, 1981; Rowthorn, 1977). This approach often made explicit the class confl ict innate in Keynesian accounts of the Phillips curve trade-off and in considerations of the natural rate of unemployment.2 While the details of these studies varied, the argument, with slight modifi cations, remained the same: workers and capitalists had opposing and irreconcil-able diff erences in the trade-off between infl ation and unemployment. Such ‘confl ict theories’ not only gave political explanations for the trajec-tory of infl ation and unemployment (most notably in the work of Hibbs (1977)) but also thereby provided direct predictions for the preferences of individuals towards anti-unemployment and anti-infl ation policies based on their position within the social stratifi cation. Specifi cally, the working class, broadly classifi ed, is more concerned about reducing unemployment than fi rm owners. Unemployment reduces the lifetime income of workers directly and also exerts a downward pressure on wages by reducing the bargaining power of workers.3 Infl ation, on the other hand, worked to the advantage of those workers with low savings.

With the rise of new classical macroeconomics in the 1970s and its


subsequent hegemony any trade-off between infl ation and unemploy-ment came to be seen as essentially short term and certainly could not be utilized by policy makers to aff ect macroeconomic outcomes without incurring severe macroeconomic costs. As a consequence, much of the macroeconomic literature moved away from class-based, political economy models of infl ation and unemployment towards what Iversen and Soskice (2006, p. 425) elegantly call macroeconomics which ‘[f]ocuses attention on what democratic governments can do wrong in the short term’. The prescription that followed was to replace government with independent central banks, and discretionary macroeconomic policy with rules-based approaches such as infl ation targeting (Barro and Gordon, 1983; Cukierman, 1992).

While post-Keynesian and heterodox approaches never abandoned the idea of demand management, the advent of New Keynesian economics restored the space in mainstream economics for economic policy to have benefi cial macroeconomic outcomes. As opposed to rational expectations, there is a role for demand management and other policies when imperfec-tions arise due to wage and price rigidities whereby involuntary unem-ployment can be reduced. The now long literature on the non-accelerating infl ation rate of unemployment (NAIRU) continues to suggest that com-bating infl ation and unemployment involves two independent and poten-tially opposing targets.4

Given that there has been a restitution of space for macroeconomic policy, this chapter seeks to make a contribution in reconstructing the distributional consequences of anti-infl ationary and anti-unemployment policies, and attitudes towards them. Recent research on the impact of such policies on the poor and the rich has used social survey data to answer part of this question. Scheve (2004) and Jayadev (2006) both fi nd strong results that the relatively poor are less likely than the relatively rich to prioritize tackling infl ation as opposed to tackling unemploy-ment, presumably because they take the latter to be a more important problem facing them. Other research (Jayadev, 2008) continues this line of research by assessing the class content of disinfl ationary and expansionary policy. This research fi nds support for the contention that those in contrary class positions respond very diff erently to policies engineered to combat infl ation versus those that are designed to combat unemployment. There are substantial class-based diff erences in what may be termed ‘relative aversion’ to infl ation and unemployment (a preference that policy is designed to keep infl ation down rather than unemployment down). Capitalists and highly skilled workers are more likely to display relative infl ation aversion than less skilled and unskilled workers. I briefl y extend the analysis to look at the relationship between

Anti-infl ation and anti-unemployment policies 73

relative infl ation aversion and other attitudes towards macroeconomic policy. Individuals who support broadly pro-labor, solidaristic and redistributionary policies for the government are signifi cantly less rela-tively infl ation averse (or what is the same thing, more relatively unem-ployment averse).

This chapter summarizes the fi ndings of these two lines of research (that is, the income and class-based preferences for combating infl ation rather than unemployment). As such, the chapter provides some indi-cation for the degree and kind of political support that can be drawn for policies that aim to combat infl ation versus those which combat unemployment. The results have implications for the debate on anti-infl ationary policies and infl ation targeting in particular as they apply to diff erent countries.

I very briefl y summarize relevant research on the income and class impacts of infl ation and unemployment and attitudes towards these prob-lems in Section 4.2. Section 4.3 describes the data that I use. Section 4.4 presents the results of various logistic regressions. Section 4.5 summarizes and concludes.

4.2 INCOME AND ATTITUDES TOWARDS INFLATION AND UNEMPLOYMENT

The theoretical impetus for the shift in focus mentioned in Section 1 was provided by the rational expectations revolution of the 1970s and sub-sequent neo and new classical analyses of the state which suggested that expansionary policies were short-term palliatives at best, and in the medium to long run, simply infl ationary. That is to say, expansion, when driven ‘artifi cially’ by the state using fi scal or monetary authorities, may well have a positive eff ect on the poor in the short run by increasing the growth and employment rates in the economy; but in the long run will simply lead to more infl ation and a fall in employment back to its natural rate. The impact on the poor, it follows, is possibly a net deterioration in their welfare, as unemployment remains at previous levels and goods become more costly. Thus the standard prescription of expansionary policy in the 1960s and 1970s was replaced by the orthodoxy of infl ation targeting in the 1980s and 1990s, at least partly as a manner in which the poor were to be protected from a perhaps well meaning, but misguided government.

The theoretical linkages between infl ation and poverty do not, however, lead to the direct conclusion that infl ation is necessarily bad for the poor. A brief account of the channels by which expansionary and potentially infl ationary policy aff ects the poor as identifi ed in Romer and Romer


(1998), Easterly and Fischer (2001) and Cardoso (1992) among others is summarized in Table 4.1.

Given these ambiguous theoretical relationships, both at the macroeco-nomic and microeconomic levels, the link between infl ation and the poor is strongly an empirical matter, and certainly highly dependent on con-ditioning structural factors.

One of the more infl uential papers linking anti-infl ationary policy and the poor is that by Easterly and Fischer ( 2001). In a novel analysis the authors turn to an unimpeachable authority on the subject of infl ation

Table 4.1 Some hypothesized linkages between infl ation and poverty

Channel Eff ect on welfare of poor

Short run

Growth Expansionary demand policy, in raising average income (without a change in the distribution of income) has a direct eff ect on reducing poverty.

Real wage and transfers To the extent that wages and transfers are unindexed to infl ation, and that certain commodities (for example, food) form a larger portion of the consumption basket of the poor, the real income of the poor falls.

Medium to long term

Growth Romer and Romer (1998) suggest a strong positive correlation between high infl ation and low growth, as well as high infl ation and instability of demand, thus suggesting that short-run gains for the poor from expansionary demand policy is eroded in the long run.

Credit market It is assumed that unanticipated infl ation is positive for debtors and negative for creditors. Given that the poor fall overwhelmingly in the former category, it should be the case that they benefi t to some extent. However, the size of this eff ect depends on how much the poor owe and how much the real wage declines.

Uncertainty Uncertainty caused by infl ation causes fi nancial market dislocations, increases capital fl ight and reduces the rate of growth through reducing physical and human capital investment; all of which have a negative impact on the poor.


and the poor – the poor themselves. Easterly and Fischer use a large social survey conducted in 38 countries in 1996 to analyse the attitudes towards infl ation of individuals at diff erent income levels. Their study suggests a robust eff ect: the self-identifi ed poor were more likely than the self- identifi ed rich to mention infl ation5 as among the top three concerns for the economy. They fi nd, by contrast, that there is no statistically signifi cant diff erence between the rich and the poor in terms of their attitude towards recession and unemployment. From their study it is easy to draw the conclusion that anti-infl ationary policy is preferred by the poor to policy designed to combat unemployment or recession. Expansionary policy which leads to infl ation can be thought of as being akin to a tax:6 redistri-bution from the public at large to facilitate increased public expenditure. As such, asking about people’s preferences towards infl ation without presenting the alternative is similar to asking them about their preferences towards lowering taxes.

From the macroeconomic policy viewpoint, an important and relevant opportunity cost of lower infl ation is higher unemployment and informal-ization, especially when it is achieved by reducing government spending. As such, a more useful question to ask would be to ask the poor and the rich their relative preferences between infl ation and unemployment. Ideally, the question that should be posed is of the form: ‘if the govern-ment was to reduce infl ation, even if it meant increasing unemployment a little, would you support it?’ While the Roper Starch Survey used by Easterly and Fischer does not unfortunately provide such an option, there is another dataset which poses the relevant question more closely and directly (more on which in Section 4.3). As a result, I use this survey to analyse the attitudes of the poor towards infl ation and unemployment.

4.3 CLASS AND ATTITUDES TOWARDS INFLATION AND UNEMPLOYMENT

The radical political economy approaches to macroeconomics mentioned in the introduction saw unemployment as acting as ‘a regulator of class confl ict’ (Rowthorn, 1977). Unemployment was the central fulcrum of the labor-capital confrontation. Specifi cally, increases in unemployment maintained a downward pressure on wages while tight labor markets increased factor income going to labor by exerting upward pressure on wages (Boddy and Crotty, 1975).7

Confl ict theories of infl ation suggested that the rise in infl ation was seen as the result of perpetual claims by workers for wage increases ahead of productivity. The detrimental impact of infl ation was primarily on those


who had nominally denominated assets whose value was eroded with increasing prices, although there were also pressures on fi rms which were unable to pass on higher wages as higher prices. To that extent the nega-tive impact of infl ation was more pronounced on the capitalist and rentier class. As a result these theories would predict pronounced diff erences in attitudes towards infl ation and unemployment depending on the respon-dent’s class position.

Drawing from this political economy research, empirical studies in the 1970s and 1980s had begun to establish the class character of individuals’ preferences for infl ation versus unemployment aversion. Given objective evidence that periods of relatively low unemployment and relatively high infl ation coincided with an equalization of the personal distribution of income, a larger share of national income going to labor versus capital and a reduction in poverty as well as losses to those with savings in nominally denominated assets (typically the rich), attitudes towards infl ation and unemployment had a class character. Hibbs (1977, p.1470) summarizes the central fi ndings from US and UK surveys:

Popular concern about unemployment and infl ation is class-related. Low and middle income and occupational status groups are more averse to unemploy-ment than infl ation, whereas, upper income and occupational status groups are more concerned about infl ation than unemployment . . . it does appear that the subjective preferences of class or status groups are at least roughly in accord-ance with their objective economic interests . . .

Recent studies of attitudes towards infl ation and unemployment have largely ignored class. While there is substantial empirical evidence from opinion research that both infl ation and unemployment are seen by respondents as disutilities (see, among others, Di Tella et al., 2001; Easterly and Fischer, 2001; Shiller, 1997), there is less consensus on the relative importance that individuals in diff erent classes place on reducing each of these. Part of the issue is simply that there have been few surveys done which explicitly ask the respondent to rate their aversion to infl ation versus unemployment if these were alternative outcomes. As such, the data has limited researchers’ agenda. Equally, there has been little interest in the characteristics of individuals who might support a policy designed to combat infl ation versus one that tackled unemployment. To the extent that this has been undertaken, it has been to assess infl ation aversion among the rich versus the poor (Easterly and Fischer, 2001; Jayadev, 2006).8 There has been little or no recent work which attempts to look at politics and in particular class politics in the determination of preferences towards anti-infl ation and anti-unemployment policy. It is to this exercise that we now turn.


4.4 DATA

4.4.1 Measuring Infl ation Aversion

The International Social Survey Program (ISSP) conducted by the Inter-university Consortium for Political and Social Research in 1996 focuses on the preferences of more than 30 000 individuals in 279 diff erent countries regarding the role of government in society. The countries included include OECD economies, former eastern bloc economies and unfortunately do not include any low income countries apart from China.10 Among the questions asked in this survey is the following:

If the government had to choose between keeping down infl ation and keeping down unemployment to which do you think it should give highest priority?

This is the key variable of analysis for the rest of the chapter. I defi ne a variable ‘relative infl ation aversion’ as taking a value of one when the respondent prefers that the government prioritize reducing infl ation rather than unemployment and zero when the opposite holds. Summary statistics for all variables used are provided in Table 4.1.

While this provides a direct measure of the weights placed in an indi-vidual welfare function on infl ation as opposed to unemployment, it is not without some limitations. Ideally, such a question might ask how much infl ation the individual might accept for reducing the level of unemploy-ment and vice versa so as to have a more direct calibration of the marginal rate of substitution in the social and individual welfare function. However, the measure is certainly superior to questions which ask about infl ation without reference to unemployment or any other macroeconomic policy objective, thereby providing no implicit budget constraint.

Figure 4.1 summarizes the average preference for keeping infl ation rather than unemployment down by country.

Some interesting observations suggest themselves. Nearly 42 percent of the overall sample report being relatively infl ation averse. However, this masks large diff erences in the average relative infl ation aversion between countries, from a low of less than 20 percent of respondents in France to a high of above 60 percent of respondents in the Czech Republic. In only fi ve countries out of 20 is the percentage of relatively infl ation averse respon-dents over half (and only in two countries – West Germany and the Czech Republic – is the percentage overwhelmingly above the midway mark).

In order to assess the relative infl ation aversion of the rich and the poor, I use the income question of the survey which asks the respondent their personal income in the last year. The income of the individual is grouped


according to within country quintiles and quintile dummies take a value of 1 if the income of the respondent is in that quintile (and zero otherwise). I also defi ne a ‘relative infl ation aversion’ dummy that takes the value of 1 if the respondent prefers the government to keep down infl ation rather than unemployment.

4.4.2 Measuring Class

Constructing readily comparable and objective measures of social class are fraught with diffi culties (see Leiulfsrud et al. (2005) and Wright (1997) for an exposition on some of these). Contemporary measures of class diff er based on the elements that researchers consider important to their theoretical approach – for example, along lines of ownership, manage-ment, career prospects, income, status, education or other such categories (Carchedi, 1977; Erikson and Goldthorpe, 1993; Esping-Andersen, 1992; Wright, 1985). However, not all of these are likely to bear directly upon the question of relative infl ation aversion. Class has relevance in as much as it refl ects the labor market position of the respondent and therefore their preferences to policies which enhances their position. It is desirable,

18.3 20

.6

21.5

22.4

30.0 34

.0

34.9 39

.5 42.4 44

.0 46.3

47.0

47.1 49

.5

50.2 53

.4

54.8

61.8 64

.9

41.547

.6

0

10

20

30

40

Per

cent

age

(%)

50

60

70

Franc

e

Israe

l-Ara

bsSpa

inIta

ly

Latvi

a

Irelan

dChin

a

Canad

a

Austra

lia

Russia

Israe

l-Jew

s

United

Sta

tes

Poland

New Z

ealan

d

Sloven

ia

Germ

any (

East)

Japa

n

Germ

any (

Wes

t)

Hunga

ry

Czech

Rep

ublic

Overa

ll

Figure 4.1 Percentage of respondents who prefer that the government keeps infl ation down rather than unemployment down


therefore, to utilize a defi nition which closely refl ects the respondent’s occupation and position as employer or worker.

In this chapter, I utilize a few diff erent measures of class. The ISSP dataset provides an occupational variable based on the International Standard Classifi cations of Occupations (ISCO) 1988 classifi cation which has previously been used by researchers to construct a wide array of meas-ures of stratifi cation. A fi rst (crude) method is to utilize the traditional Marxist division between fi rm owner and employee. If the respondent is classifi ed as self-employed with employees, they are classifi ed as a fi rm owner, and an employee otherwise. I defi ne a dummy variable called fi rm owner which takes the value of 1 if the respondent is self-employed and has more than one employee, and 0 otherwise.

Sociologists have expended enormous eff ort in providing more useful and sophisticated categorizations of class. In order to perform a more sat-isfactory class analysis, I replicate Wright’s (1985) scheme which divides the labor force into owners and wage laborers and wage laborers in turn into three categories – experts, skilled and low skilled. Wright uses this to operationalize his idea of class locations and contradictions therein. Workers may be divided according to their relative privilege in the labor process. Based on a careful cataloging of occupations, Wright defi nes experts as those whose jobs require skills (and in particular accredited or

credentialed skills) and who are in scarce supply relative to their demand

by the market. Semi-skilled and unskilled class positions by contrast are held by those who have uncredentialed or no skills and who are thus in abundant supply. Using this approach has signifi cant advantages. It makes theoretical sense for the question at hand to conceive of class measures which refl ect the respondent’s relationship to the labor market and therefore to their bargaining power and probability of continued employment. An expert, for example, will typically enjoy a credential rent and be more likely to have both a higher level of bargaining power and a lower probability of being replaced than a low-skilled worker. They may therefore have opposing ideological and political interests based on other workers. At the same time both an unskilled worker and a skilled worker are more concerned about unemployment than a capitalist.11 Appendix 4.A1 details the creation of the class variables.

Figure 4.2 shows the average relative infl ation aversion in each country for each grouping of wage laborers. As is evident, in most coun-tries ‘experts’ are more relatively infl ation averse than semi-skilled and unskilled workers,12 as might be expected given the logic that more highly skilled workers enjoy greater bargaining power and a lower probability of unemployment than lower skilled workers, but are equally likely to see their wages eroded by infl ation.


Another approach is to look at subjective evaluations of class cat-egories. The ISSP dataset asks respondents their own evaluation of their social class (the categories are lower middle class, upper working class, middle class, upper middle class and upper class). Unlike more objective measures, subjective perceptions of class probably depend on an amalgam of factors such as the respondent’s education, status, income and gender as well as factors which are of direct relevance to the question of relative infl ation aversion – the respondent’s position and prospects in the labor market and the asset market. Nevertheless, it is useful to look at this as a check on the robustness of the earlier measure of class. I defi ne three subjective class categories – the variable subjective lower class takes a value of 1 when the respondent identifi es as being in the lower class or in the working class and 0 otherwise. Another variable subjective middle class takes a value of 1 if the respondent is from the lower middle class or middle class and 0 otherwise. Finally, the value of the variable subjective upper class takes a value of 1 when the respondent is from the upper middle or upper class.

4.5 RESULTS AND EXTENSIONS

In order to assess the income and class character of relative infl ation aver-sion, I undertake a series of logistic regressions of relative infl ation aver-sion on the respondent’s class position. Tables 4.2, 4.3, 4.4 and 4.5 provide the detailed results of these exercises.

0

10

20

30

40

50

60

70

80

90

Austra

lia

Canad

a

China

Czech

Rep

ublic

East G

erm

any

Franc

e

Hunga

ry

Irelan

d

Israe

l Ara

bs

Israe

l Jew

s

Latvi

a

New Z

ealan

d

Poland

Russia

Slovak

iaSpa

in

United

Sta

tes

Wes

t Ger

man

y

Per

cent

age

(%)

Skilled workersLow-skilled workers

Experts

Figure 4.2 Percentage of each category who are relatively infl ation averse


Table 4.2 Logistic regression on the likelihood that a respondent will prefer the government to keep infl ation down rather than unemployment down

Variable Coeffi cient z-statistic

Lowest income quintile 0.56* (−10.5)Second income quintile 0.67* (−6.6)Third income quintile 0.76* (−4.8)Fourth income quintile 0.82* (−3.4)Number of observations 14 340

Note: * 5 signifi cant at the 1 percent level. Omitted variable is richest quintile. Country intercept dummies are included but not shown.

Table 4.3 Logistic regression on the likelihood that a respondent will prefer the government to keep infl ation down rather than unemployment down, with controls

Variable Coeffi cient z-statistic

Lowest income quintile 0.67* (−6.5)Second income quintile 0.75* (−4.5)Third income quintile 0.83* (−3.3)Fourth income quintile 0.87** (−2.3)Some primary education 0.89** (−2.1)Some secondary education 0.86* (−3.1)Person in 20’s 1.31** (2.3)Person in 30’s 1.52* (3.6)Person in 40’s 1.29* (2.2)Person in 50’s 1.25*** (1.9)Person in 60’s 1.43* (3.0)Person in 70’s 1.30** (2.0)Person above 70’s 1.26 (1.4)Unemployed 0.82* (−5.4)Female 0.72* (−3.6)Trade union member 0.81* (−4.3)Voted for right wing party 1.39* (7.6)Number of observations 14 340

Note: * 5 signifi cant at the 1 percent level, ** 5 signifi cant at the 5 percent level, *** 5 signifi cant at the 10 percent level. Omitted variable is ‘richest quintile’ for the income variables, ‘at least some university education’ for the education variables and ‘persons less than 20 years of age’ for the age variables. Country intercept dummies are included but not shown.


Table 4.2 reports the results from a logistic regression of the relative infl ation aversion dummy on the income quintile dummies. The fi fth, or richest, quintile is the omitted variable, so that the coeffi cients on the income quintile dummies measure the diff erence between the coeffi cient on that income category and on the richest quintile. In order to control for country diff erences in the national averages of relative infl ation aversion, the regression includes average individual country relative infl ation as dummies, but these are not shown in the table.

A strong and consistent result is immediately evident. Relative infl ation aversion (or the likelihood of preferring the government to keep infl ation down rather than unemployment down) is decreasing in the income of the respondent. The odds ratio falls monotonically with the fall in the income category, and the coeffi cients are all highly signifi cant. The odds ratios suggest that a respondent in the lowest quintile is only about half as likely as someone in the richest quintile to display a preference for keeping infl a-tion down rather than unemployment. Someone in the second and third quintile is about two-thirds and three-fourths as likely as someone in the top quintile to display relative infl ation aversion, respectively.

Table 4.3 repeats the analysis of Table 4.2 but controls for several other characteristics of the respondents. The fi rst control is for the level of edu-cation. The prior expectation for this variable is unclear: a lower level of education makes an individual’s welfare potentially more vulnerable to both unemployment (lower human capital reduces employment oppor-tunities) and infl ation (less education means less knowledge and ability to protect one’s income from infl ation). The coeffi cients from column 2 suggest, however, that less educated individuals are less relatively infl a-tion averse than more educated individuals. While statistically signifi cant, these eff ects are not large: both primary and secondary educated individ-uals are about 90 percent as likely as university educated people to prefer that the government keep down infl ation rather than unemployment if it had to choose one of the two.

A second control used is the age of the respondent. We may expect that the elderly are more concerned about infl ation versus unemployment than the young, since the elderly are likely to be living off accumulated assets and pensions (which can easily be eroded through infl ation) rather than wage income. The regression suggests that it is certainly the case that respondents in all age groups are more relatively infl ation averse than those in the omitted group (those below age 20). However, those in their 60’s and 70’s do not appear to be more relatively infl ation averse than those in their 20’s and 30’s.

The last four rows report some interesting results. Not surprisingly, the unemployed are less likely than those who have employment to prefer


that the government pursue a policy of keeping infl ation down rather than unemployment. Women, too, are less relatively infl ation averse than men and are about seven-tenths as likely as men to be so. Trade union members display statistically signifi cantly less relative infl ation aversion than non-union members, perhaps because union members can more easily bargain their wages upwards than non-union members in the face of rising prices. Finally, political affi liation has an extremely strong predictive eff ect on the question at hand. Those who voted for the right wing party in the country in the last election are about 1.4 times as likely as those who did not to display relative infl ation aversion.

As the coeffi cients on the income quintile dummies show, the result that the poor are less likely than the rich to prefer that the government keep infl ation down rather than unemployment remains true in the presence of controls. A respondent in the poorest quintile is about two-thirds as likely as one in the richest quintile to say so. This central result remains robust to the inclusion of several additional controls, such as occupation, household size and whether the individual works for the public or private sector.

Moving now to considering the results on class or job attachment, columns I-III in Table 4.4 are the results from performing the logistic regression on the three defi nitions of class without any controls. As is evident from column I, fi rm owners are signifi cantly more likely than workers to be relatively infl ation averse (or less likely to be relatively

Table 4.4 Logistic regression of relative infl ation aversion on class variables (without controls)

Variable I II III

Odds ratio

Z-statistic Odds ratio


Z-statistic

Firm owner 1.47*** (5.53)Unskilled 0.69*** (−6.66)Semi-skilled 0.70*** (−6.44)Subjective lower class 0.69*** (−6.37)Subjective middle class 0.84*** (−3.15)Number of observations 23 824 13 955 20 437

Note: *** 5 signifi cant at the 1 percent level, ** 5 signifi cant at the 5 percent level, * 5 signifi cant at the 10 percent level. Omitted variable is experts (high skilled) in column II, and subjective upper class in column III. Country average infl ation dummies are included but not shown.


unemployment averse). Column II uses the classifi cation for wage laborers developed by Wright. The omitted dummy is expert workers and hence the results suggest that as compared to experts, both low-skilled and skilled workers display less relative infl ation aversion suggests that in comparison with the omitted group. Similarly, in column III in comparison to those who consider themselves upper class, those who consider themselves middle and lower class are much less relatively infl ation averse.

Columns I-III in Table 4.5 show that these results persist in the pres-ence of a variety of plausible controls, including dummies for income, gender, age, employment status and union membership. Column I shows

Table 4.5 Logistic regression of relative infl ation aversion on class variables (with controls)

Variable I II III

Odds ratio



Z-statistic

Firm owner 1.20** (2.11)Unskilled 0.87** (−1.98)Semi-skilled 0.82*** (−2.86)Subjective lower class 0.75*** (−3.72)Subjective middle class 0.86** (−2.08)Lowest income quintile 0.63*** (−7.95) 0.66*** (−5.51) 0.68*** (−6.28)Second income quintile 0.73*** (−5.17) 0.75*** (−3.65) 0.76*** (−4.24)Third income quintile 0.80*** (−3.97) 0.81*** (−3.18) 0.85*** (−2.67)Fourth income quintile 0.86** (−2.50) 0.92 (−1.17) 0.89** (−1.86)Unemployed 0.70*** (−3.80) 0.74** (−1.86) 0.71*** (−3.58)Female 0.82*** (−5.43) 0.80*** (−4.75) 0.80*** (−5.76)Union member 0.79*** (−5.10) 0.81*** (−3.90) 0.80*** (−4.76)Age 1.00 (0.97) 1.00 (0.53) 1.00 (1.24)Number of observations 14 245 9 273 13 469

Note: *** 5 signifi cant at the 1 percent level, ** 5 signifi cant at the 5 percent level, * 5 signifi cant at the 10 percent level. Omitted variable are experts (high skilled) and highest income quintile in column II, and subjective upper class and highest income quintile in column III. Country average infl ation dummies are included but not shown.


that a fi rm owner is about a fourth more likely to prefer anti-infl ation to anti-unemployment policies as a worker. Column II shows that relative to ‘experts’, semi-skilled are about eight-tenths and unskilled workers are about nine-tenths as likely to report supporting anti-infl ation to anti-unemployment policies. The results in column III show that as compared to the subjective upper class, subjective lower classes display signifi cantly less infl ation aversion. A respondent who considers themself as being part of the lower or lower middle class is about three-fourths as likely as someone who is in the upper class to prefer that the government keep infl ation down. A respondent in the middle or upper middle class is about eight-tenths as likely as someone in the upper middle or upper class to prefer that infl ation be kept down rather than unemployment.

What kinds of other policy preferences are associated with relative infl a-tion aversion? If the class content narrative is correct and not simply an artifact of the occupational categorization used in this chapter to defi ne class, we should fi nd that there are other attitudes in common with respect to class politics. To test this, I use questions from the ISSP dataset which speak to the respondent’s stance towards pro-business and pro-labor poli-cies and attitudes, on the one hand, and towards redistributionary policies on the other. The dataset provides a host of questions which could be used for the purpose and while those used here are not exhaustive, using other indicators provides very similar results.

In order to gauge attitudes towards labor and towards business I con-struct several dummy variables which could be said to represent anti-labor and pro-business sentiment. I defi ne a dummy variable pro wage control when the respondent replies being strongly in favor or in favor of wage control by law. Another dummy variable anti jobs for all refers to the situ-ation when the respondent believes that it should not be the government’s responsibility to provide a job for everyone who wants one. If the respon-dent is against the idea of the government fi nancing of new jobs then they are said to be anti government jobs. The variable pro deregulation refers to the situation the respondent is for government deregulation of business. Anti protect jobs takes a value of 1 if the respondent is against the protec-tion of declining industries in order to protect jobs. Finally the dummy variables power of labor and power of business refer to situations where the respondent believes the political power of labor and the political power of business to be too strong respectively. Column I in Table 4.6 shows the results of a logistic regression of the infl ation aversion variable on these dummies. In concordance with the narrative of class-based diff erences, all the odds ratios are statistically signifi cant and in the expected direction. There is a strong correspondence between pro-business and anti-labor pref-erences, on the one hand, and relative infl ation aversion, on the other.


Attitudes towards redistribution are approximated by three dummies. A respondent is anti redistribution if they respond that the government should not redistribute wealth, and is anti equalization if they respond that it is not the government’s responsibility to reduce income diff erences between the rich and the poor. If the respondent prefers reducing taxes even if this means lower social services then the dummy variable lower social services takes a value of 1. Column II in Table 4.6 shows the results of a logistic regression of the infl ation aversion variable on these dummies. The odds ratios are all statistically signifi cant and show that relative infl a-tion aversion is stronger among respondents who are also against gov-ernment action to redistribute income and wealth and to provide social services at the potential cost of higher taxation.

4.6 CONCLUSION

In the last three decades the hopeful message of Keynesian demand management has fallen out of favor with policy makers; as more faith has been placed in market-based solutions, in independent central banks and in microeconomic interventions to handle the problems of infl ation and unemployment. The theoretical impetus for this shift was provided by the rational expectations revolution of the 1970s and subsequent new

Table 4.6 Logistic regression of relative infl ation aversion on other attitudes

Variable I II

Odds ratio Z-statistic Odds ratio Z-statistic

Pro wage control 1.10** (2.55)Anti jobs for all 1.51*** (10.03)Anti government jobs 1.49*** (5.39)Pro deregulation 1.37*** (2.70)Anti protect jobs 1.56*** (9.96)Power of labor 1.12*** (2.97)Power of business 0.86*** (−3.22)Anti redistribution 1.32*** (6.28)Anti equalization 1.45*** (8.98)Lower social services 1.40*** (10.26)Number of observations 14 533 17 743

Note: *** 5 signifi cant at the 1 percent level, ** 5 signifi cant at the 5 percent level, * 5 signifi cant at the 10 percent level. Country average infl ation dummies are included but not shown.


classical analyses of the state. These in turn argued that democratic gov-ernments were often bad for macroeconomic effi ciency as they would tend to increase defi cits and be unable to credibly tackle infl ationary pressures. Worse still, any attempt to artifi cially reduce the unemployment rate would lead inevitably to higher infl ation with no eff ect beyond the very short term on the unemployment rate. Thus, the standard prescription of earlier times was replaced by the orthodoxy of central bank independence and increasingly a narrow focus on infl ation targeting (Bernanke et al., 1999). For proponents of this view, delegating responsibility to an author-ity which can credibly commit to a single target is benefi cial in increasing macroeconomic effi ciency and protecting the public from a perhaps well meaning, but misguided government.

A movement away from commitment to full employment and towards low targeted infl ation has potentially profound distributional conse-quences. Despite the claims made by some that anti-infl ationary policy is, for example, pro poor, it is an empirical question as to whether infl ation or unemployment is seen as a bigger problem by diff erent individuals if there is a trade-off between these. Both infl ation and unemployment are more likely, for example, to aff ect the poor, but this study as well as recent research on infl ation versus unemployment aversion (Jayadev, 2006; Scheve, 2004) found that the rich are more relatively infl ation averse than the poor. Other results presented in this chapter further strengthen this idea by fi nding that relative infl ation aversion is more pronounced among the privileged or elite broadly defi ned in class terms. Since class is a critical variable in determining an individual’s labor market opportunities as well as the source and variability of their income, the results make sense. The fi ndings are in concordance with confl ict-based models of unemployment and infl ation which argue that macroeconomic policies may have system-atically diff erential eff ects on the welfare of workers and owners (as well as on diff erent segments of the working class) and that the preferences of individuals in separate class positions refl ect these diff erences. These fi nd-ings have important consequences for research on the implications of anti-infl ation versus anti-unemployment policies in general and on the more current debate around infl ation targeting in particular.

NOTES

1. The author is grateful to Gerald Epstein, Christian Weller, Sam Bowles, Kade Finnoff , Ozgur Orhangazi, James Heintz, Suresh Naidu and Erinc Yeldan for useful comments.

2. Indeed, as Pollin (1998) points out, class confl ict is the implicit mechanism that drives the natural rate of unemployment even in orthodox neoclassical accounts. As he puts


it ‘Marx and Kalecki also share a common conclusion with natural rate proponents, in that they would all agree that positive unemployment rates are the outgrowth of class struggle over distribution of income and political power. . . . Of course, Friedman and the New Classicals reach this conclusion via analytic and political perspectives that are diametrically opposite to those of Marx and Kalecki. To put it in a nutshell, mass unemployment results in the Friedmanite/New Classical view when workers demand more than they deserve, while for Marx and Kalecki, capitalists use the weapon of unemployment to prevent workers from getting their just due’ (Pollin, 1998).

3. The idea of unemployment as a labor disciplining device has, of course, a provenance from Marx. Marxist theories of the labor market maintain this as a key fact of the labor market (for example, Bowles, 1985).

4. For a recent review see Ball and Mankiw (2002). 5. At the outset, it is important to note a potentially serious problem with the data. The

choice for infl ation as worded in the survey is for ‘infl ation and high prices’. As Easterly and Fischer themselves note, this raises the problem that in complaining about high prices, the poor are only naturally complaining about low incomes. They suggest that the low correlation between this variable and another option available to the respon-dent ‘money not enough to live by’ is proof that the poor are indeed protesting infl ation and not low income. This is not, however, truly satisfactory. It is not entirely clear that a poor respondent would list both ‘infl ation and high prices’ and ‘money not enough to live by’ as being within the top three problems facing the economy if they interpret the fi rst option as being about their low income.

6. Indeed it is often spoken about as being a cruel ‘tax’, whereby the public’s real earnings are eroded and transferred to the state.

7. Recent empirical studies, especially about the US experience of the 1990s boom, support these ideas indirectly. Bernstein and Baker (2003) fi nd that the low unemployment period in the US economy of the late 1990s aided in improving the welfare of workers according to several metrics. Real wages increased after a generation of decline, and the infl ation adjusted income of low income families grew by twice the amount that they did in the 1980s expansion (when average unemployment was higher). Abraham and Haltiwanger (1995) who review mainly US evidence suggest that real wages are more likely to be pro-cyclical than counter-cyclical.

8. Scheve (2003, 2004) remains an exception in providing more detailed evidence for the characteristics of individuals supporting each policy.

9. In the survey respondents from Israel and Germany are divided in two separate cat-egories each. The former is divided between Israeli Arabs and Israeli Jews and the latter is split between East and West Germans.

10. Survey was conducted during 2006.11. This is an example of what Wright terms a contradictory class location.12. The major exception is Israel – perhaps because of the impact that the period of hyper-

infl ation in the 1970s had upon even skilled workers.

REFERENCES

Abraham K.G. and J.C. Haltiwanger (1995), ‘Real wages and the business cycle’, Journal of Economic Literature, 33, 1215–64.

Bach, G. and J. Stephenson (1974), ‘Infl ation and the redistribution of wealth’, Review of Economics and Statistics, 61, 1–13.

Ball, L. and G.N. Mankiw (2002), ‘The NAIRU in theory and practice’, Johns Hopkins University working papers no. 475.

Barro, R. and D. Gordon (1983), ‘A positive theory of monetary policy in a natural rate model’, Journal of Political Economy, 31, 589–610.


Bernstein, J. and D. Baker (2003), The Benefi ts of Full Employment: When Markets Work for People, Washington, DC: Economic Policy Institute.

Bernanke, B., T. Laubach, F. Mishkin and A.S. Posen (1999), Infl ation Targeting: Lessons from the International Experience, Princeton, NJ: Princeton University Press.

Boddy, R. and J. Crotty (1975), ‘Class confl ict and macro-policy: the political busi-ness cycle’, Review of Radical Political Economics, 7, 1–19.

Bowles, S. (1985), ‘The production process in a competitive economy: Walrasian, non-Hobbesian, and Marxian models’, American Economic Review, 75 (1), 16–36.

Carchedi, G. (1977), The Economic Identifi cation of Social Class, London: Routledge and Kegan Paul.

Cardoso, Eliana (1992), ‘Infl ation and Poverty’, National Bureau for Economic Research working paper no. 6793, March, accessed at http://ssrn.com/abstract5293237.

Cukierman, A. (1992), Central Bank Strategy, Credibility, and Independence, Cambridge, MA: MIT Press.

Di Tella, R., R. MacCulloch and A. Oswald (2001), ‘Preferences over infl ation and unemployment: evidence from surveys of happiness’, American Economic Review, 91, 335–41.

Easterly, W. and S. Fischer (2001), ‘Infl ation and the poor’, Journal of Money, Credit, and Banking, 33 (2), 160–78.

Erikson, R. and J.H. Goldthorpe (1993), The Constant Flux, Oxford: Oxford University Press.

Esping Andersen, G. (ed.) (1992), Changing Classes: Stratifi cation and Mobility in Post-Industrial Societies, London: Sage.

Hibbs, D. (1977), ‘Political parties and macroeconomic policy’, American Political Science Review, 71, 1467–87.

Iversen, T. and D. Soskice (2001), ‘An asset theory of social preferences’, American Political Science Review, 95 (4) (December), accessed at www.people.fas.harvard.edu/~iversen/data/ISCO_conversion_tables.htm.

Iversen, T. and D. Soskice (2006), ‘New macroeconomics and political science’, Annual Review of Political Science, 9, 425–53.

Jayadev, A. (2006), ‘Diff ering preferences between anti-infl ation and anti-unemployment policy among the rich and the poor’, Economics Letters, 91, 67–71.

Jayadev, A. (2008), ‘The class content of preferences towards anti-infl ation and anti-unemployment policies’, International Review of Applied Economics, 22 (2), 161–72.

Johnson, H. (1968), ‘Problems of effi ciency in monetary management’, Journal of Political Economy, 76, 986.

Leiulfsrud, H., Bison, I. and H. Jensberg (2005), ‘Social class in Europe’, European Social Survey 2002/3, NTNU Social Research Ltd., unpublished, accessed http://ess.nsd.uib.no/fi les/2003/ESS1SocialClassReport.pdf.

Pollin, R. (1998), ‘The natural rate of unemployment: it’s all about class confl ict’, Dollars and Cents Magazine, (September/October).

Romer, Christina D. and David H. Romer (1998), ‘Monetary policy and the well-being of the poor’, National Bureau for Economic Research working paper no. 6793.

Rosenberg, S. and T. Weisskopf (1981), ‘A confl ict theory approach to infl ation in


the postwar US economy’, American Economic Review Papers and Proceedings, 71 (2), 42–7.

Rowthorn, R.E. (1977), ‘Confl ict, infl ation and money’, Cambridge Journal of Economics, 1 (3), 215–39.

Scheve, K. (2003), ‘Public demand for low infl ation’, Bank of England working paper no. 172, London.

Scheve, K. (2004), ‘Public infl ation aversion and the political economy of macro-economic policymaking’, International Organization, 58 (1), 1–34.

Shiller, R. (1997), ‘Why do people dislike infl ation?’ in C. Romer and D. Romer (eds), Reducing Infl ation: Motivation and Strategy, Chicago, IL: University of Chicago Press, pp. 13–71.

Wright, E.O. (1985), Classes, London: New Left Books.Wright, E.O. (1997), Class Counts: Comparative Studies in Class Analyses,

Cambridge: Cambridge University Press.


APPENDIX 4.A1 CLASS MEASURES

The ISSP dataset provides ISCO-88 classifi cations for all but four of the countries. For each of these countries we have ISCO-68 classifi cations. Iversen and Soskice (2001) provide a bridge between these coding mecha-nisms based on previous work by Ganzebloom. Using this code (available from Iversen’s webpage at the Harvard School of Government), I recode all respondents as per ISCO-88 codes. I drop the 300 or so observations for which there is no bridge available.

Wright (1997) provides a mechanism by which to classify ISCO-88 classifi cations into three skill categories, experts, skilled and low skilled workers. Using the codes provided by Leilsfrud et al. (2005), I replicate these categorizations for the ISSP dataset.

The code is available upon request.

Table 4.X Summary statistics

Variable Source Obs Mean Std dev.

Min. Max.

Infl ation down V63: ISSP 23 824 0.41 0.49 0 1Expert Constructed from

V202: ISSP35 313 0.07 0.25 0 1

Semi-skilled Constructed from V202: ISSP

35 313 0.23 0.42 0 1

Unskilled Constructed from V202: ISSP

35 313 0.24 0.43 0 1

Subjective lower class V221: ISSP 35 313 0.31 0.46 0 1Subjective middle class V221: ISSP 35 313 0.49 0.50 0 1Subjective upper class V221: ISSP 35 313 0.07 0.26 0 1Firm owner Constructed from

V213 and V214: ISSP dataset

35 313 0.04 0.19 0 1

First income quintile

Constructed from V217: ISSP

35 313 0.15 0.36 0 1

Second income quintile


35 313 0.11 0.31 0 1

Third income quintile


35 313 0.13 0.34 0 1

Fourth income quintile


35 313 0.11 0.31 0 1


Table 4.X (continued)

Variable Source Obs Mean Std dev.

Min. Max.

Fifth income quintile


35 313 0.10 0.30 0 1

Union member V222: ISSP 35 313 0.18 0.38 0 1Female V200: ISSP 35 228 0.52 0.50 0 1Unemployed V206: ISSP 35 313 0.06 0.23 0 1Age V201: ISSP 35 109 44.82 16.65 15 97Pro wage control Constructed from

V17: ISSP33 798 0.33 0.47 0 1

Anti jobs for all Constructed from V36: ISSP

33 820 0.27 0.44 0 1

Anti government jobs


34 188 0.06 0.24 0 1

Pro deregulation Constructed from V21: ISSP

28 350 0.92 0.28 0 1

Anti protect jobs Constructed from V23: ISSP

34 035 0.21 0.40 0 1

Power of labor Constructed from V33: ISSP

30 614 0.65 0.48 0 1

Power of business Constructed from V34: ISSP

29 484 0.84 0.37 0 1

Anti redistribution


33 542 0.24 0.43 0 1

Anti equalization Constructed from V42: ISSP

33 096 0.30 0.46 0 1

Lower social services


26 652 0.52 0.50 0 1

93

5. The gendered political economy of infl ation targeting: assessing its impacts on employment1

Elissa Braunstein and James Heintz

5.1 INTRODUCTION

Central banks in developing countries are increasingly maintaining low infl ation rates as the central goal of monetary policy, without much con-sideration of how these policies impact real outcomes like employment, investment and growth (Epstein, 2003). Although targeting very low infl ation rates seems to have done little to raise economic growth, these policies remain a key feature of neoliberal approaches to monetary policy (Epstein, 2000). Gerald Epstein and Juliet Schor argue that anti-infl ation policy and neoliberal approaches to central banking emerged from the ‘contested terrain’ of central banks – the class and intra-class confl icts over the distribution of income and power in the macroeconomy (Epstein, 2000; Epstein and Schor, 1990). Their work underscores the importance of understanding monetary policy from a political economy perspective, as the distribution of the gains and costs of economic policy proff ers insight into both a policy’s genesis and its longer term consequences.

In this chapter we build on their analysis by considering the employ-ment costs of infl ation reduction in developing countries from a gender perspective. We explore two empirical questions: (1) what is the impact, if any, of infl ation reduction on employment, and is the impact diff erent for women and men, and (2) how are monetary policy indicators con-nected to defl ationary episodes and gender-specifi c employment eff ects? We fi nd a common pattern among countries undergoing what we term contractionary infl ation reduction, or periods of declining infl ation that are accom panied by a loss of employment. After controlling for long-term employment trends, we fi nd that the ratio of women’s to men’s employ-ment tends to decline during these periods in the majority of countries examined. During periods of expansionary infl ation reduction, however, there are no clear patterns to the relative changes in women’s and men’s


employment. In terms of monetary policy, we fi nd that countries that respond to infl ation by raising real interest rates or tightening the real money supply (both relative to their long-run trends) are also more likely to experience employment contractions, with concomitantly higher costs for women’s employment. Conversely, those that maintain competitive real exchange rates are likely to reverse the negative impact of contraction-ary infl ation reduction on women’s relative employment.

That the costs of infl ation reduction, at least in terms of employment, are inequitably distributed means that the contested terrain of monetary policy is also gender specifi c, with the result that the costs of implementing these sorts of policies are actually quite diff erent – and potentially higher – than is generally presumed. After a discussion of the literature on the diff erences in women’s and men’s unemployment in developing countries, and presenting our empirical results, we will develop this last point in more detail in the closing section.

5.2 GENDER DIFFERENCES IN EMPLOYMENT AND UNEMPLOYMENT IN SEMI-INDUSTRIALIZED COUNTRIES

In thinking about the diff erences between women’s and men’s unemploy-ment, it is helpful to think in terms of supply-side factors and demand-side factors (Seguino, 2003). On the supply side, diff erences in human capital are probably the most commonly considered. However, gender-based diff erences in education, skill and experience are themselves rooted in workers’ productive roles outside the factory door and the institutional, social and material contexts in which they live. A useful way of consider-ing these diff erences is through what Folbre (1994) terms ‘structures of constraint’ – the preferences, norms, assets and rules that shape individual choice.

Women and men make decisions about whether or not to look for wage work. However, self-perception, what individuals value, and what choices they perceive as possible are constituted by the social world, and are dif-ferent for women and men (Sen, 1990), and so individual preferences must be understood in this light. Norms are the traditional structures of gender and kinship that constitute the social expectations of women and men in the household. For example, women are primarily associated with the care of the family, and much of their work time is spent outside of the market, whereas men’s work is typically viewed as directly productive and more fully incorporated into the market sphere. Assets also engender labor supply. Systematic diff erences by gender in ownership of assets are

The gendered political economy of infl ation targeting 95

common, and partly determine how much wage employment women seek. Similarly, laws that confer patriarchal property rights, where eldest men have the right to claim and apportion the fruits of household members’ labor time, can create incentives for high fertility and lower female labor force participation (Braunstein and Folbre, 2001). Conversely, not having a legal claim on a spouse’s income in the event of separation means that a paying job can be an insurance policy against loss of that support (Folbre, 1997).

On the demand side, gender discrimination results in diff erential access to employment opportunities. The presumption that men should bear the primary fi nancial responsibility for provisioning families has been linked with higher unemployment for women relative to men in OECD countries (Algan and Cahuc, 2004). Women are laid off fi rst because employers presume that it is more important for men to be able to fulfi ll their tra-ditional breadwinning responsibilities (Azmat et al., 2004). Diff erential hiring practices also contribute to gender segregation in employment. For example, labor-intensive exporters prefer to hire women both because women’s wages are typically lower than men’s, and because employ-ers consider women to be more productive in these types of jobs due to employer perceptions that women possess ‘nimble fi ngers’, are less prone to worker unrest, are better suited to tedious work and are more reliable workers than men (Anker and Hein, 1985; Elson, 1996; Elson and Pearson, 1981; Fernández-Kelly, 1983).2 By extension, women may lose their com-parative advantage when industries upgrade, leading to a defeminization of employment as has happened in Mexico, India, Ireland and Singapore (Elson, 1996; Fussell, 2000; Ghosh, 2001; Joekes, 1999).

As a result of these gender diff erences in labor demand and supply, changes in macroeconomic structure and policy have diff erential eff ects on men’s and women’s work (Seguino, 2003). For example, feminists have long argued that the interaction between gender relations and structural adjustment programs (SAPs) have implications both for the distribution of costs and benefi ts between diff erent groups of women and men, and for the achievement of the economic objectives of the SAPs themselves (Bakker, 1994; Benería and Feldman, 1992; Benería and Roldan, 1987; Cagatay et al., 1995; Elson, 1991, 1995).

Macroeconomic policies and variables have been shown to infl uence women’s labor supply decisions. Cagatay and Ozler (1995) pool cross country data for 1985 and 1990 to show that SAPs led to increased femin-ization of the labor force via worsening income distribution and openness. Others have also argued that, in the case of Latin America, macroeco-nomic crises place pressures on household resources which increase labor force participation and hours worked, particularly for women (Arriagada,


1994; Cerrutti, 2000). Similarly, research into the determinants of women’s labor supply in post-apartheid South Africa has shown that women’s labor force participation responds positively to growing unemployment, thereby further increasing the country’s average unemployment rate (Casale, 2003).

Trade liberalization and structural adjustment policies can cause a reallocation of paid employment from traditional activities, which are adversely aff ected, to export-oriented production, which is encouraged (Cagatay and Ozler, 1995; Standing, 1999). In addition, female-intensive export-oriented industries are often more cyclically volatile than men’s industries, resulting in higher overall rates of unemployment (Howes and Singh, 1995). Emphasis on export-oriented industrialization has also been associated with increases in informalization of certain types of employ-ment due to increases in competitive pressures (Carr et al., 2000; Standing, 1999). As female labor force participation rises in the context of crisis and structural adjustment, the increasing dominance of informal work has become a key new reality for women (Arriagada, 1994; Benería, 2001; Patnaik, 2003).

Similar work has been done on the gendered employment eff ects of the Asian fi nancial crisis in 1997–98 (Aslanbeigui and Summerfi eld, 2000; Lim, 2000; United Nations, 1999). Women were typically the fi rst to be laid off both because they worked in more cyclically volatile fi rms, such as small export-oriented enterprises, and because of eff orts to protect the jobs of ‘male breadwinners’. The lack of formal employment opportuni-ties and pressures for greater labor force participation led to an increase in women’s informal employment in some cases. A slightly diff erent pattern was found by Lim (2000) in the Philippines, where the post-crisis decline increased male unemployment more than female unemployment despite a rapid displacement of women from the manufacturing sector (especially in traded goods). The reason was the relative resilience of the service and trade sectors, which are more female intensive. Women did, however, increase their labor force participation to deal with male unemployment, and their total work hours relative to men increased as well.

These gender dynamics have implications for macroeconomic outcomes. For example, increases in women’s employment that are associated with lower unit labor costs have important implications for infl ation dynamics. For the non-tradable and import-competing sectors in which some degree of competition exists, lower unit labor costs associated with growth in women’s employment reduce infl ationary pressures in the countries where the women work.

Clearly, there are signifi cant structural diff erences between women’s and men’s labor markets on both the supply and demand sides that are


diff erently aff ected by macroeconomic structure and policy. The literature reviewed above on semi-industrialized countries suggests that economic contractions have a larger negative eff ect on women’s formal employment than men’s, though women tend to increase their labor force participation at the same time to protect household income. In the next section we turn to a more direct test of this employment hypothesis, looking at whether the process of infl ation reduction in particular can be associated with gender diff erentials in employment.

5.3 INFLATION REDUCTION AND WOMEN’S EMPLOYMENT

The empirical exercise we present here explores the eff ects of infl ation reduction on women’s and men’s formal employment. We compiled data for 51 ‘infl ation reduction episodes’ in 17 low- and middle-income coun-tries.3 To assess the employment eff ects of infl ation reduction periods, we looked at actual employment trends during each infl ation reduc-tion episode, disaggregated by gender, and compared these to long-run employment trends, estimated by applying a Hodrick-Prescott fi lter to the employment series. We also examine indicators that suggest how mon-etary policy responded during infl ation reduction episodes using a similar approach. We compare average short-term real interest rates, growth rates of the real money supply, and indicators of the real exchange rate to their long-run trends to see if these variables deviated from trend during infl ation reduction episodes. We drew from three diff erent data sources in conducting this exercise: employment data came from the International Labor Organization’s (ILO) LABORSTA online database, and the macro-economic data was compiled from World Development Indicators 2005 (Washington, DC: World Bank) and International Financial Statistics (Washington, DC: IMF, October 2005).4

The methodology used is drawn from the literature on measuring ‘sac-rifi ce ratios’ – that is, the loss of output or employment associated with a given reduction in infl ation. Sacrifi ce ratios can be measured simply as the slope of the Phillips Curve showing an employment-infl ation trade-off . However, this approach assumes that the sacrifi ce ratio remains constant over long periods of time. Other approaches examine changes in employ-ment or output over specifi c defl ationary periods. This technique allows the sacrifi ce ratio to change over time and makes it possible to analyse the behavior of economic variables during specifi c defl ationary periods. This is the approach followed here.

Ball (1993) suggests one method for identifying defl ationary periods.


Infl ation fi gures are notoriously volatile, making the identifi cation of turning points diffi cult. A moving average of infl ation – in our case, a three-year moving average, encompassing one previous year and one subsequent year – can be used to smooth the series. Peaks and troughs in the smoothed infl ation series are identifi ed. Peaks occur when the value in a particular year exceeds the values of immediately adjacent years. Troughs occur when the value in a given year falls below the values of the adjacent years. A defl ationary period runs from a peak year to the next trough year.

For the purposes of this exercise, we use the term ‘infl ation reduction episode’ to refer to these defl ationary periods. The reason for this is that, during some of the periods identifi ed, employment actually expands more rapidly than its long-run trend. It seems confusing to refer to these periods as ‘defl ationary’. Therefore, we use the terms ‘expansionary infl ation reduction episode’ and ‘contractionary infl ation reduction episode’.

The infl ation reduction episodes identifi ed for the 17 countries exam-ined are remarkably similar in terms of the timing of the episodes. That is, infl ation reduction episodes occur in many of these countries simulta-neously. Figure 5.1 illustrates this pattern. The fi gure shows, for each year

0

10

20

30

40

50

60

70

80

90

100

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

Per

cent

of s

ampl

e co

untr

ies

unde

rgoi

ng d

isin

flatio

n

0

2

4

6

8

10

12

14

Inte

rest

rat

e (p

erce

nt)

Interestrate

Figure 5.1 Infl ation reduction episode frequencies and the real US interest rate, 1971–2003


1970–2003, the percent of all countries in our study in which an infl ation reduction episode occurred. Infl ation reduction episodes are concentrated over certain sub-periods: 1974–77, 1980–86, 1991–94 and 1997–2000. Also Figure 5.1 clearly demonstrates that infl ation reduction episodes were much more common throughout the 1990s than throughout the prior two decades, the 1970s and 1980s.

This pattern of infl ation reduction suggests that one or more common factors determine infl ation rates and monetary variables in the diff erent countries examined. One obvious possibility is the world interest rate. Figure 5.1 shows the real yield on US Treasury bills over the entire period as a proxy for short-term, infl ation-adjusted world interest rates. An increase in the real yield of US Treasury bills preceded three out of the four common infl ation reduction episodes. Only in the most recent episode, from approximately 1997 to 2000, did interest rates fail to increase before-hand.5 Because the data for the diff erent countries show similar trends in the variables analysed, we emphasize deviations from long-run trends, rather than absolute levels or changes, in order to investigate common patterns and divergent trends.

We restricted the countries used as follows:

Only low- or middle-income countries were examined. ● 6

Countries must have at least 20 years of gender-disaggregated ●

employment data (it would be hard to estimate a meaningful ‘long-run trend’ with a shorter series).Time series with missing values were used. However, time series with ●

two consecutive missing observations were rejected. Missing values were estimated for the purposes of computing a long-run trend by extrapolating between the previous and subsequent values in the series.

Changes in employment across infl ation reduction episodes were calcu-lated as the annualized value of the overall rate of change in employment across the entire peak-to-trough period. As mentioned earlier, values for the long-run employment trends were computed by applying a Hodrick-Prescott fi lter to the actual employment time series for each country (men, women and total). The employment time series used most likely underestimates the magnitude of informal employment in these countries. Therefore, the results of this analysis should be interpreted bearing this in mind.

Table 5.1 summarizes the results for all the infl ation reduction episodes studied. The table shows the country name, the dates of each infl ation reduction episode, and the deviation from the long-run trend for women’s


Table 5.1 Infl ation reduction episodes and deviations from long-run employment trends, disaggregated by gender

A. Contractionary infl ation reduction episodes

period Defl ation as percent

of avg. infl ation

Deviations from long-run employment trends (percent)

W M Ratio

Barbados 1980–86 −143 −1.8 −0.8 −1.01990–94 −113 −2.5 −2.6 0.1****1996–99 −117 −1.0 0.3 −1.4

Brazil 1993–99 −305 −0.7 −0.1 −0.6Colombia 1980–85 −29 −3.2 −2.5 −0.7Costa Rica 1982–85 −123 −1.5 −0.1 −1.4India 1973–77 −175 −0.2 −0.4 0.2****

1982–86 −37 −0.1 −0.2 0.1****1991–94 −26 0.1 −1.2 0.2****1997–02 −93 −1.0 0.9 −0.2

Jamaica 1974–76 −45 −0.5 −0.2 −0.31992–2000 −190 −0.5 0.1 −0.6

Kenya 1975–80 −36 −2.2 −0.1 −2.11981–87 −66 0.8 −0.3 1.1****

Malaysia 1981–86 −181 −0.4 −0.8 0.4****Mauritius 1980–86 −182 −0.6 −1.6 0.9****

1989–93 −46 −1.3 –0.3 −0.91994–96 −21 −1.8 −0.9 –0.8

Philippines 1973−76 −69 −1.6 −0.4 –1.21980−82 −55 0.2 −0.3 0.5****1984–87 −130 −2.4 0.0 –2.3

Singapore 1974−76 −171 −6.7 −0.7 −5.91981–86 −267 −1.8 −2.0 0.1****

South Korea 1980–85 −196 −1.4 −0.9 −0.51991–94 −45 −0.4 0.0 −0.4

1997–2000 −82 −1.2 −1.1 −0.1Sri Lanka 1981–86 −106 −0.7 0.1 −0.8

1997–99 −56 −0.7 −2.6 1.9****Taiwan 1974–76 −106 −4.9 0.3 −5.1

1980–85 −243 0.6 −0.5 1.1****1991–02 −197 −0.4 −0.2 0.2****

Thailand 1974–76 −88 −1.3 −0.8 −0.51980–85 −180 −2.6 −0.7 −1.81990–93 −31 −0.8 0.1 −0.91997–00 −136 −0.8 −0.7 −0.1

Trinidad and Tobago 1980–87 −56 −1.0 −0.6 −0.4


employment, men’s employment, and the female to male employment ratio. Negative values indicate that the series grew more slowly than the long-run trend (a negative value could also indicate a more rapid decrease in the actual value compared to the long-run trend). Table 5.1 is divided into contractionary and expansionary infl ation reduction episodes. During contractionary infl ation reduction episodes the rate of increase of total employment fell below its long-run trend. During expansionary infl ation reduction episodes, the rate of increase of total employment was equal to or greater than its long-run trend.

In 67 percent of contractionary infl ation reduction episodes the rate of change of the female to male employment ratio fell below its long-run trend, indicating that women’s employment was disproportionately aff ected by the slowdown. If India were excluded from the sample, this proportion would rise to over 72 percent.7 However, in expansionary


Brazil 1989–92 −39 1.9 −0.8 2.8****Chile 1984–88 −39 0.8 2.3 −1.5Costa Rica 1991–93 −42 0.2 1.6 −1.4Jamaica 1979–82 −101 0.1 0.5 −0.4

1985–88 −90 2.9 0.5 2.3****Kenya 1993−96 −134 0.9 −0.3 1.3****Malaysia 1992–96 −27 0.6 1.4 −0.8Mauritius 1974–77 −59 3.9 1.6 2.1****Philippines 1990–94 −60 0.2 0.3 −0.1Singapore 1990–99 −139 0.1 0.1 0.0Sri Lanka 1974–76 −100 1.9 3.1 −1.1

1989–94 −50 8.9 3.6 4.9****Trinidad and

Tobago1974–77 −48 0.2 1.0 −0.81989–92 −37 3.8 1.6 2.1****1993–96 −69 0.7 0.6 0.2****

Note: **** Infl ation reduction episodes in which the ratio of women’s to men’s employment increased more rapidly than the long-run trend.

Summary:

Contractionary infl ation reduction episodes with declining female:male employment ratios relative to the trend: 67 percent.Contractionary infl ation reduction episodes with declining female:male employ ratios relative to the trend (excluding India): 72 percent.Expansionary infl ation reduction episodes with increasing female:male employment ratios relative to the trend: 53 percent.


infl ation reduction episodes there was no clear distinction. The female to male employment ratio increased faster than trend in 53 percent of cases and at or below trend in 47 percent of cases – nearly an even split.

The contractionary infl ation reduction episodes were associated with a larger decline in infl ation from peak to trough relative to the average infl ation rate for the entire episode. Averaging across all contractionary episodes, the decline in infl ation was 115 percent of the average infl ation rate during the episode in question. For expansionary episodes, the decline was 69 percent.

The diff erence in employment experiences across countries during infl ation reduction episodes – for example, expansion or contraction of employment – might be explained, in part, by policy choices. For example, if real interest rates rose above the long-run trend in reaction to an accel-eration of infl ation, this could trigger a contractionary infl ation reduction episode. However, if real interest rates were not raised above the long-run trend (for example, they were kept in line with the long-run trends of global interest rates), a contraction of employment might be avoided.

To examine this possibility, we looked at average real short-term inter-est rates across the infl ation reduction episodes.8 In most cases short-term rates linked directly to monetary policy choices were used (for example, a discount rate or bank rate). If these rates were unavailable, yields on short-term (three-month) Treasury bills were calculated instead.9 If actual average real interest rates were negative, these infl ation reduction episodes were tabulated separately.

Table 5.2 shows patterns in short-term interest rates over expansion-ary infl ation reduction episodes. Average actual real interest rates are compared to the average interest rates associated with the long-run trend, calculated by applying a Hodrick-Prescott fi lter to the real interest rate series. The diff erence in average real interest rates (actual rates minus the rates associated with the long-run trend) is expressed as a percentage of the average long-run trend in real interest rates over the infl ation reduction episode. Only infl ation reduction episodes with positive actual average real interest rates over the episode in question are included in Table 5.2.

In almost all of the infl ation reduction episodes, actual real interest rates were, on average, kept below the long-run trend. This is consistent with the argument that countries that do not raise interest rates in the face of growing infl ationary pressures are less likely to experience employment losses, or a slow-down in the growth rate of employment, during infl ation reduction episodes. There were only two exceptions: Sri Lanka 1974–76 and Trinidad and Tobago 1989–92.

One problem with looking at interest rates over the entire infl ation reduction episode is that policy makers might increase interest rates in


anticipation of infl ation or in response to domestic economic conditions prior to the year in which infl ation peaks. Therefore, we also examined average interest rates over an alternative periodization in order to see if the timing of the interest rate changes infl uenced our interpretation of the nature of the infl ation reduction event. We looked at the period one year before the beginning of an infl ation reduction episode and extending half-way into the episode. For example, if the infl ation reduction episode spanned the years 1980 to 1984, then the alternative period over which interest rates were compared would be 1979 to 1982.

This alternative periodization for the real interest rate analysis shows that real interest rates were negative on average in Sri Lanka from 1973 to 1975. Therefore, the increase in real interest rates documented in Table 5.2 might represent an eff ort by policy makers to move from negative real interest rates to positive (yet still relatively low) rates. Raising negative interest rates (that is, making them more positive) could have a diff erent impact on employment than raising positive real interest rates – a point to which we will return shortly. However, the alternative periodization does not shed much light on the case of Trinidad and Tobago from 1989 to 1992.

Table 5.3 shows a similar set of calculations for another sub-set of

Table 5.2 Patterns in average real short-term interest rates over infl ation reduction episodes in which total employment expanded (only episodes with positive average real interest rates included).

Period Diff erence in actual and long-run average real interest rates as

a percent of the long-run average

Notes

Chile 1984–88 −31.3Costa Rica 1991–93 −9.8Kenya 1993–96 −2.1Malaysia 1992–96 −9.7Philippines 1990–94 −42.8Singapore 1990–99 −6.9Sri Lanka 1974–76 124.0 1973–75: avg. neg.

interest1989–94 −5.6

Trinidad and Tobago

1989–92 16.3 1988–91: 15.4 per cent

1993–96 −3.0


infl ation reduction episodes. This time, contractionary infl ation reduction episodes are featured. As with Table 5.2, only infl ation reduction episodes with positive actual average real interest rates over the episode in question are included.

In most cases the opposite pattern observed previously in Table 5.2 is

Table 5.3 Patterns in average real short-term interest rates over infl ation reduction episodes in which total employment contracted (only episodes with positive average real interest rates included)

Period Diff erence in actual and long-run average real interest rates as

a percent of the long-run average

Notes

Barbados 1980–86 136.71990–94 14.41996–99 −3.3 1995–98: 11.0

percentColombia 1980–85 −1.1 1979–83: 134.6

percentIndia 1982–86 114.2

1991–94 −9.3 1990–93: −19.3 percent

1997–02 −18.4 1996–2000: −12.8 percent

Jamaica 1992–2000 −8.4 Interest rates are t-bill rates. 1991–96: avg. neg. interest

Malaysia 1981–86 151.9Mauritius 1989–93 −79.2 1988–91: −76.3

percent1994–96 116.4

Singapore 1981–86 117.9Sri Lanka 1981–86 1398

1997–99 12.9Taiwan 1980–85 124.3

1991–02 −6.5 1990–97: −11.9 percent

Thailand 1980–85 16.61990–93 15.2

1997–2000 114.2


evident. During contractionary infl ation reduction episodes actual inter-est rates were kept above the long run trend on average. In two cases, Barbados 1996–99 and Colombia 1980–85, actual real interest rates fell below the long-run trend during the entire infl ation reduction episode. However, if we apply the alternative periodization we can see that real interest rates increased before infl ation peaked and during the earlier stages of the infl ation reduction period. For example, in Colombia from 1979 to 1983 average actual real interest rates were nearly 35 percent higher than the long-run average over this same period.

There were fi ve contractionary infl ation reduction episodes in which real interest rates behaved diff erently: India 1991–94, India 1997–2002, Jamaica 1992–2000, Mauritius 1989–93 and Taiwan 1991–2002. In one case – Jamaica – the alternative periodization shows that interest rates were negative on average going into the infl ation reduction episode. This alters the impact on employment of keeping rates below their long-run trend. However, in the other four cases interest rates and employment exhibited a diff erent pattern when compared to the other episodes in Table 5.3.

This discussion suggests that negative real interest rates are important in analysing the employment trends across infl ation reduction episodes. Table 5.4 summarizes trends in real interest rates for those infl ation reduc-tion episodes in which average actual real interest rates were negative.

In the majority of the cases in which average real interest rates were negative, interest rates were kept below the long-run average and employ-ment grew slower than the trend rate of growth. This suggests that keeping interest rates negative and below the long-run trend will not help to increase employment. In two cases employment expanded despite low, negative real interest rates: Jamaica 1979–82 and Trinidad and Tobago 1974–77. However, these appear to be the exception rather than the rule. One reason why negative real interest rates tend to be associated with a contraction in employment relative to its long-run trend is that many of these infl ation reduction episodes are ‘stagfl ationary’. That is, supply-side shocks produce a situation in which infl ation accelerates, economic growth slows down, resulting in negative real interest rates.

Real interest rates are only one set of variables that potentially link monetary policy to infl ation reduction and employment dynamics. Two other possibilities are the real exchange rate and the growth rate of the real money supply. Table 5.5 examines trends in these two variables across the various infl ation reduction episodes. Specifi cally, it shows dif-ferences between the actual average annual growth rate and the growth rate associated with the long-run trend for the real exchange rate and the real money supply across various infl ation reduction episodes. The


infl ation reduction episodes are separated into contractionary and expan-sionary periods.

In terms of the real exchange rate, for the purposes of this study, we measured the rate as the nominal US dollar exchange rate adjusted for changes in the US GDP defl ator relative to the specifi c country’s GDP defl ator. A decrease in the value of our real exchange rate measurement, therefore, represents an appreciation in the real exchange rate. Likewise, an increase in value represents a depreciation.

There do not appear to be any systematic patterns with respect to changes in the real exchange rate across infl ation reduction episodes and whether the episode was contractionary or expansionary. In 34 percent of infl ation reduction episodes the average annual percent change in the real exchange rate was below that of the long-run exchange rate (that is, the exchange rate appreciated relative to its long-run trend); in 60 percent of the episodes the diff erence in average growth rate was positive (that is, the actual real exchange rate depreciated relative to the long-run trend); and in 6 percent

Table 5.4 Patterns in average real short-term interest rates over infl ation reduction episodes in which average real interest rates were negative

Period Actual interest rates above or below long-run trend

Employment: expansionary or contractionary

Costa Rica 1982–85 below contractIndia 1973–77 below contractJamaica 1974–76 below contract

1979–82 below expand1985–88 above expand

Kenya 1975–80 below contractMauritius 1980–86 below contractPhilippines 1973–76 below contract

1980–82 below contract1984–87 below contract

Singapore 1974–76 below contractSouth Korea 1980–85 below contract

1991–94 below contract1997–2000 above contract

Taiwan 1974–76 below contractTrinidad and Tobago 1974–77 below expand

1980–87 below contract


Table 5.5 Diff erences in the average actual annual growth rate and the average annual growth rate associated with the long-run trend for the real exchange rate (RER) and the real money supply across infl ation reduction episodes

Country Period RER (percent) Money supply (percent)

Total employment contracts on averageBarbados 1980–86 –0.3 −1.1

1990–94 11.1 −1.51996–99 −1.2 13.0

Brazil 1993–99 11.1 10.3Colombia 1980–85 12.2 10.2Costa Rica 1982–85 −6.7 −9.4India 1973–77 11.6 −1.6

1982–86 10.8 10.61991–94 13.6 −0.4

1997–2002 10.4 11.7Jamaica 1974–76 −9.0 –0.7

1992–2000 −0.6 10.5Kenya 1975–80 −3.5 11.9

1981–87 11.7 11.8Malaysia 1981–86 14.5 −1.4Mauritius 1980–86 0.0* −2.2

1989–93 11.8 −1.91994–96 −4.0 −1.4

Philippines 1973–76 −3.9 −1.61980–82 10.8 −0.41984–87 0.0 −10.2

Singapore 1974–76 13.9 −0.61981–86 12.8 −2.4

South Korea 1980–85 17.0 −1.91991–94 −0.4 −2.7

1997–2000 16.7 16.2Sri Lanka 1981–86 −0.7 −1.1

1997–99 11.7 0.0Thailand 1974–76 −0.2 −0.3

1980–85 13.7 −2.51990–93 −1.9 12.8

1997–2000 17.0 −0.4Trinidad and Tobago 1980–87 12.7 −1.0

Total employment expands on averageBrazil 1989–92 0.0 14.9Chile 1984–88 12.2 12.7


of the episodes there was no diff erence between the growth rate of the actual and long-run real exchange rates. These ratios were approximately the same for contractionary and expansionary infl ation reduction episodes.10

However, real exchange rates appear to have an impact on the gender bias observed in contractionary infl ation reduction episodes. Recall that, in the majority of cases, women’s formal employment was disproportion-ately aff ected by the slow-down in employment growth. However, about a third of the time the ratio of women’s to men’s employment actually improved when compared to its long-run trajectory. In each of these cases the real exchange rate either depreciated or showed no deviation relative to its long-run trend. In other words, maintaining a competitive exchange rate may off set some of the gender bias observed during contractionary infl ation reduction. Why would this be the case? As previously noted, in many countries the growth of women’s employment – particularly formal employment and wage employment – has tended to be concentrated in tradable sectors, either export-oriented or import-competing (Benería, 2003; Elson, 1996; Elson and Pearson, 1981; Kabeer and Mahmud, 2004). A real depreciation of the exchange rate favors tradable sectors and could help protect women’s employment in certain cases.

Turning to the real money supply,11 there is some indication that money


Country Period RER (percent) Money supply (percent)

Costa Rica 1991–93 −4.8 −1.1Jamaica 1979–82 −2.2 11.9

1985–88 −7.8 14.9Kenya 1993–96 15.7 111.4Malaysia 1992–96 −4.6 15.1Mauritius 1974–77 n.a. 15.4Philippines 1990–94 −1.8 −0.5Singapore 1990–99 10.4 11.4Sri Lanka 1974–76 −1.0 −4.8

1989–94 −0.3 −0.7Trinidad and Tobago 1974–77 −1.0 −1.0

1989–92 −3.8 −2.71993–96 11.4 15.4

Note: * Based on 1981–86 average, due to data limitations.

Source: Authors’ calculations based on data from the IMF publication International Financial Statistics. Real US interest rate is T-bill rate less the infl ation rate.


supplies grew more slowly during contractionary episodes. In 67 percent of all contractionary episodes for which data is available actual average annual growth rates in the real money supply fell below the long-run trend. In 60 percent of all expansionary episodes actual annual growth rates in the real money supply were greater than the average for the long-run trend. Relative to the gendered eff ects of infl ation reduction then, tightening the real money supply may have a tendency to be associated with greater sacrifi ces in women’s employment.

This analysis suggests a number of preliminary fi ndings. However, it is important to recognize that a myriad of factors aff ect the variables exam-ined here – in particular, employment, infl ation, interest rates, exchange rates and the money supply – and therefore any generalizations must be tentative. Although some commonalities do arise, the diversity of country experiences prevents us from drawing defi nitive conclusions without ad-ditional in-depth analysis and country-specifi c case studies.

Nevertheless, we can make some general observations from the analysis presented.

Infl ation reduction episodes occur simultaneously across a large ●

number of countries. This suggests that there are common, external factors that infl uence infl ation dynamics in low- and middle-income countries.If employment contracts during an infl ation reduction episode, it ●

is likely that women will experience a larger loss of employment, in percentage terms, than men. However, during infl ation reduction episodes in which employment expands, the gender-specifi c impact is ambiguous.Countries that respond to infl ationary pressures by raising real ●

interest rates above the long-run trend are more likely to experience a slow-down in the growth of employment relative to those coun-tries that keep interest rates in line with or below the long-run trend, with concomitantly higher losses for relative female employment. However, countries with negative real interest rates do not appear to be able to increase employment growth by lowering real interest rates still further.We did not fi nd a link between changes in the real exchange rate ●

and the impact of infl ation reduction on employment in general. However, we did fi nd that real exchange rates seem to impact the gender bias of contractionary infl ation reduction episodes. In all cases where women experienced relative employment gains during employment contractions, exchange rates either depreciated or showed no deviation relative to long-run trends.


Tightening the real money supply also seems to be negatively asso- ●

ciated with employment in general and women’s employment in particular.

These results suggest that contractionary monetary policy aimed at reducing infl ation often has a disproportionately negative impact on women’s employment, an eff ect that may be eased by maintaining a com-petitive exchange rate. Conversely, non-contractionary infl ation reduc-tion is not necessarily favorable to women’s formal employment in all circumstances.

5.4 DISCUSSION AND CONCLUSIONS

The empirical analysis presented here concerns the short-run, gender- specifi c impacts of policy responses during infl ation reduction episodes. The results say little about the long-run impact of diff erent policy responses. Supporters of infl ation targeting frequently acknowledge that short-run trade-off s might exist, but the long-run benefi ts of low infl a-tion for growth and development are more signifi cant. This argument is problematic when transitory policy shocks have long-run consequences for real economic variables (Fontana and Palacio-Vera, 2004). Similarly, short-term gender-specifi c shocks can have long-run eff ects for a country’s human and economic development.

A number of empirical studies suggest that gender-based inequities in employment and unemployment have implications for long-term develop-ment. For example, this body of research shows that a positive relationship exists between gender equality (measured most commonly as educational equity) and economic growth in developing countries (Dollar and Gatti, 1999; Hill and King, 1995; Klasen, 1999). Some of the eff ects are quite large: Klasen (1999), in a panel data study between 1960 and 1992, fi nds that had South Asia and sub-Saharan Africa had more gender equity in education, growth would have been 0.9 percent per year faster. Investing in girls makes for a higher productivity workforce, but higher rates of unemployment and cyclical volatility in women’s jobs will discourage these types of investments at both the individual and community levels.

In a related sense, lower incomes and higher income volatility for women could lead to lower investments in human capital overall, thereby lowering long-term growth. Theory and evidence have aptly demonstrated a higher co-incidence between a mother’s income and the family’s basic needs than a father’s income (Benería and Roldan, 1987; Blumberg, 1991; Chant, 1991), a fi nding underlying what has been termed the ‘good mother


hypothesis’. Income that is controlled by women is more likely to be spent on children’s health and nutrition (Dwyer and Bruce, 1988; Hoddinott et al., 1998). In many countries a large proportion of fathers provide little or no economic support for their children (Folbre, 1994). But faced with cyclically higher rates of unemployment during disinfl ation, ‘good mothers’ will have fewer opportunities to invest in their children.

In the nearer term systematic discrimination in hiring and fi ring results in a misallocation of workers as the most able are not matched with the most appropriate jobs. Such misallocations can result in effi ciency losses with eff ects on output and growth. Tzannatos (1999) evaluates the eff ects of eliminating gender-based occupational segregation in a number of developing countries, and fi nds on average a one-time gain in output of 6 percent. His results are signifi cant but incomplete. He only considers occupational segregation, not industrial segregation, which can be a much more extensive part of gender segregation in some countries, as in China. In addition, Tzannatos only considers the static picture, and not the dynamic effi ciency losses from being on a particular growth path charac-terized by a high degree of gender segregation (Elson, 1999).

Less instrumental and ultimately more central than the economic impli-cations of gender-biased infl ation reduction discussed above is the issue of equity, as we fi nd that women as a group shoulder a disproportionate share of the costs of contractionary infl ation reduction. From a social justice perspective, then, it is important to better understand and redress these gender-biased outcomes. One way of doing so would be to focus research and policy on better understanding the link between infl ation targeting and overall social welfare. To get at these connections, central banks should incorporate gender-specifi c indicators in the creation of targets, such as gender disaggregated employment fi gures or gender aware infl ation rates (that account for gendered consumption and employment patterns). These market-based targets should also be supplemented by longer term gender aware human development targets to ultimately get at the links among infl ation targeting, equity and well-being.

Treating gender equity as secondary to other, more ‘general’ economic concerns may also be instrumental in the political sustainability of infl a-tion targeting in monetary policy. As our analysis implies, incorporating concerns over gender equity into monetary policy formulation would involve a move away from infl ation targeting as it is currently practiced and could harm the interests of those invested in a low infl ation, high interest rate environment. Moreover, if women’s labor force participa-tion keeps unit labor costs and infl ation lower than it would otherwise be, then a focus on gender equity within the context of sustainable levels of infl ation could require other mechanisms for price control that are more


consistent with long-run development. Such a move might be resisted by those that benefi t (perhaps only in the short run) from women’s more pre-carious employment – for example, their employers and employed men. From this perspective, gender-biased central bank policy may help solve the political problems introduced by infl ation targeting in that gender bias concentrates the costs of these policies on a less powerful segment of society – women. As such, infl ation targeting policies should be considered in terms of their social content (for example, what are the social structures that underlie this policy) as well as its social impact (Elson and Cagatay, 2000). Such an approach underscores the gendered nature of the contested terrain in macroeconomic policy making.

NOTES

1. This chapter is part of the PERI/Bilkent project on ‘Alternatives to Infl ation Targeting’. We thank the Ford Foundation, the UN-DESA, the Rockefeller Brothers Fund and the Political Economy Research Institute at the University of Massachusetts for fi nancial support. We would also like to thank Gerald Epstein, Erinc Yeldan and Diane Elson for their help and comments, as well as the participants (and our hosts) of the AIT meeting at CEDES in Buenos Aires in June 2005. All mistakes are our own, of course.

2. Similar reasoning can be applied to the ‘pink collar’ aspects of the international pro-duction of services, a sector where women predominate in the administrative support aspect of the trade.

3. Our choice of countries was limited to those for which reasonably reliable, gender-disaggregated formal employment data were available.

4. In addition, the GDP defl ator for the USA used to construct the real exchange rate measure was taken from the online database of the US Bureau of Economic Analysis.

5. Other real US interest rates did increase around the time of the last infl ation reduction episode, for example, the yield on three-year government bonds.

6. The sample of countries includes Singapore which could arguably be classifi ed as a high-income country today. However, for much of the period considered in this chapter, 1970–2003, Singapore can be considered a middle-income country.

7. India is notable in terms of its frequent deviation from the behavior exhibited by other countries, both with respect to the gender dynamics of slower employment growth and the observed trends in real interest rates during infl ation reduction episodes.

8. Due to its extreme volatility, Brazil was excluded from this analysis. 9. Only in the case of Jamaica were the treasury-bills used to determine actual interest

rates and to estimate long-run trends.10. In contractionary episodes 36 percent of the episodes showed an appreciation, 58

percent a depreciation and 6 percent no diff erence in the average growth of the real exchange rate relative to its long-run trend. In expansionary episodes the percentages were 34 percent, 60 percent and 7 percent, respectively.

11. The defi nition of the money supply used is M2, money plus quasi-money, which includes cash; checking accounts; and relatively liquid deposits like savings and money market deposit accounts.


REFERENCES

Algan, Y. and P. Cahuc (2004), ‘Job protection: the macho hypothesis’, Society for Economic Dynamics 2004 meeting papers no. 332, September.

Anker, R. and C. Hein (1985), ‘Why Third World urban employers usually prefer men’, International Labour Review, 124 (1), 73–90.

Arriagada, I. (1994), ‘Changes in the urban female labour market’, CEPAL Review, 53, 92–110.

Aslanbeigui, N. and G. Summerfi eld (2000), ‘The Asian crisis, gender, and the international fi nancial architecture’, Feminist Economics, 6 (3), 81–103.

Azmat, Ghazala Maia Gvell and Alan Mangging (2004), ‘Gender gaps in unem-ployment rates in OECD countries’, London School of Economics Centre for Economic Policy discussion paper no. 604.

Bakker, I. (ed.) (1994), The Strategic Silence: Gender and Economic Policy, London. Zed Books with the North-South Institute.

Ball, L. (1993), ‘What determines the sacrifi ce ratio?’, National Bureau for Economic Research working paper no. 4306.

Benería, L. (2001), ‘Shifting the risk: new employment patterns, informalization, and women’s work’, International Journal of Politics, Culture and Society, 15 (1), 27–53.

Benería, L. (2003), Gender, Development, and Globalization: Economics as if All People Mattered, London: Routledge.

Benería, L. and S. Feldman (eds) (1992), Unequal Burden: Economic Crises, Persistent Poverty, and Women’s Work, Boulder, CO: Westview Press.

Benería, L. and M. Roldán (1987), The Crossroads of Class and Gender: Industrial Homework, Subcontracting, and Household Dynamics in Mexico City, Chicago, IL and London: University of Chicago Press.

Blumberg, R. (1991), ‘Income under female versus male control’, in R. Blumberg (ed.), Gender, Family and Economy: The Triple Overlap, Newbury Park, CA: Sage Press, pp. 97–127.

Braunstein, E. and N. Folbre (2001), ‘To honor and obey: effi ciency, inequality and patriarchal property rights’, Feminist Economics, 7 (1), 25–44.

Cagatay, N. and S. Ozler (1995), ‘Feminization of the labor force: the eff ects of long-term development and structural adjustment’, World Development, 23 (11), 1883–94.

Cagatay, N., D. Elson and C. Grown (eds) (1995), ‘Gender, adjustment and macro economics’, World Development, 23 (11) (special issue), 145–390.

Carr, M., M. Chen and J. Tate (2000), ‘Globalization and home-based workers’, Feminist Economics, 6 (3), 123–42.

Casale, D. (2003), ‘The rise in female labour force participation in South Africa: an analysis of household survey data, 1995–2001’, PhD dissertation, Division of Economics, University of KwaZulu-Natal.

Cerrutti, M. (2000), ‘Economic reform, structural adjustment and female labor force participation in Buenos Aires, Argentina’, World Development, 28 (5), 879–91.

Chant, S. (1991), Women and Survival in Mexican Cities: Perspectives on Gender, Labour Markets and Low-Income Households, New York: Manchester University Press.

Dollar, D. and R. Gatti (1999), ‘Gender inequality, income and growth: are good


times good for women?’, World Bank policy research report on gender and development working paper series no. 1, Washington, DC.

Dwyer, D. and J. Bruce (1988), A Home Divided: Women and Income in the Third World, Palo Alto, CA: Stanford University Press.

Elson, D. (ed.) (1991), Male Bias in the Development Process, Manchester: Manchester University Press.

Elson, D. (1995), ‘Gender awareness in modeling structural adjustment’, World Development, 23 (11), 1851–68.

Elson, D. (1996), ‘Appraising recent developments in the world market for nimble fi ngers’, in A. Chhachhi and R. Pittin (eds), Confronting State, Capital and Patriarchy: Women’ Organizing in the Process of Industrialization, New York: St Martin’s Press, pp. 35–55.

Elson, D. (1999), ‘Labor markets as gendered institutions: equality, effi ciency and empowerment issues’, World Development, 27 (3), 611–27.

Elson, D. and N. Cagatay (2000), ‘The social content of macroeconomic policies’, World Development, 28 (7), 1347–64.

Elson, D. and R. Pearson (1981), ‘Nimble fi ngers make cheap workers: an analy-sis of women’s employment in Third World export manufacturing’, Feminist Review, 7, 87–107.

Epstein, G. (2000), ‘Myth, mendacity and mischief in the theory and practice of central banking, mimeo, University of Massachusetts Amherst.

Epstein, G. (2003), ‘Alternative to infl ation targeting monetary policy for stable and egalitarian growth: a brief research summary’, Political Economy Research Institute working paper no. 62.

Epstein, G. and J. Schor (1990), ‘Macropolicy in the rise and fall of the golden age’, in S. Marglin and J. Schor (eds), The Golden Age of Capitalism: Reinterpreting the Postwar Experience, Oxford: Clarendon Press, pp. 126–52.

Fernández-Kelly, M. (1983), For We are Sold, I and My People: Women and Industrialization in Mexico’s Frontier, Albany, NY: SUNY Press.

Folbre, N. (1994), Who Pays for the Kids? Gender and the Structures of Constraint, London and New York: Routledge.

Folbre, N. (1997), ‘Gender coalitions: extrafamily infl uences on intrafamily inequality’, in L. Haddad, J. Hoddinott and H. Alderman (eds), Intrahousehold Resource Allocation in Developing Countries. Models, Methods, and Policy Baltimore, MD: Johns Hopkins University Press, pp. 263–74.

Fontana, G. and A. Palacio-Vera (2007), ‘Are long-run price stability and short-run output stabilization all that monetary policy can aim for?’, Metroeconomica, 58 (2), 269–98.

Fussell, E. (2000), ‘Making labor fl exible: the recomposition of Tijuana’s maqui-ladora female labor force’, Feminist Economics, 6 (3), 59–79.

Ghosh, J. (2001), ‘Globalisation, export-oriented employment for women and social policy: a case study of India’, paper prepared for United Nations Research Institute for Social Development project on Globalization, Export-Oriented Employment for Women and Social Policy, Geneva.

Hill, M. and E. King (1995), ‘Women’s education and economic well-being’, Feminist Economics, 1 (2), 21–46.

Hoddinott, J., H. Alderman and L. Haddad (eds) (1998), Intrahousehold Resource Allocation in Developing Countries: Methods, Models and Policy, Baltimore, MD: Johns Hopkins University Press.

Joekes, S. (1999), ‘A gender-analytical perspective on trade and sustainable


development’, in Trade, Sustainable Development and Gender, New York and Geneva: United Nations Conference on Trade and Development, pp. 33–59.

Kabeer, N. and S. Mahmud (2004), ‘Globalization, gender, and poverty: Bangladeshi women workers in export and local markets’, Journal of International Development, 16, 93–109.

Klasen, S. (1999), ‘Does gender inequality reduce growth and development? Evidence from cross-country regressions’, World Bank policy research report on gender and development working paper series no. 7, Washington, DC.

Lim, Joseph (2000), ‘The eff ects of the East Asian crisis on the employment of women and men: the Philippine case’, World Development, 28 (7), 1285–306.

Patnaik, U. (2003), ‘Global capitalism, defl ation and agrarian crisis in developing countries’, United Nations Research Institute for Social Development, social policy and development programme paper no. 15.

Seguino, S. (2003), ‘Why are women in the Caribbean so much more likely than men to be unemployed?’, Social and Economic Studies, 52 (4), 83–120.

Sen, A. (1990), ‘Gender and cooperative confl icts’, in I. Tinker (ed.), Persistent Inequalities: Women and World Development New York and Oxford: Oxford University Press, pp 123–49.

Standing, G. (1999), ‘Global feminization through fl exible labor: a theme re-visited’, World Development, 27 (3), 583–602.

Tzannatos, Z. (1999), ‘Women and labor market changes in the global economy: growth helps, inequalities hurt and public policy matters’, World Development, 27 (3), 551–69.

United Nations (1999), World Survey on the Role of Women in Development. Globalization, Gender and Work, New York: United Nations.

116

6. Infl ation and economic growth: a cross-country non-linear analysisRobert Pollin and Andong Zhu1

6.1 INTRODUCTION

This chapter presents new cross-country evidence between 1961 and 2000 on the relationship between infl ation and economic growth. Despite the central importance of this infl ation-growth relationship for macroeco-nomic theory and policy, there is nothing close to a professional consensus as to what the empirical evidence tells us about this relationship.

The results we present here have direct relevance to the debate on infl a-tion targeting as an appropriate framework for conducting monetary policy. Over the past decade governments throughout the world have embraced infl ation targeting as a dominant policy framework. For the most part, this specifi cally means that they have set a low band of accept-able infl ation rates as a primary target in the conduct of economic policy. This band is usually between a 3–5 percent annual infl ation rate. They have then maintained suffi ciently high short-term interest rates as the intermediate policy instrument for preventing infl ation from exceeding that target band. Higher interest rates are aimed, in turn, at reducing eco-nomic growth. Slower economic growth should then dampen infl ationary pressures. At least in the short run, the costs in terms of slower growth of containing infl ation within this 3–5 percent band are evident. But pro-ponents of infl ation targeting hold that, over a longer term framework, maintaining low infl ation will itself yield benefi ts for growth that exceed these short-term costs.2

Some limitations of infl ation targeting have been widely recognized by mainstream economists and even US central bankers Ben Bernanke and Alan Blinder (see Bernanke et al., 1999; Blinder, 1998). The Bernanke/Blinder view is that infl ation targeting does not provide an approach to maintaining low infl ation that is clearly superior to other approaches. This is true as such, but this concern about infl ation targeting as an operating procedure alone begs a more important question. This crucial question is whether maintaining infl ation within a band of 3–5 percent itself is, as a

Infl ation and economic growth: a cross-country analysis 117

generalization, supportive of economic growth, regardless of the technique being used to maintain infl ation within that low band. It is this broader question that we address in this chapter. That is, are countries making sac-rifi ces in terms of their economic growth path by focusing macroeconomic policy on maintaining infl ation at no more than 3–5 percent?

In Section 6.2 we briefl y review the overarching and longstanding analytic debates on the relationship between infl ation and economic growth, then focus specifi cally on the recent econometric research that has explored that relationship. In Section 6.3 we present basic descriptive data from our data sample, then examine the main results from our various economic exercises. In the concluding Section 6.4 we consider the broader implications of our fi ndings, especially as they relate to policy debates around infl ation targeting and possible alternative approaches to infl ation control.

6.2 LITERATURE OVERVIEW

6.2.1 Analytic Perspectives

We begin by separating out the phenomenon of hyperinfl ation, which we broadly defi ne as being annual infl ation rates in excess of 40 percent per year. Hyperinfl ations occur through a variety of specifi c factors. But regardless of their specifi c origins, hyperinfl ations represent a breakdown of economic functionings. We will assume that hyperinfl ations corre-spond with, and are detrimental to, a positive economic growth path. We are therefore leaving aside here the possibility that there may be some positive correspondence between infl ation above 40 percent and economic growth.

Hyperinfl ations aside, the relationship between infl ation and growth has been at the very center of macroeconomic theory debates since the mon-etarist counterrevolution against Keynesianism beginning in the 1960s.3 The main progeny of that counterrevolution – the ‘natural rate of unem-ployment’, the vertical Phillips Curve, and New Classical Economics more generally – have been focused largely around demonstrating that there can be no positive benefi ts for economic growth or employment of operating an economy at anything above a minimal infl ation rate in the range of 2–3 percent. From this perspective, infl ation impedes effi cient resource alloca-tion by obscuring the signaling role of relative price changes, which, in turn, is the most important guide to effi cient economic decision making.

This position contrasts sharply with the Keynesian perspective and the early Phillips Curve models, which held that infl ation and economic


growth can be positively associated when infl ationary pressures emerge as a byproduct of rising aggregate demand. In this Keynesian framework it is not the case that infl ation is itself a positive engine of growth, certainly not a primary growth-inducing force. The point is rather that if rising aggregate demand is leading to increased growth, then some infl ation-ary pressures are likely to emerge in this scenario as a relatively benign byproduct. Within this Keynesian framework, there could also be reasons for infl ation and growth to be negatively correlated. This would occur when infl ation results from monopolistic pricing practices, exchange rate volatility or supply shocks. These problems can also be compounded when adequate policy interventions do not occur to dampen the infl ation-ary impulses induced by monopolistic pricing, exchange rate volatility or supply shocks.

6.2.2 Recent Empirical Studies

Probably the most infl uential recent contribution to the econometric lit-erature on infl ation and growth is that of Bruno and Easterly (with results presented in both Bruno (1995) and Bruno and Easterly (1998)). Bruno and Easterly examined the relationship between infl ation and economic growth for 127 countries between 1960 and 1992. Their examination of this data set is historical and descriptive. They do not present a formal econometric model.

Their key conclusion was that there is no robust evidence from this data sample demonstrating a trade-off between output growth and infl ation. More specifi cally, only on the basis of two conditions could one observe a negative growth-infl ation relationship at all in their data sample. These were: (1) the inclusion in the data sample of very high infl ation experi-ences, that is, rates of infl ation of 40 percent and higher and (2) increasing the frequency of the data observations. As they write, ‘The results get stronger as one goes from the cross-section to ten year averages to fi ve year averages to annual data’ (1998, p. 4).

Once one controls for these two factors, Bruno and Easterly found that average growth rates fell only slightly as infl ation rates moved up to 20–25 percent. For infl ation rates below 20 percent, Bruno concluded that ‘there is no obvious empirical evidence for signifi cant long-run growth costs’ (Bruno, 1995, p. 38). Moreover, of particular importance for our concerns with aggregate demand eff ects on infl ation and growth, Bruno found that during 1960–72, economic growth on average increased as infl ation rose, from negative or low rates to the 15–20 percent range. This is because, as Bruno explained, ‘in the 1950s and 1960s, low-to-moderate infl ation went hand in hand with very rapid growth because of investment demand


pressures in an expanding economy’ (ibid., p. 35). Thus, infl ation that results directly from economic expansion does not, according to Bruno’s fi ndings, create any signifi cant barriers to expansion.

Despite these fi ndings, Bruno still makes clear in his single-authored paper that he does not advocate complacency with respect to infl ation rates in the 20 percent region. According to Bruno, once infl ation moves into the 20 percent region, it is diffi cult to contain at this level. This is because, within the 20 percent infl ation region, the systems of indexing wages and fi nancial assets, as well as exchange rate adjustments, become more fre-quent. This then creates a momentum toward accelerating infl ation.

Neither Bruno alone nor Bruno and Easterly provide systematic evi-dence on behalf of Bruno’s concerns about infl ation within the 20 percent region. Nevertheless, Bruno is clear in his conclusion that ‘getting infl ation down to single digits is important even for longer-term growth reasons’ (ibid., p. 38). But even within this less systematic discussion on the dangers of infl ation in the 20 percent range, it is still notable that Bruno never sug-gests that infl ation needs to be pushed below a single-digit threshold – and specifi cally down into the 3–5 percent range advocated by proponents of infl ation targeting.

Since the Bruno and Easterly study, various researchers have examined the output growth-infl ation relationship through more formal techniques than those employed by Bruno and Easterly while still searching out, as with Bruno and Easterly, potential non-linearities. For example, in a 1998 paper, International Monetary Fund (IMF) economists Atish Ghosh and Steven Phillips combine panel regression techniques with non-linear treat-ment of the infl ation-growth relationship. They also utilize a decision-tree technique that, in their view, is more robust to outliers and non-linearities than is standard regression analysis. Their model draws from a data sample of IMF member countries over 1960–96. According to this model, they fi nd evidence of a negative infl ation threshold at 2.5 percent. But they also acknowledge that thresholds of 5 or 10 percent generate statistical results very similar to the 2.5 percent threshold.

A 2001 publication by two separate IMF economists, Moshin Khan and Abdelhak Senhadji, off ers two innovations relative to Ghosh and Phillips (Khan and Senhadji, 2001). The fi rst is their use of conditional least squares, a new non-linear estimation technique. The second, and more straightforward, innovation was to divide their data sample into industrial and developing countries. Based on this approach, they fi nd that the threshold level above which infl ation signifi cantly slows growth is 1–3 percent for industrial countries and 11–12 percent for developing countries.

More recently still, a 2004 paper by Burdekin et al. followed Khan and


Senhadji in allowing for diff erent threshold eff ects among the industrial and developing countries (Burdekin et al., 2004). They also allow for non-linearities in the growth-infl ation relationship through utilizing spline estimation techniques. The results from this research diverge sharply from Khan and Senhadji. In terms of point estimates, they found that the turning point for industrial countries was 8 percent while that for develop-ing countries was 3 percent.

In short, all of these studies are in broad concurrence with Bruno and Easterly as to the presence of non-linearities in the growth-infl ation rela-tionship. They also broadly concur with Bruno’s conclusion that the nega-tive eff ects of infl ation will occur somewhere below a 20 percent threshold, most likely in the single-digit range. However, they diverge sharply as to where the turning point occurs within a range of roughly 12 percent infl a-tion or less. Moreover, the two studies that adopted the simple innova-tion of dividing the sample between industrial and developing countries reached opposite conclusions as to which set of countries had a higher infl ation threshold. Thus, despite the deployment of sophisticated tech-niques for capturing the impact of non-linearities in the growth-infl ation relationship, major questions remain unresolved. In particular, there remains no robust evidence in support of a policy goal of maintaining an infl ation target in the range of 3–5 percent.

6.3 DESCRIPTIVE DATA AND ECONOMETRIC EVIDENCE

Our own model is a straightforward panel data model, in which we aim to isolate the eff ects of infl ation on economic growth through including a series of control variables as well as allowing for a non-linear component to the growth-infl ation relationship. Our data sample runs from 1961 to 2000, including data from a total of 80 countries. We have excluded from the model countries whose population is less than 2 million people. We do this to focus our empirical exercises on countries whose economies are minimally large enough so that the countries’ patterns of economic activ-ity can be understood as having features that are distinct to that country. Appendix 6.1A provides a full list of the countries in our data sample.

6.3.1 Descriptive Statistics

We fi rst provide some basic descriptive statistics from our data sample in both Table 6.1 and Figure 6.1. Table 6.1 shows both means and standard deviations for infl ation and growth, for the full sample, and broken out


according to our three income-level groupings. For all countries in the sample, as we see, the average rate of GDP growth is 1.9 percent and the average infl ation rate is 10.2 percent. However, from the standard deviations – 2.7 percent for GDP growth and 7.2 percent for infl ation – we also see that there are wide disparities among the observations in the sample.

The disparities do diminish as we break out the full sample of countries into income-level groupings. Not surprisingly, the OECD countries expe-rience the highest average rate of economic growth (virtually by defi nition; see Note 4) and the lowest average infl ation rates. Average growth is sig-nifi cantly faster in the middle-income countries relative to the low-income countries, but average infl ation is somewhat lower in the low-income countries.

The four scatter plots in Figure 6.1 show the range of values for our data sample more fully. No strong patterns at all emerge from these fi gures in terms of the infl ation/GDP growth relationship. Of course, these data plots do not control for factors other than infl ation that could be aff ecting economic growth.

We label in the four diagrams in Figure 6.1 the data points that emerge as outliers through simple observation. This provides some useful perspec-tive. For example, with the full set of countries, the most rapid growth spurt was experienced by Haiti from 1996 to 2000. Haiti grew on average by 15.2 percent in this period, even while infl ation was rising at an average

Table 6.1 Descriptive statistics on GDP growth and infl ation 80 country sample, 1961–2000

All countries

(80 countries) (%)

OECD countries

(21 countries) (%)

Middle-income countries

(32 countries) (%)

Low-income countries

(25 countries) (%)

GDP growthMean 1.9 2.6 1.8 0.9Standard deviation 2.7 1.8 2.7 3.4Infl ationMean 10.2 7.0 12.8 11.3Standard deviation 7.2 4.9 8.3 6.6

Note: Israel and Singapore are not included in the country groupings because they are non-OECD high-income countries.

Source: See Appendix 6.1A.


of 14.9 percent. In terms of other outliers in the all-country diagram, we see that the very high infl ation and/or very low growth outliers are all low- or middle-income countries, with Rwanda, Nicaragua and Zimbabwe all experiencing severe political confl icts during their low growth/high infl a-tion years.

With the OECD and middle-income country diagrams, we see that the countries able to experience the most rapid economic growth rates were Japan, Ireland, South Korea and China. In all cases the rapid GDP growth was tied to reaching new levels of export success. Infl ation in these

–8

–4

0

4

8

12

16

–10 0 10 20 30 40 50

Inflation rate (%)

GD

P g

row

th r

ate

(%)

All countries

Haiti1996–2000

Venezuela1986–90

Ecuador1991–95

Rwanda1991–95

Nicaragua1976–81

Costa Rica1981–85

Zimbabwe1996–2000

–2

0

2

4

6

8

10

12

0 4 8 12 16 20 24

Inflation rate (%)

GD

P g

row

th r

ate

(%)

OECD countries

Japan1966–70

S. Korea1986–90

Ireland1996–2000

Greece1986–90

Puerto Rico1976–81

Puerto Rico1981–85

S. Korea1966–70

New Zealand1976–80

Finland 1991–95

Switzerland 1991–95

–8

–4

0

4

8

12

–10 0 10 20 30 40 50

Inflation rate (%)

GD

P g

row

th r

ate

(%)

Middle-income countries

China1991–95

Venezuela1986–90

Ecuador1971–75

Costa Rica1981–85

Ecuador 1991–95

Venezuela1971–75 Jordan 1986–90

–8

–4

0

4

8

12

16

0 10 20 30 40

Inflation rate (%)

GD

P g

row

th r

ate

(%)


Haiti1996–2000

Rwanda1991–95

Nicaragua 1976–80

Ghana1986–90

Zimbabwe1996–2000

Note: Annual data are grouped into fi ve-year averages.

Source: See Appendix 6.1A.

Figure 6.1 Infl ation and economic growth, 1961–2000


countries over the relevant years ranged widely, between 2.6 and 13.2 percent. Clearly, it is diffi cult to off er generalizations from these fi gures as to the interrelationship between infl ation and economic growth. It is evident that we need to examine this relationship more systematically, the task to which we now turn.

6.3.2 Econometric Model

Our approach has been to build a formal model that is still consistent with the main strength of the Bruno-Easterly framework, which is its simplicity. To do this, we work with a panel model that incorporates non-linearities through two relatively simple procedures.

The fi rst feature of our non-linear model is to simply exclude from our data set all observations in which infl ation exceeded 40 percent. As we mentioned above, we accept the fi nding of Bruno and Easterly that infl a-tion in that high range will produce negative eff ects on growth. We are therefore eff ectively asking with our model whether an annual infl ation rate below 40 percent exerts a negative eff ect on economic growth, and if so, at what point are such negative eff ects likely to emerge?

The second way that we introduce non-linearity in our model is to include the squared term on infl ation as an explanatory variable, which means we are estimating the regression equations as a second-degree polynomial. This is a straightforward, widely used technique for estimat-ing non-linear relationships, through allowing for changes in slopes as a function of changes in the independent variable. In this case the slope of the estimating equation can vary with changes in the infl ation rate. This enables us to observe turning points in the relationship between infl ation-growth and infl ation-equality. We can observe such possible turning points through this calculation:

Turning point 5 2((inflation coefficient) /

(2 * (inflation 2 squared coefficient)) .

Within this framework, we then also pursue robustness tests through three sets of straightforward procedures:

1. We utilize four diff erent panel data techniques: pooled ordinary least squares (OLS), between eff ects fi xed eff ects, and random eff ects. In principle, researchers are supposed to establish through diagnostic exercises which of the four techniques is appropriate with a given data sample. In practice, however, it is frequently diffi cult to know


which technique is the most reliable. Each of the techniques has both strengths and weaknesses. A pooled OLS model implicitly assumes there are no problems of omitted variables in a model, which is not likely to be true, even through frequently the problems may not be serious enough to substantially distort one’s results. A between eff ects model averages the data for each country into one observa-tion. It is therefore testing more narrowly for variation on a country-by-country basis, as opposed to considering variation between time periods as well as countries. With the fi xed eff ects model, we are allowing for intercept shifts to occur for each country, based on the range of possible omitted variables in evaluating country-by-country determinants of economic growth. But the fi xed eff ects model eff ectively creates dummy variables for each country in the sample, which reduces degrees of freedom. Finally, the random eff ects model also allows for a diff erent intercept for each country in the sample. But the random eff ects model isolates these individual country eff ects in the error term, and therefore does not reduce degrees of freedom in the manner of the fi xed eff ects estimator. But, at the same time, to be an unbiased estimator, the random eff ects model requires that the omitted variable eff ects will be uncorrelated with the explanatory variables.

Given this range of concerns with the various techniques, we report results utilizing all four techniques. By examining results generated by all four techniques, we are able to assess the robustness of our fi ndings across the range of panel data estimators.

2. We run regressions both on the full set of countries as one sample, then through dividing the countries into three groupings, OECD countries, middle-income countries and low-income countries. We are therefore able to observe the extent to which diff erences in the results are due to broad diff erences in the various countries’ level of development, as distinct from the individual country diff erences that we control for through the fi xed and random eff ects models.4

3. We decompose the full time period into four decade-long sub-periods. This enables us to examine how the relationships may have changed over time. We are especially interested in following up on Bruno’s observation that infl ation and growth were positively correlated from the 1960s up until the 1973 oil shock. This was a period in which, as Bruno said, infl ation emerged out of explicit eff orts to stimulate aggre-gate demand.

Beyond these distinct features of our model, we also incorporate a set of control variables in each specifi cation of the model. These control variables are standard in cross-country estimates of the


determinants of economic growth. They include (1) the initial level of GDP; (2) the share of investment spending in GDP; (3) the share of government spending in GDP; (4) the fi scal defi cit; (5) educational levels; (6) the level of overall health, as measured by life expectancy; (7) the change in terms of trade; (8) the eff ects of natural disasters; and (9) the eff ects of wars. In addition, we include dummy variables for each year in the pooled OLS, fi xed eff ects and random eff ects models to control for the time eff ects within each set of country observations. Full descriptions of each of the control variables is reported in the Appendix 6.1B. We do not report here the full set of results on the control variables, but these results are available on request.5

We report the key fi ndings of our econometric models in Tables 6.2 and 6.3. Both tables report the coeffi cients and t-statistics for the infl a-tion and infl ation-squared variables only for each of the regressions. We also report the turning points estimated by each equation when infl ation switches from becoming a positive to negative, or negative to positive, infl uence on growth.

Results for the full time periodThe results for the full time period are presented in Table 6.2. Considering fi rst the data for all countries in the sample, we see that the sign of the infl ation coeffi cient is consistently positive across specifi cations, and is statistically signifi cant in all but the fi xed eff ects specifi cation. Moveover, the coeffi cient values across specifi cations are similar, ranging between 0.09–0.15. The coeffi cients on the infl ation-squared terms are also similar and statistically signifi cant in all cases. With the coeffi cients on the infl a-tion and infl ation-squared terms, we can then calculate the turning points as being between a 15.2 and 18.6 infl ation rate. Overall, this fi rst set of tests with the full data sample suggest that the rate of economic growth rises by between about 0.1 and 0.15 percent for every percentage point increase in the infl ation rate up to a 15–18 percent threshold. Infl ation then becomes a damper on growth beyond this threshold.

The clear fi ndings we obtain with the full data set is, however, not main-tained when we consider OECD, middle-income and low-income coun-tries separately. With the data grouped by income levels, we expect that the signifi cance levels will go down due to the smaller sample sizes. And we do indeed observe generally lower signifi cance levels with the results grouped by income levels.

More specifi cally, in the case of the OECD countries, none of the coef-fi cients for infl ation or infl ation-squared are statistically signifi cant in any of the specifi cations. Moreover, the signs on the infl ation variable shift to negative in the pooled OLS, between eff ects and random eff ects models. In

126

Tab

le 6

.2

Infl a

tion

and

econ

omic

gro

wth

, 196

1–20

00

Mod

elA

ll co

untr

ies

OE

CD

cou

ntri

es

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

No.

of o

bser

vatio

ns:

356

356

356

8013

513

513

521

Infl a

tion

0.11

*0.

091

0.11

*0.

149*

−0.

055

0.02

5−

0.03

4−

0.13

0(2

.49)

(1.6

1)(2

.55)

(1.9

9)(−

0.66

)(0

.23)

(−0.

36)

(−0.

30)

(Infl

atio

n)−

0.00

3**

−0.

003*

−0.

003*

*−

0.00

4*−

0.00

5−

0.00

7−

0.00

50.

001

(−3.

73)

(−2.

41)

(−3.

42)

(−2.

35)

(−1.

43)

(−1.

79)

(−1.

37)

(0.0

6)T

urni

ng P

oint

18.3

15.2

18.3

18.6

−5.

51.

8−

3.4

65.0

Mod

elM

iddl

e-in

com

e co

untr

ies

Low

-inco

me

coun

trie

s

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

No.

of o

bser

vatio

ns12

712

712

732

8686

8625

Infl a

tion

0.06

0.02

80.

057

0.12

90.

359

0.55

9*0.

386*

0.23

8(1

.12)

(0.3

1)(0

.84)

(1.1

9)(1

.38)

(2.3

8)(2

.29)

(0.7

3)(I

nfl a

tion)

−0.

002*

−0.

001

−0.

002

−0.

004

−0.

01−

0.01

2*−

0.01

*−

0.00

8(−

2.45

)(−

0.84

)(−

1.67

)(−

1.71

)(1

.55)

(2.2

0)(2

.36)

(0.7

2)T

urni

ng P

oint

15.0

14.0

14.3

16.1

18.0

23.3

19.3

14.9

Not

es:

1.

Dat

a gr

oupe

d as

fi ve

-yea

r ave

rage

s; de

pend

ent v

aria

ble

is G

DP

grow

th; t

-sta

tistic

s in

pare

nthe

ses;

p ,

0.0

5 5

*, p

, 0

.01

5 *

*.2.

A

ll co

untr

ies

with

less

than

2 m

illio

n pe

ople

wer

e ex

clud

ed fr

om th

e sa

mpl

e. A

ll ob

serv

atio

ns w

ith in

fl atio

n ab

ove

40 p

erce

nt w

ere

excl

uded

. F

ive-

year

per

iod

aver

age

data

wer

e us

ed in

stea

d of

yea

rly d

ata

in th

e re

gres

sion.

‘All

coun

trie

s’ re

fers

to a

ll th

e co

untr

ies i

n th

is ta

ble.


short, we do not obtain any reliable results on the infl ation-growth rela-tionship for the OECD countries.

With the middle-income countries, the signs on the infl ation coeffi cient are all positive. However, the coeffi cients are insignifi cant in all cases, the coeffi cient values correspondingly jump from 0.06 to 0.129. However, the estimated turning points in these equations are within a tight band of between 14–16 percent.

Finally, with the low-income countries, we do again obtain consistently positive coeffi cient values from the infl ation variable. These coeffi cients are also signifi cant in the fi xed eff ects and random eff ects models. The coeffi cient values in these regressions are substantially higher than with the other country groupings, ranging between 0.24 and 0.56. The infl ation-squared terms are also strongly signifi cant in the fi xed and random eff ects models. The turning point estimates range between 15–23 percent.

Results by decadeIn Table 6.3 we report results from regressions run separately for each of the four decades in our data sample. These regressions are run with annual data rather than fi ve-year averages in order to generate a larger number of observations. In doing this, we recognize the point, emphasized by Bruno and Easterly, that we should expect more of a negative correspond-ence between infl ation and growth as we move to higher frequency data samples. This is because of the likelihood that negative eff ects on growth will occur through short bursts of high infl ation rates that approach our cut-off fi gure of 40 percent. Such short bursts of high infl ation and slow or negative growth will be smoothed out when data are grouped at lower frequencies.

One key point emerges from the results in Table 6.3: that the evidence for a positive association between growth and infl ation is far stronger in the 1961–70 decade than in subsequent decades. For 1961–70, the coef-fi cient values on infl ation are all positive, though statistically signifi cant only in the between eff ects model. The fi xed eff ects model stands apart with a low infl ation coeffi cient value of 0.065, but otherwise the coeffi c-ients for the other specifi cations are high, at 0.11 for the pooled OLS and random eff ects models and a very high 0.61 for the between eff ects model.

With the 1971–80 sample, the infl ation coeffi cients remain positive, but the coeffi cient values and levels of signifi cance fall off , especially with the random eff ects and between eff ects models. For 1981–90, the infl ation coeffi cients all turn negative, and there is a statistically signifi cant negative value in the fi xed eff ects model. Finally, for 1991–2000, we obtain negative infl ation coeffi cients with two tests and close to zero coeffi cients for the other two.

128

Tab

le 6

.3

Infl a

tion

and

econ

omic

gro

wth

all

coun

trie

s by

deca

des

Yea

rs19

61–7

019

71–8

0

Mod

elP

oole

dO

LS

Fixe

d eff

ect

sR

ando

m

eff e

cts

Bet

wee

n eff

ect

sP

oole

dO

LS

Fixe

d eff

ect

sR

ando

m

eff e

cts

Bet

wee

n eff

ect

s

No.

of o

bser

vatio

ns48

048

048

059

620

620

620

71In

fl atio

n0.

112

0.06

50.

109

0.61

*0.

084

0.06

0.08

40.

047

(1.0

4)(.0

70)

(1.2

4)(2

.08)

(0.9

9)(0

.61)

(0.9

2)(0

.20)

(Infl

atio

n)−

.000

5−

0.00

7*−

0.00

6−

0.02

4*−

0.00

5*−

0.00

6*−

0.00

5*0.

001

(1.4

7)(2

.10)

(1.7

4)(2

.08)

(2.0

0)(2

.21)

(2.0

3)(0

.13)

Tur

ning

Poi

nt11

.24.

69.

112

.78.

45.

08.

4−

23.5

Yea

rs19

81–9

019

91–2

000

Mod

el

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

Poo

led

OL

SFi

xed

eff e

cts

Ran

dom

eff

ect

sB

etw

een

eff e

cts

No.

of o

bser

vatio

ns71

871

871

885

698

698

698

106

Infl a

tion

−0.

016

−0.

118*

−0.

016

−0.

016

0.02

5−

0.13

0.01

−0.

189

(0.3

1)(1

.98)

(0.3

3)(0

.17)

(0.3

1)(1

.69)

(0.1

5)(1

.32)

(Infl

atio

n)0.

000.

001

0.00

0.00

3−

0.00

3−

0.00

1−

0.00

30.

004

(0.0

08)

(0.6

1)(0

.08)

(0.9

1)(1

.30)

(0.4

3)(1

.70)

(0.8

7)T

urni

ng P

oint

–59

.0–

2.7

4.2

−65

.01.

723

.6

Not

es:

1.

Ann

ual d

ata;

dep

ende

nt v

aria

ble

is G

DP

grow

th; t

-sta

tistic

s in

pare

nthe

ses;

p ,

0.0

5 5

*, p

, 0

.01

5 *

*).

2.

All

coun

trie

s w

ith le

ss th

an 2

mill

ion

peop

le w

ere

excl

uded

from

the

sam

ple.

All

obse

rvat

ions

with

infl a

tion

abov

e 40

per

cent

wer

e ex

clud

ed.

Fiv

e-ye

ar p

erio

d av

erag

e da

ta w

ere

used

inst

ead

of y

early

dat

a in

the

regr

essio

n. ‘A

ll co

untr

ies’

refe

rs to

all

the

coun

trie

s in

this

tabl

e.


These results provide broad support for Bruno’s observation cited above, about infl ation and growth moving positively together during the 1960s in correspondence with, as he put it, ‘very rapid growth because of investment demand pressures in an expanding economy’ (Bruno, 1995, p. 35). During the 1970s, demand management policies were still in favor to support growth. But the positive associations between growth and infl a-tion as a byproduct of growth in this period were undermined by the two oil price shocks in 1973 and 1979. The overall infl ation experience of this decade therefore is a combination of demand-pull eff ects from growth and supply side shocks. It is therefore not surprising that the infl ation coeffi c-ients in the 1970s fall in value and lose signifi cance.

The 1980s marked the beginning of what Angus Maddison (2001), among others, has term the ‘neoliberal era’. Probably the single defi ning feature of this era is the virtual abandonment by governments throughout the world of Keynesian demand management policies as a tool for stimu-lating growth and employment. Thus, as a broad generalization, the infl a-tion that is experienced in the 1980s and 1990s emerges almost entirely as a result of supply shocks and inertia, as opposed to demand-pull pressures. Within this context it is also not surprising that the infl ation coeffi cients become consistently negative, albeit generally not to a statistically signifi -cant extent.

6.4 CONCLUSIONS

Considering fi rst our full data set of 80 countries between 1961 to 2000, we have consistently found that higher infl ation is associated with moderate gains in GDP growth up to a roughly 15–18 percent infl ation threshold.

However, the fi ndings diverge when we divide our full data set accord-ing to income levels. With the OECD countries, no clear pattern emerges at all with either the infl ation coeffi cient or our estimated turning point. Both the signs on the infl ation coeffi cients as well as the turning points are highly sensitive to specifi cations. With the middle-income countries, by contrast, we return to a consistently positive pattern of infl ation coef-fi cients, though none are statistically signifi cant. However, the turning points range within a narrow band in this sample, between 14.0 and 16.3 percent. With the low-income countries, we obtain positive and higher coeffi cient values on the infl ation coeffi cient than with the middle-income countries. These coeffi cients are also statistically signifi cant with the fi xed and random eff ects models.

Finally, with the groupings by decade, the results broadly indicate that infl ation and growth will be more highly correlated to the degree that


macroeconomic policy is focused on demand management as a stimulus to growth as opposed to macroeconomic austerity and infl ation targeting.

Overall, there is no evidence from this research supportive of a policy of maintaining infl ation within a low band of about 3–5 percent, to the degree that government policy makers are interested in promoting eco-nomic growth and employment, rather than merely low infl ation as an end in itself. At the same time, there is also no evidence that governments should allow infl ation to rise above a 15–20 percent range in an eff ort to spur growth.

This suggests that there is still a wide range of infl ation rates that are very likely to be associated positively with economic growth. Certainly for the middle- and low-income countries, our results strongly suggest that allowing infl ation to be maintained in the range of 10 percent or somewhat higher is very likely to be consistent with higher rates of economic growth. This is most especially the case when infl ation is resulting from, as Bruno put it, ‘investment demand pressures in an expanding economy’.

For the OECD countries, the primary conclusion that we can reach from our results is a negative one: that no generalization about the infl ation-growth relationship is likely to fi nd robust support from the available evidence. What appears likely for the OECD countries is that the wide range of relationships that emerge from the data refl ect the dif-ferences in the sources of infl ation – that is, whether infl ation has resulted primarily from Keynesian type demand-pull forces as opposed to supply shocks and inertia.

Some broad policy implications fl ow from these results. The fi rst is that there is no justifi cation for infl ation targeting policies as they are currently being practiced throughout the world, that is, to maintain infl ation with a 3–5 percent band and to adjust short-term interest rates as needed to dampen infl ationary pressures beyond that targeted band. As a corollary, there is very likely to be positive growth benefi ts in middle- and low-income countries from allowing infl ation to rise to a high single-digit range or even in some cases up to about 15 percent rather than dampening infl ationary pressures through raising short-term interest rates. This is especially true to the extent that infl ation within this range is resulting from demand-pull forces as opposed to supply shocks and inertia.6

A second implication is that researchers are likely to make productive contributions through giving increased attention on the infl ation-growth relationship to some relatively underexplored aspects of the issue. The fi rst is to be able to sort out with increased specifi city the sources of infl ationary pressures, given the likely wide disparities in the infl ation-growth relation-ship depending on what is fueling infl ation. A second is to focus more on policy measures for dampening infl ation not at very low levels, but rather


at levels approaching the upper limit of the positive growth-infl ation association. This would be in the range of 10–15 percent for middle- and low-income countries. With the OECD countries, the acceptable range is likely to depend entirely on what are the primary sources of infl ationary pressures.

One well-known policy tool for maintaining infl ationary pressures within a positive threshold range is some variation of incomes policies. Incomes policies have been widely used as an infl ation control tool in a variety of contexts. One common situation has been in bringing down infl ation after it has risen to a range above 40 percent. For example, in their paper ‘Moderate infl ation’ Dornbusch and Fischer (1991) describe how Mexico in the 1980s drew upon experiences in Argentina, Brazil, Peru and Israel in developing a strategy to bring infl ation down from the 100 percent range to something closer to 20 percent. Dornbusch and Fischer reported that the Mexicans learned two lessons from these experiences, (1) ‘that disinfl ation without fi scal discipline was unsustainable’; but that (2) ‘disinfl ation without incomes policy, relying solely on tight money and tight budgets, would be unnecessarily expensive’ (p. 31). In analysing the Israeli experience with disinfl ation over the 1980s, Bruno documents in detail the major contributions of incomes policies to the success of the eff ort (Bruno 1993, Chapter 5).

A more directly relevant set of experiences with respect to infl ations at more moderate levels have been the Nordic countries. This is because, in these countries, incomes policies have been used successfully as a tool for maintaining relatively low infl ation over long periods of time rather than as primarily an instrument of disinfl ation after infl ation exceeded 40 percent, as was true with Mexico and Israel. Sweden, for example, suc-ceeded in maintaining unemployment at an average rate below 2 percent between 1951 and 2000 while still holding infl ation at a 4.4 percent average rate. The application of incomes policies in Sweden, moreover, primarily took the form of centralized bargaining between unions and business, through which the aim of infl ation control was recognized in the bargain-ing process. As such, the government did not have to rely on setting man-dates for acceptable wage and price increases. The government did also utilize fi scal and monetary policies as tools for controlling infl ation. But they did not have to apply these tools stringently, precisely because they were able to rely on their well-developed system of incomes policies as a complement to monetary and fi scal policies.7

The most basic critique of incomes policies is that, in order for the approach to have any chance of success, it is necessary that a country operate with a high level of organization among workers, and that there be some reasonable degree of common ground between workers and


business. Otherwise, there will be no realistic prospect for economy-wide bargaining to yield results that will be honored widely. By its very nature, the relationship between unions and business in capitalist economies is likely to be highly contentious. But this could possibly diminish to the extent that both sides see the benefi ts accelerated economic growth and employment expansion as opposed to maintaining tight monetary policy for the purpose of holding infl ation within a 3–5 percent band.

This point brings us to a fi nal issue for further research that includes both purely analytic as well as policy oriented implications. This is to examine the relationship between infl ation and inequality in addition to the infl ation-growth relationship. To the extent that infl ation is associ-ated with faster economic growth, it is likely to also be correlated with faster employment growth and thereby increased equality. At the same time, to the extent that wage agreements and social benefi ts do not include adequate cost-of-living adjustments, even a growth-generated infl ation could yield greater inequality. In terms of policy implications, the issues that are central in the exploration of the infl ation-inequality relationship will also be closely linked to the question of infl ation control policies. For example, are incomes policies or infl ation targeting a more eff ective means of promoting greater equality as well as economic growth? These are crucial questions that deserve substantial additional research in an eff ort to design more eff ective analytic foundations for the conduct of macro-economic policy.

NOTES

1. This chapter was written as part of a broader project on pro-poor macroeconomic poli-cies in sub-Saharan Africa, under the sponsorship of the United Nations Development Programme (UNDP). We are most grateful for the support of the UNDP. We are also grateful for comments on this specifi c aspect of the broader project from our co-workers Jerry Epstein, James Heintz and Leonce Ndikumana as well as from Michael Ash.

2. An excellent survey of infl ation targeting and related issues in global monetary macro-economics is Saad Filho (2005).

3. The literature on this issue is of course vast. Three references off ering diff erent perspec-tives are Cross (1995), Krueger and Solow (2001), and Saad Filho (2005).

4. Grouping the countries in the sample by average GDP levels does raise the potential for signifi cant bias in the regression. This is because the dependent variable in the model is GDP growth. Strictly speaking, we are not dividing the sample based on the dependent variable, but there is obviously a close correspondence between the growth of GDP, our dependent variable and GDP levels, the variable on which we truncate the sample. To test for bias here, we have also divided the full sample based on pre-1960 GDP level groupings – that is, on the basis of data points that precede in time our sampling period. In this case the division is between current OECD countries and the non-OECD coun-tries – that is, the demarcation between middle- and low-income countries was not so evident in the pre-1960 data, and only becomes evident over the 40 years that constitute


our data sample. However, the results of this exercise do not vary substantially from those reported with the three GDP-level groupings reported here. This suggests that any potential bias from the GDP-level groupings is not a serious problem for our substantive understanding of the fi ndings.

5. By exploring the infl ation-growth relationship within this framework of a standard cross-country growth regression model, we are building in an assumption that causality in the relationship is running from infl ation to growth. We do not explore the issue of simultaneity or reverse causality in this exercise, while we recognize it as an important issue for further research.

6. Even some of the most recent work by IMF economists has recognized that, at least for the low-income countries, infl ation in the range of 5–10 percent is likely to be supportive of economic growth. See IMF (2005).

7. Diff erent perspectives on the Nordic experiences are presented in Calmfors (1993), Pekkarinen et al. (1992), Flanigan (1999), Marshall (1994) and Iversen et al. (2000).

REFERENCES

Bernanke, B., T. Laubach, A.S. Posen and F.S. Mishkin (1999), Infl ation Targeting: Lessons from the International Experience, Princeton, NJ: Princeton University Press.

Blinder, A.S. (1998), Central Banking in Theory and Practice, Cambridge, MA: MIT Press.

Bruno, M. (1993), Crisis, Stabilization, and Economic Reform: Therapy by Consensus, Oxford: Clarendon Press.

Bruno, M. (1995), ‘Does infl ation really lower growth?’, Finance and Development, 32 (3) (September), 35–8.

Bruno, M. and W. Easterly (1998), ‘Infl ation crises and long-run growth’, Journal of Monetary Economics, 41, 3–26.

Burdekin, R.C.K., A.T. Denzau, M.W. Keil, T. Sitthiyot and T.D. Willett (2004), ‘When does infl ation hurt economic growth? Diff erent nonlinearities for diff er-ent economies’, Journal of Macroeconomics, 26, 519–32.

Calmfors, L. (1993), ‘Centralization of wage bargaining and macroeconomic per-formance’, OECD Economic Studies, 21 (Winter), 161–91.

Cross, R. (ed.) (1995), The Natural Rate of Unemployment: Refl ections on 25 Years of the Hypothesis, Cambridge: Cambridge University Press.

Dornbusch, R. and S. Fischer (1991), ‘Moderate infl ation’, National Bureau of Economic Research working paper no. 3896.

Flanigan, R.J. (1999), ‘Macroeconomic performance and collective bargaining: an international perspective’, Journal of Economic Literature, 37 (3) (September), 1150–75.

Ghosh, A. and S. Phillips (1998), ‘Warning: infl ation may be harmful to your growth’, IMF Staff Papers, 45 (4), 672–86.

International Monetary Fund (IMF) (2005), ‘Monetary and fi scal policy design issues in low-income countries’, 8 August draft, Policy Development and Review Department and Fiscal Aff airs Department, IMF, Washington, DC.

Iversen, T., J. Pontusson and D. Soskice (eds) (2000), Unions, Employers, and Central Banks, Cambridge: Cambridge University Press.

Khan, M.S. and A.S. Senhadji (2001), ‘Threshold eff ects in the relationship between infl ation and growth’, IMF Staff Papers, 48 (1), 1–21.


Krueger, A. and R.M. Solow (eds) (2001), The Roaring Nineties: Can Full Employment Be Sustained?, New York: The Russell Sage Foundation and the Century Foundation.

Maddison, A. (2001), The World Economy: A Millennial Perspective, Paris: Development Centre of the Organisation for Economic Co-operation and Development.

Marshall, M. (1994), ‘Lessons from the experience of the Swedish model’, in P. Arestis and M. Marshall (eds), The Political Economy of Full Employment, Aldershot, UK and Brookfi eld, US: Edward Elgar, Ch. 10.

Pekkarinen, P., M. Pohjola and B. Rowthorn (1992), Social Corporatism: A Superior Economic System?, New York: Oxford University Press.

Saad Filho, A. (2005), ‘Pro-poor monetary and anti-infl ation policies: developing alternatives to the new monetary policy consensus’, manuscript, Department of Development Studies, School of Oriental and African Studies, University of London.


APPENDIX 6.1A COUNTRIES INCLUDED IN DATA POOL FOR ANALYSING THE INFLATION-ECONOMIC GROWTH RELATIONSHIP

Data Sample is 1960 to 2001

OECD countries

Middle-income countries Non-OECD high-income countries


Australia Algeria Paraguay Israel BangladeshAustria Argentina Peru Singapore BurundiBelgium Bolivia Philippines CameroonCanada Brazil Poland Central Afr. R.Denmark Chile South Africa CongoFinland China Sri Lanka GhanaFrance Colombia Syria HaitiGreece Costa Rica Thailand IndiaIreland Dominican

Rep.Tunisia Indonesia

Italy Ecuador Uruguay KenyaJapan Egypt Venezuela LesothoKorea El Salvador MalawiNetherlands Guatemala MaliNew Zealand Honduras NepalNorway Hungary NicaraguaPortugal Iran, I.R. of NigerSpain Jamaica PakistanSweden Jordan Papua New

GuineaSwitzerland Malaysia RwandaUnited

KingdomMexico Senegal

United States Panama Sierra LeoneTogoUgandaZaireZimbabwe


APPENDIX 6.2B SPECIFICATIONS OF VARIABLES IN THE FULL INFLATION/ECONOMIC GROWTH MODEL

Economic growth. Real GDP per capita (Constant price: Laspeyres) growth rate. The nth year’s growth rate is calculated as the log value of the ratio of the nth year’s per capita GDP to the (n-1)th year’s per capita GDP (Penn World Table (PWT) 6.1, http://pwt.econ.upenn.edu/).

Infl ation. The increase of consumer price index (World Bank World Development Indicators (WDI), 2003).

Initial output level. The log value of per capita GDP (Constant price: Laspeyres) at the beginning year of each period (PWT 6.1, http://pwt.econ.upenn.edu/).

Investment. The share of gross investment in GDP (current prices) (PWT 6.1, http://pwt.econ.upenn.edu/).

Fiscal policy. (1) The share of government consumption in GDP (current prices) (PWT6.1, http://pwt.econ.upenn.edu/). (2) Government budget defi cit as percentage of GDP (WDI, 2003).

Life expectancy. Life expectancy at birth (WDI CD-ROM, 2003, World Bank).

Education level. Average years of secondary schooling of the total popu-lation aged 25 and over (R. Barro and J.W. Lee, 2000, http://www2.cid.harvard.edu/ciddata/barrolee/panel_data.xls).

Terms of trade. The change of terms of trade weighted by foreign trade dependence ratio (the sum of exports and imports divided by GDP) (Easterly, William and Mirvat Sewadeh (2001), Global Development Network Growth database, www.worldbank.org/research/growth/GDNdata.htm).

Natural disaster. The share of population aff ected by the natural disas-ters happened in the year weighted by the share of agricultural output in GDP. Unreported natural disasters, if any, are treated as 0 (The natural disaster data come from the Centre for Research on the Epidemiology of Disasters (CRED), the OFDA/CRED International Disaster data-base, http://www.cred.be/emdat/intro.htm. The agricultural data are from WDI, 2003).

War. A war is defi ned as an armed confl ict with more than 25 deaths. Value 1 is given to those countries that experienced war within its border; –1 is given to those countries involved in war in other countries. Other situations are given value 0 (Gleditsch et al., 2002, Armed Confl ict 1946–2002 database, http://www.prio.no/cwp/ArmedConfl ict/).

PART III

Infl ation targeting: critiques and country-specifi c alternatives

139

7. Infl ation targeting in Brazil: 1999–2006Nelson H. Barbosa-Filho

Brazil adopted infl ation targeting after a brief period of exchange rate targeting that ended up in a major currency crisis. More specifi cally, in 1994–98 the Brazilian government used high domestic interest rates and privatization to attract foreign capital and sustain an appreciated exchange rate peg. The main objective of the Brazilian economic policy at that time was to reduce infl ation and the main side eff ect of exchange rate appreciation was a substantial increase in the country’s current account defi cit and net public debt. Similar to what happened in Mexico and Argentina during the 1990s, the Brazilian macroeconomic stabilization strategy was heavily dependent on the continuous infl ow of foreign capital and, as a result, the international fi nancial position of Brazil became increasingly fragile after the contagion eff ects from the East Asian cur-rency crises of 1997 and the Russian currency crisis of 1998. In fact, by the end of 1998 Brazil’s current account defi cit reached 4.5 percent of GDP and the low stock of foreign reserves of the Brazilian Central Bank (BCB) did not allow a defense of the Brazilian currency, the real, in case another speculative attack hit the country.

The inevitable currency crisis came in the beginning of 1999 and resulted in a ‘maxi-devaluation’ of the real. In numbers, the Brazilian real/US dollar exchange rate rose 57 percent in just two months, that is, from 1.21 in December 1998 to 1.90 in February 1999. After that the exchange rate dropped a little and then remained around 1.84 during the rest of 1999, that is, the exchange rate stabilized at a level 52 percent higher than before the crisis. The initial response of the BCB to such an abrupt devaluation of the real was a substantial increase in its base interest rate to stop the capital fl ight from Brazil and to reduce the pass-through of exchange rate depreciation to infl ation.1 Then, after the exchange rate stabilized at a higher level in mid 1999, the govern-ment announced that it would start to target infl ation. At that time the basic justifi cation for such a move was that the government needed to substitute a price target for the former exchange rate target in order


to coordinate market expectations and control infl ation in a context of fl oating exchange rates.

In terms of the classic policy ‘trilemma’ of open economies, the option for infl ation targeting meant that Brazil chose to have an indepen dent monetary policy, free capital fl ows and a fl oating exchange rate. In practice the situation was obviously diff erent because the exchange rate was an important determinant of the Brazilian infl ation rate during the period under analysis. On the one hand, the change in the domestic price of tradable goods in Brazil was (and it continues to be) basically deter-mined by foreign infl ation and exchange rate variations. On the other hand, the prices of some key non-tradable goods in Brazil were also tied to the exchange rate because, during the privatizations of the 1990s, the government allowed the price of some public utilities (basically in the telecommunication and energy sectors) to follow a price index that is heavily infl uenced by the exchange rate. The inevitable result was that a major part of the Brazilian infl ation rate was linked to the exchange rate in 1999–2006.2

Given the centrality of the exchange rate for Brazilian infl ation, it is no surprise that infl ation targeting resulted in a disguised and loose exchange rate targeting by the BCB in 1999–2006. In theory the exchange rate was free to fl oat, but in practice the country ended up with an asymmetric dirty fl oating regime, that is, a regime in which the BCB had to fi ght devalu-ations but to tolerate revaluations in order to meet the infl ation targets set by the Brazilian government. As a result, the interaction between the international fi nancial conditions, on the one side, and the Brazilian inter-est rates, on the other side, explains most of the behavior of the exchange rate in Brazil in 1999–2006, which in its turn explains most of the successes and failures of infl ation targeting during that period. The objective of this chapter is to describe and analyse this experience.3

7.1 THE BRAZILIAN INFLATION TARGETING REGIME

The institutional basis of the Brazilian infl ation targeting system can be described as follows:

The National Monetary Council (Conselho Monetário Nacional or ●

CMN), formed by the Minister of Finance, the Minister of Planning and the President of the BCB, establishes the infl ation targets based on the recommendations of the Finance Minister. All three members are appointed by the president and do not have fi xed mandates.

Infl ation targeting in Brazil: 1999–2006 141

In June of every year the CMN establishes the infl ation targets, and ●

their corresponding intervals of tolerance, for the next two years. The target consists of the desired variation of a consumer price index (the IPCA index) estimated by the government’s statistical branch (the Instituto Brasileiro de Geografi a e Estatística or IBGE).The BCB is responsible for achieving the target, but no specifi c ●

instrument or strategy is specifi ed.4

The Monetary Policy Committee (Comitê de Política Monetária ●

or Copom), formed by the president and the directors of the BCB, decides the level of the BCB’s base interest rate, the so-known SELIC (Sistema Especial de Liquidação e Custodia) rate, needed to achieve the infl ation target. Occasionally, additional actions such as changes in the banks’ reserve requirements can be taken.5

The target is considered met whenever the observed accumulated ●

infl ation during each calendar year falls within the interval of toler-ance specifi ed by the CMN.If the targets are not met, the president of the BCB has to issue an ●

open letter to the Minister of Finance explaining the causes of the failure, the measures to be adopted to ensure that infl ation returns to the target level, and the period of time needed for this to happen.

Since the BCB is not independent and its penalty for not meeting the infl ation target is just to explain why that happened, we can conclude that the Brazilian infl ation targeting regime is basically a loose way for the federal government to assure society, especially fi nancial markets, that it will not let infl ation run out of control.

7.2 MACROECONOMIC PERFORMANCE

The macroeconomic performance of Brazil since the adoption of infl a-tion targeting has been mixed. First, considering infl ation itself, the gov-ernment’s targets were met when the international fi nancial conditions allowed it, that is, infl ation targeting was successful when the exchange rate dynamics helped the BCB eff orts to control infl ation. Second, when compared to the period of exchange rate targeting, infl ation target-ing brought a reduction in the base domestic real interest rate, but the Brazilian rate remained well above the international standards because the country needed to appreciate the exchange rate in order to meet its infl a-tion targets. Third, the high real interest rate put an expansionary pressure on the net public debt and this had to be compensated by an increase in the government’s primary surplus, that is, the government budget surplus


excluding net interest payments. Fourth, the growth performance of the economy under infl ation targeting did not improve much when compared to the previous period of exchange rate targeting, and this happened even though the international environment was much more favorable for eco-nomic growth in 1999–2006 than in 1994–98.6 To facilitate the analysis, we look at each one of these issues separately below.

Starting with infl ation itself, Table 7.1 presents the annual infl ation targets set by the CMN and the eff ective rates of infl ation and exchange rate variation in 1999–2006. If we consider the whole period the infl ation targets were met in fi ve out of eight years, which on a fi rst look seems to be a very good starting performance of infl ation targeting in Brazil. However, when we look closer the real story cannot be considered that successful because the infl ation targets were frequently changed according to the shocks that hit the Brazilian economy and, most importantly, disinfl ation was obtained mostly through exchange rate appreciation. In fact, if we exclude 1999 from the analysis, the infl ation targets were met only when the exchange rate appreciated in nominal terms. To illustrate this point, Figure 7.1 shows that in 2000, 2004, 2005 and 2006 the target was met and the exchange rate appreciated, in 2001 and 2002 the target was not met and the exchange rate depreciated, and in 2003 the target was not met and the exchange rate appreciated.

Now, before we move to the real interest rate, let us make a brief pause to present some of the history behind the numbers shown in Table 7.1. First, since the Brazilian infl ation targeting regime started immediately after a currency crisis, there was a 7.2 percentage point (pp) increase in infl ation in 1999 when compared to 1998.7 Despite such acceleration, the government infl ation target for 1999 was met for two idiosyncratic reasons: fi rst, the 1999 target was set in June of that year, when the gov-ernment authorities already knew half of the annual infl ation rate of 1999 and, second, the target for 1999 was not ambitiously low in order not to undermine the credibility of the new policy regime.8

Continuing with our infl ation story, in 2000 the Brazilian exchange rate stabilized at a new and higher level, both in real and nominal terms, and this helped the BCB to reduce infl ation and meet the infl ation target set for that year. Then, in 2001, Brazil was hit by two major shocks that made the 6.0 percent infl ation target unfeasible. First, because of an unexpected drought, there was a shortage in the supply of electricity from hydro- electric plants during 2001 and, since most of Brazilian electrical power comes from such a source, the adverse energy shock pushed infl a-tion upwards. Second, and most important, the 2001 currency crisis in Argentina resulted in capital fl ight from and exchange rate depreciation in Brazil, pushing the domestic price of tradable goods up. Altogether

143

Tab

le 7

.1

Infl a

tion,

exc

hang

e ra

tes,

GD

P g

row

th a

nd in

tere

st ra

tes i

n B

razi

l

1998

%19

99 %

2000

%20

01 %

2002

%20

03 %

2004

%20

05 %

2006

%

Eff e

ctiv

e in

fl atio

n ra

te

End

of t

he p

erio

d 1

.78.

96.

07.

712

.59.

37.

65.

73.

1

Ave

rage

of t

he p

erio

d 3

.24.

97.

06.

88.

414

.76.

66.

94.

2

Tar

get i

nfl a

tion

rate

(end

of t

he p

erio

d)

Initi

al ta

rget

n.a.

8.00

6.00

4.00

3.50

4.00

5.50

4.50

4.50

C

eilin

gn.

a.10

.00

8.00

6.00

5.50

6.50

8.00

7.00

6.50

F

loor

n.a.

6.00

4.00

2.00

1.50

1.50

3.00

2.00

2.50

W

as th

e ta

rget

revi

sed?

n.a.

NO

NO

NO

NO

YE

SN

OY

ES

NO

R

evise

d ta

rget

8.5

5.1

W

as th

e ta

rget

met

?n.

a.Y

ES

YE

SN

ON

ON

OY

ES

YE

SY

ES

Nom

inal

exc

hang

e ra

te v

aria

tion

E

nd o

f the

per

iod

8.3

36.7

–2.2

16.0

35.1

–15.

3–4

.8–2

0.5

–1.3

A

vera

ge o

f the

per

iod

7.7

56.4

0.9

28.4

24.3

5.4

–4.9

–16.

8–1

0.6

GD

P g

row

th ra

te 0

.00.

34.

31.

32.

71.

15.

72.

93.

7

Bas

e in

tere

st ra

te

Nom

inal

bas

e in

tere

st ra

te28

.825

.617

.417

.319

.223

.316

.219

.115

.1

Rea

l bas

e in

tere

st ra

te26

.615

.310

.88.

96.

012

.88.

012

.711

.6

Not

e:

The

infl a

tion

targ

ets a

re th

e ta

rget

s set

at l

east

one

yea

r in

adva

nce.

The

exc

eptio

n is

1999

, whe

n th

e ta

rget

was

set i

n Ju

ne o

f tha

t yea

r. T

he

revi

sed

targ

et is

the

targ

et s

et in

the

corr

espo

ndin

g ye

ar. T

he e

xcha

nge

rate

var

iatio

n re

fers

to th

e B

razi

lian

real

/US

dolla

r ex

chan

ge r

ate,

and

the

base

inte

rest

rate

is th

e SE

LIC

inte

rest

rate

, set

by

the B

CB

, and

cum

ulat

ed o

ver t

he y

ear.

The

real

inte

rest

rate

was

obt

aine

d by

defl

atin

g th

e nom

inal

in

tere

st ra

te b

y th

e in

fl atio

n ra

te c

umul

ated

ove

r the

yea

r, ac

cord

ing

to th

e IP

CA

inde

x.

Sour

ce:

Bra

zilia

n C

entr

al B

ank

(htt

p://w

ww

.bcb

.gov

.br)

and

aut

hor’s

cal

cula

tion.


the fi nal result of these two adverse shocks was a 1.7 pp increase in the Brazilian infl ation rate in 2001, when compared to 2000.9

The macroeconomic turmoil of 2001 was followed by more instability in 2002, this time because of the Brazilian political cycle and the speculative international capital fl ows associated with it. More specifi cally, in 2002 Brazil had its presidential election and the looming victory of a leftist can-didate, Lula, resulted in massive capital outfl ows and an unprecedented cut in Brazil’s access to foreign lines of credit. The exchange rate shot up as a result of such a move, reaching its highest level in real terms since the debt crisis of the early 1980s. The Brazilian infl ation rate followed soon after and reached double-digit levels at the end of 2002. In fact, the exchange rate depreciation and infl ation acceleration were so sharp and fast that they even made the Brazilian monthly base interest rate tempor-arily negative at the end of 2002. Altogether the infl ation rate increased in another 4.8 pp in 2002, when compared to 2001.

Lula was elected president in 2002 and his fi rst term in offi ce started with a sharp monetary and fi scal crunch in order to reduce infl ation and restore the country’s access to foreign fi nance. Lula’s move and the very own excesses of the speculative attack to the Brazilian currency in 2002 quickly resulted in a return to normality, that is, a return to capital

–25

–20

–15

–10

–5

0

5

10

15

20

25

30

35

40

1998 1999 2000 2001* 2002* 2003* 2004 2005 2006

Per

cent

(%

)

Exchange rate variationInflation

Note: * Target not met. The infl ation rate is the variation in the IPCA index, December to December, and the exchange rate variation the rate of change in the Brazilian real/US dollar exchange rate, also December to December.

Source: Brazilian Central Bank (http://www.bcb.gov.br).

Figure 7.1 Consumer infl ation rate and exchange rate variation in Brazil


infl ows and exchange rate appreciation. Infl ation decelerated to one-digit levels in 2003, with a reduction of 3.2 pp in relation to 2002. Despite this disinfl ation, the infl ation target for 2003 was not met, even after the Lula administration revised it upwards because of the exchange rate deprecia-tion of 2002.10

After the failure in 2001–03, the Brazilian infl ation target regime started to perform well again in 2004–06. Most of this success can be credited to the exchange rate appreciation during these three years, which in its turn resulted from both the high domestic interest rates of Brazil and the favor-able international trade and fi nancial conditions in the rest of the world. More specifi cally, on the one hand, the BCB continued to practice high real interest rates in 2004–06 and this resulted in a gradual appreciation of the Brazilian exchange rate, in nominal and in real terms, as shown in Figures 7.2 and 7.3.11 On the other hand, the exchange rate apprecia-tion of 2004–06 was also driven by the boom in Brazilian exports, pulled by the rise in the world demand for commodities, which in their turn increased the country’s trade and current account surpluses. It should be noted that the improvement in the Brazilian balance of payments hap-pened despite the exchange rate appreciation for two reasons: fi rst, the increase in international prices off set the reduction in the exchange rate for many sectors and, second, the depreciation in 2001 and 2002 was so

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Jan/

98

Jul/9

8

Jan/

99

Jul/9

9

Jan/

00

Jul/0

0

Jan/

01

Jul/0

1

Jan/

02

Jul/0

2

Jan/

03

Jul/0

3

Jan/

04

Jul/0

4

Jan/

05

Jul/0

5

Jan/

06

Jul/0

6

Jan/

07

Jul/0

7

Source: IPEADATA (http://www.ipeadata.gov.br).

Figure 7.2 Nominal exchange rate in Brazil – Brazilian real per US dollar


intense that only at the end of 2005 did the Brazilian real exchange rate return to the level verifi ed in 1999.

Altogether the infl ation targeting performance of Brazil in 1999–2006 can be divided into three phases: (1) implementation, in 1999–2000, when the favorable international conditions and the modest infl ation targets proved to be very successful; (2) crisis, in 2001–03, when a combination of adverse supply shocks and fi nancial crises made the ambitiously low infl a-tion targets unfeasible; and (3) consolidation, in 2004–06, when the very favorable international conditions and the high real domestic interest rates resulted in a quick reduction in infl ation. Let us now turn to the evolution of the Brazilian base real interest rate throughout this process.

Figure 7.4 shows the real base interest rate of Brazil since 1994 in order to allow a comparison between the exchange rate targeting and the infl a-tion targeting regimes. The fi rst thing that draws one’s attention is the very high real interest rate of Brazil during its exchange rate targeting regime and the substantial reduction brought by infl ation targeting. In numbers, the average annual real base interest rate of Brazil was 21.9 percent in 1994–98, and 10.7 percent in 1999–2006.12 Despite such a substantial reduction in the real interest rate, it should also be noted that during the infl ation targeting regime the Brazilian rate continued to be extremely high

60

80

100

120

140

160

180

Jan/

98

Jul/9

8

Jan/

99

Jul/9

9

Jan/

00

Jul/0

0

Jan/

01

Jul/0

1

Jan/

02

Jul/0

2

Jan/

03

Jul/0

3

Jan/

04

Jul/0

4

Jan/

05

Jul/0

5

Jan/

06

Jul/0

6

Jan/

07

Jul/0

7

Source: Central Bank of Brazil (http://www.bcb.gov.br).

Figure 7.3 Index of the real eff ective exchange rate of Brazil (1992 5 100)


by international standards and, more important, the Brazilian real interest rate was extremely high when compared to the average GDP growth rate of the economy in 1999–2006, that is, 2.7 percent.

Focusing our analysis on 1999–2006, Figure 7.4 shows that infl ation targeting started with very high real interest rates in Brazil. As we men-tioned earlier, the reason for this was the Brazilian 1999 currency crisis, which led the BCB to increase interest rates abruptly and substantially to stop the capital outfl ows from the country. Then, after infl ation targeting was formally introduced, in mid 1999, the real base interest rate started to fall until it reached 9.5 percent at the end of 2000. After that the real inter-est rate fl uctuated between 8 percent and 10 percent until the end of 2002, when the devaluation of the real associated with Lula’s election and the subsequent increase in infl ation made BCB’s base interest rate temporarily negative. ‘Normality’ was quickly restored at the beginning of 2003, when the Lula administration raised the real interest rate of the economy to 14.3 percent in order to stop devaluation and decelerate infl ation. When it became clear that such an eff ort was successful, in mid 2003 the real base interest rate started to fall gradually. The interest rate cuts persisted until the end of 2004, when the acceleration of GDP growth made the BCB fear that the market’s expectation for infl ation would accelerate above the

0

5

10

15

20

Per

cent

(%

)

25

30

1994

-1

1995

-1

1996

-1

1997

-1

1998

-1

1999

-1

2000

-1

2001

-1

2002

-1

2003

-1

2004

-1

2005

-1

2006

-1

2007

-1

EXCHANGE RATETARGETING

INFLATIONTARGETING

Note: The real interest rate is the SELIC interest rate, set by the BCB, cumulated in the past 12 months, defl ated by the consumer infl ation rate also cumulated in the past 12 months, measured by the IPCA index.

Source: Central Bank of Brazil (http://www.bcb.gov.br) and author’s calculation.

Figure 7.4 Real annual base interest rate in Brazil


target set for 2005.13 The BCB’s response to such an expectational threat was another interest rate hike, which made the Brazilian real interest rate reach 13.4 percent in mid 2006.

The increase in the interest rate in 2005 resulted in a sharp appreciation of the real and a deceleration of GDP growth. By the end of 2005 it was clear that the BCB had exaggerated in its response to the increase in the level of economic activity at the end of 2004, as well as that the huge dis-crepancy between Brazilian and international interest rates could compro-mise the country’s fi scal stability. The BCB’s response was another round of gradual interest rate cuts, which made the Brazilian real interest fall to ‘just’ 11.6 percent at the end of 2006, that is, a still abnormally high real interest rate by international standards and well above the GDP growth rate of the economy in that year, that is, 3.7 percent.

Moving to fi scal policy, the main side eff ect of the Brazilian high real interest rates was the expansionary impact of net interest payments on public expenditures and debt. Table 7.2 presents the main fi scal numbers for the period, from which we can identify two distinct phases in the evolu-tion of the net public debt of Brazil. First, in the fi rst four years of infl ation targeting most of the government net debt was directly or indirectly tied to the exchange rate because the Brazilian government was a net debtor in foreign currency and most of its domestic debt was formally indexed to either the exchange rate or the interest rate.14 So, in the wake of the 1999 currency crisis and the interest rate hike that followed it, the net interest payments by the Brazilian public sector shot up to 12.5 percent of GDP, and the ratio of its net public debt to GDP increased in 5.6 pp in just one year. Then, as the exchange rate stabilized at a new and higher level, the net interest payments by the public sector fell to less than 10 percent of GDP in 2000–01, but the net domestic debt of the government continued to rise gradually in relation to GDP because of the huge diff erence between the real interest rate, on the one side, and the growth rate of the economy, on the other side. The fi rst phase of debt dynamics ended in 2002, when Brazil experienced another sharp and abrupt exchange rate depreciation, the net interest payments by the public sector shot up to 13 percent of GDP, and the net public debt of the country rose to 50.5 percent of its GDP.

The second phase in the dynamics of public debt began in 2003, when the exchange rate appreciation started to reduce the government net inter-est payments. The ratio of net public debt to GDP followed the same path in 2004–06, when the Brazilian government was able to repay most of its foreign debt and accumulate a huge amount of foreign reserves. This movement was obviously facilitated by the favorable international trade and fi nancial conditions of the period and, at the end of 2006, the Brazilian government became a net creditor in foreign currency. In contrast, on

149

Tab

le 7

.2

Fisc

al p

olic

y in

Bra

zil

1998

1999

2000

2001

2002

2003

2004

2005

2006

Pub

lic d

ebt i

n %

of G

DP

N

et d

omes

tic d

ebt

33.2

35.2

36.5

38.9

37.5

41.7

40.2

44.1

47.6

M

onet

ary

base

4.0

4.2

3.9

3.9

4.2

4.2

4.4

4.7

5.1

N

on-m

onet

ary

debt

29.2

31.0

32.7

34.9

33.3

37.5

35.8

39.4

42.5

N

et fo

reig

n de

bt5.

89.

49.

09.

613

.010

.76.

82.

3–2

.7

Net

dom

estic

and

fore

ign

debt

38.9

44.5

45.5

48.4

50.5

52.4

47.0

46.5

44.9

Bud

get b

alan

ce o

f the

pub

lic se

ctor

in %

of G

DP

Pr

imar

y de

fi cit

0.0

−3.

0−

3.3

−3.

4−

3.7

−3.

9−

4.2

−4.

4−

3.9

N

et in

tere

st p

aym

ents

7.4

12.5

7.4

8.2

13.0

7.2

6.5

7.1

6.8

On

fore

ign

debt

0.3

0.9

0.8

1.1

1.2

1.1

0.9

0.7

0.3

On

dom

estic

deb

t7.

111

.56.

67.

111

.86.

15.

66.

56.

5

Nom

inal

defi

cit

7.4

9.5

4.1

4.8

9.3

3.3

2.3

2.8

2.9

Pri

mar

y bu

dget

of t

he fe

dera

l

gove

rnm

ent i

n %

of G

DP

R

even

ues

15.8

16.4

16.5

17.2

17.9

17.4

18.1

18.8

19.4

E

xpen

ditu

res

15.1

14.5

14.7

15.6

15.7

15.1

15.6

16.4

17.2

N

et e

rror

and

om

issio

ns−

0.3

0.2

0.0

0.0

0.0

0.0

0.2

0.1

0.1

B

alan

ce0.

52.

11.

71.

72.

22.

32.

72.

62.

2

Not

e:

The

prim

ary

defi c

it is

the

gove

rnm

ent b

udge

t defi

cit

excl

udin

g ne

t int

eres

t pay

men

ts. T

he n

omin

al d

efi c

it is

the

tota

l bud

get d

efi c

it, th

at is

, th

e pr

imar

y de

fi cit

plus

net

inte

rest

pay

men

ts b

y th

e pu

blic

sect

or. T

he n

umbe

rs fo

r the

who

le B

razi

lian

publ

ic se

ctor

refe

r to

the

fede

ral,

stat

e an

d m

unic

ipal

gov

ernm

ents

, and

the

stat

e-ow

ned

ente

rpris

es.

Sour

ce:

Bra

zilia

n C

entr

al B

ank

(htt

p://w

ww

.bcb

.gov

.br)

and

aut

hor’s

cal

cula

tion.


the domestic front, the high real interest rate continued to increase the Brazilian net public domestic debt, which reached the unprecedented high level of 47.6 percent of GDP at the end of the period. Altogether, we can say that in this second phase of debt dynamics infl ation targeting was char-acterized by an increase in the ratio of net public domestic debt to GDP and a substitution of domestic debt for foreign debt.

To complete our fi scal outlook, Table 7.2 also shows the evolution of the federal government primary surplus and nominal budget defi cit in the infl ation targeting period.15 Compared to the last year of exchange rate targeting, most of the infl ation targeting period was characterized by a gradual increase in the federal primary surplus.16 Most of this increase happened in 1999, when the Brazilian federal government cut its spend-ing and increased its revenues in proportion to GDP in order to avoid an explosive increase in its public debt. After that the federal government’s revenues and expenditures tended to grow together in relation to GDP, which in its turn had a small positive impact on economic growth through the balanced-budget multiplier. The second main fi scal crunch happened in 2003, when the federal government once more cut its spending in pro-portion to GDP in order to avoid an explosive increase in the country’s net public debt. However, diff erently from what happened in 1999, the federal government revenues also fell in proportion to GDP in 2003, which ended up stabilizing the primary surplus in terms of GDP.17 The federal government revenues and expenditures resumed growing faster than GDP in 2004–06, and fi scal policy became expansionary at the end of the period.18

Finally, considering GDP growth, the infl ation targeting period had a slower rate of expansion than the exchange rate period, but a smaller vola-tility as well. To illustrate this point, Figure 7.5 shows the moving-average annual GDP growth rate of Brazil in 1994–2006. In numbers, on the one hand, the average GDP growth rate was 3.8 percent during the exchange rate targeting period, against 2.7 percent during the infl ation targeting period. On the other hand, the maximum and minimum growth rates during exchange rate targeting were 8.5 percent and zero, respectively, whereas during infl ation targeting the maximum and minimum rates were 5.7 percent and minus 0.8 percent, respectively. Finally, it should also be noted that the trend growth rate was stationary or declining during the exchange rate targeting period, but rising during the infl ation targeting period.

Figure 7.5 also reveals clearly the sequence of booms and busts expe-rienced by the Brazilian economy since 1994. The downturns are usually associated with adverse shocks, and the upturns with expansionary macro economic policy. More specifi cally, the period starts with the boom


brought by infl ation reduction in 1994. This boom ended in 1995, when the Brazilian economy suff ered the contagion eff ect from the Mexican crisis and the Brazilian government adopted restrictive macroeconomic measures to avoid exchange rate depreciation and infl ation acceleration. The second boom started after the Mexican crisis was absorbed by Brazil and lasted until 1997. Then, in 1997–99, the Brazilian economy experi-enced another growth deceleration because of the contagion eff ects from the East Asian crises of 1997 and the Russian crisis of 1998, and because of the recessive impact of their very own Brazilian crisis of 1999. The economy resumed growth in 2000, but the expansion was quickly cur-tailed by the Argentine crisis and the Brazilian energy rationing of 2001. The Brazilian economy started another recovery in 2002, but the expan-sion was short-lived because of the speculative attack to the Brazilian currency during the presidential elections of that year. The exchange rate depreciation and the restrictive macroeconomic measures adopted by the Brazilian government pushed economic growth down in 2003, and only in the beginning of 2004 did the economy start to recover. However, diff erently from all of the previous episodes, the 2004 expansion was cut for domestic reasons, namely the fear of the BCB that the economy was overheating at the end of that year. As we saw earlier, interest rates were raised and economic growth decelerated in approximately two

–2

0

2

4

6

8

10

1994

-1

1995

-1

1996

-1

1997

-1

1998

-1

1999

-1

2000

-1

2001

-1

2002

-1

2003

-1

2004

-1

2005

-1

2006

-1

2007

-1

INFLATION TARGETINGEXCHANGE RATETARGETING

Per

cent

(%

)

Note: The growth rate is the rate of change of the average GDP during period t through t-3, in relation to the average GDP during period t-4 through t-7.

Source: IBGE (http://www.ibge.gov.br) and author’s calculation.

Figure 7.5 Moving-average of the annual GDP growth rate of Brazil


percentage points in 2005, but it still remained well above the previous troughs. Then, at the end of the period under analysis, in the second half of 2006, the growth rate of the Brazilian economy started once again to accelerate.19

7.3 INFLATION TARGETING, REAL EXCHANGE RATES AND POTENTIAL OUTPUT

Whether or not Brazil started a new growth cycle in recent years is an open question that depends, among other things, on the management of monetary policy. From the recent history of infl ation targeting in Brazil, there are two main macroeconomic challenges to the Brazilian authori-ties that will eventually have to be addressed. First, so far the success of infl ation targeting depended heavily on a favorable behavior of exchange rate and this cannot go on indefi nitely. By defi nition real exchange rate appreciation cannot go on forever, otherwise the country’s currency will become infi nitely expensive and, more important, an excessively appreci-ated real exchange rate can reduce the growth prospects of the economy. Second, so far infl ation targeting has been characterized by a slow GDP growth, both in comparison to Brazil’s previous history and to the growth rate of the rest of the world. Since also by defi nition any growth acceler-ation results in a temporary increase in the level of economic activity, it tends to create infl ationary pressures and requires compensatory meas-ures by monetary policy. However, if monetary policy is too conservative in setting a low infl ation target and a fast speed of convergence to it, the economy may end up locked in a slow-growth equilibrium where the BCB ‘kills’ any growth acceleration for fears of rising infl ation. To complete our analysis of infl ation targeting in Brazil, let us look at these two chal-lenges in more detail.

First, regarding exchange rate, stable infl ation requires a stable real exchange rate, but the level of the exchange rate is not determined a priori. There are multiple equilibria and, therefore, the real exchange rate can be stabilized at a competitive or an uncompetitive level. Thus, when the infl ation target is ambitiously low and the speed of convergence to it too fast, there will be a tendency to real exchange rate appreciation, which in its turn may end up increasing the fi nancial fragility of the economy in the medium run, especially if appreciation is sustained by speculative capital infl ows. The usual logical sequence here is that high real interest rate leads to capital infl ows and exchange rate appreciation, which in their turn leads to a reduction in the trade and current account balance of the economy, which in their turn make the economy more vulnerable to shifts in the


international liquidity. So, when infl ation targeting is too ambitious and fi nanced by speculative capital infl ows, the result tends to be a sequence of booms and busts, during which GDP growth fl uctuates around a low rate. Another possibility is that infl ation targeting stabilizes the real exchange rate at a new low level without compromising the balance of payments of the economy.20 However, even in this case ambitious infl ation targeting can still result in a slow growth rate because an appreciated real exchange rate does not stimulate the development of the domestic tradable sector. In other words, productivity growth is usually faster in the tradable sector and, therefore, an appreciated real exchange rate may end up reducing the overall growth rate of the economy.21

The natural solution to the exchange rate challenge outlined above is to combine infl ation targeting with an asymmetric dirty fl oating that main-tains the country’s real exchange rate stable and competitive in the long run. The dirty fl oating will have to be asymmetric because the instruments to combat appreciation are diff erent than the ones use to combat depreci-ation in order to keep the real exchange rate competitive. In other words, in face of depreciation the government should use its traditional restrictive macroeconomic tools to stop the process and avoid an increase in infl a-tion, without selling much of its foreign reserves. However, in the face of appreciation the government should buy the foreign currency to slow down the process, but with no open compromise with a specifi c exchange rate. The auxiliary measures to combat appreciation are obviously to increase imports and reduce domestic interest rates so that, at the end of the day, the asymmetric dirty fl oating ends up creating a sliding fl oor for the exchange rate.22

Second, regarding growth, infl ation targeting is usually aimed at stabil-izing economic growth at its potential level, so that it eliminates excess demand pressures that tend to increase infl ation, and insuffi cient demand pressures that do the opposite. The logical form of the orthodox stand-point is that the long-run growth rate of the economy is given from outside of macroeconomic issues, by preferences and institutions, and the only thing a responsible central bank can do is to make the economy grow at its potential rate in the long run. So, according to the orthodox view of eco-nomic policy, attempts to grow faster than such supply-determined limits end up increasing infl ation, with either no or negative permanent real eff ects on the economy. The main problem with such a view is that it fails to recognize that the potential output of an economy is an endogenous variable and, therefore, it can and usually is aff ected by macroeconomic policy itself, including monetary policy.

More specifi cally, potential output is an unobservable variable that is estimated from the past and expected behavior of the economy.23 Since


expectations are usually heavily infl uenced by the recent past, the estimates of potential output tend to extrapolate current trends to the near future and, in this way, they may create a self-fulfi lling monetary policy. An example helps to illustrate the point. Suppose that a conservative central bank estimates a slow potential growth rate for the economy and, based on such a pessimistic estimate, it combats a growth acceleration for fearing that it will increase infl ation above the pre-specifi ed target. In doing so the very own actions of the central bank reduce the eff ective growth rate of the economy, especially of investment, and when potential output is re-estimated again ex post, the data will confi rm that the central bank was right not necessarily because its initial estimates were right, but because the growth deceleration produced by the central bank is automatically translated in a lower estimate of the growth potential of the economy.24 In short, since the estimates of potential output are highly uncertain at the end of the sample and continuously revised every time a new observation is incorporated into the analysis, the actions of the central bank may seem to be correct ex post because its very own actions produced the scenario that justifi ed it ex ante.

The endogeneity of potential output at the end of the sample is espec-ially dangerous for an economy, as Brazil, that aims to accelerate its growth rate. The danger here is that because the growth rate in the recent past was low, any strong growth acceleration tends to generate an appar-ent excessive output gap in the short run. So, if the central bank stops the process too soon, there won’t be enough time for the estimates of poten-tial output to pick the structural change in the growth prospects of the economy. The fi nal result is either that the economy never takes off , or that it takes off very slowly. On the other hand, the endogeneity of potential output does not obviously mean that authorities live in a sort of ‘fi eld of dreams’, in which, if you believe, growth will come.25 There are objective limits to the growth rate of the economy that cannot be avoided by opti-mistic expectations like, for instance, the maximum supply of energy, the stock of foreign reserves and a zero unemployment rate.

The endogeneity of potential output only means that there may be more than one equilibrium position for the GDP growth rate and, therefore, the speed of convergence to the infl ation target is also a relevant variable for economic growth. As usually happens in economics, central banks face a trade-off in this area: on the one hand, quick disinfl ation may lock the economy in a slow-growth equilibrium, but, on the other hand, slow disinfl ation may lock the economy in a high-infl ation equilibrium. Even though this trade-off has yet to be emphasized by modern macroeconomic theory, it is an unavoidable matter for policy makers who deal with the real world.


7.4 CONCLUSION

Infl ation targeting represents an improvement in relation to the previ-ous monetary policy regime adopted by Brazil, but it can and should be improved in order to increase the growth rate of the economy. In general terms the main conclusions from the previous sections can be summarized in ten points. (1) Infl ation targeting managed to reduce infl ation in Brazil after its 1999 and 2002 currency crises, with a substantial help of exchange rate appreciation. (2) Economic growth was slower under infl ation target-ing than under exchange rate targeting, but with a smaller volatility and with an apparent upward trend. (3) Infl ation targeting reduced the real interest rate of the economy, which nevertheless remained well above international standards and more than three times higher than the GDP growth rate of Brazil in 1999–2006. (4) The high real interest rate required a substantial increase in fi scal austerity by the Brazilian government, but so far this has not been suffi cient to stop the increase in the ratio of net public domestic debt to GDP. (5) The high domestic real interest rates and the favorable international trade and fi nancial conditions in the rest of the world allowed the Brazilian government to accumulate foreign reserves, repay most of its foreign debt and reduce its dependence of foreign capital in 2003–06. (6) Infl ation targeting can be combined with an asymmetric dirty fl oating regime, in which the central bank combats exchange rate depreciation with restrictive monetary policy, and accumulates foreign reserves to slow down appreciation. (7) The asymmetric dirty fl oating should aim at a stable competitive real exchange rate, in order to promote the fast growth of the domestic tradable sector, and in this way push the country’s exports and imports up without increasing the foreign fi nancial fragility of the economy. (8) Aggregate estimates of potential output are not good guides for the demand pressures on infl ation in moments of structural growth acceleration or deceleration. (9) Because of the endo-geneity of potential output at the end of the sample, infl ation targeting should be done with moderation in order to avoid a self-fulfi lling mon-etary policy that locks the economy in a slow-growth equilibrium. (10) A fast convergence to a low infl ation target may produce a permanently low growth rate, but a slow convergence to a moderate infl ation target can produce a permanently high infl ation.

NOTES

1. In this chapter we defi ne the exchange rate as the domestic price of foreign currency, so that a depreciation of the exchange rate means an increase in the exchange rate, that is,


a devaluation of the domestic currency. By analogy, appreciation means a reduction in the exchange rate, that is, a revaluation of the domestic currency.

2. For instance, according to the estimates for that time (Belaisch, 2003), 23 percent of exchange rate variations tended to pass through to consumer prices (measured by the IPCA index) in the long run, whereas the pass-through to the general price level (meas-ured by the IGPDI index) was 71 percent.

3. For a more technical analysis of infl ation targeting in Brazil, see Bogdanski et al. (2000), Minella et al. (2003), Tombini and Alves (2006) and Bevilaqua et al. (2007).

4. According to the Brazilian Presidential Decree 3088, of 21 June 1999, it is up to the BCB ‘to execute the necessary policies to meet the specifi ed targets’.

5. In 1999–2005 the Copom met every month to determine interest rates. Starting in 2006 the Copom has been meeting eight times a year, that is, one meeting approximately every six weeks. If necessary, an extraordinary meeting can be called by the president of the BCB.

6. According to the IMF estimates, the average world GDP growth rate was 3.7 percent in 1994–98 and 4.2 percent in 1999–06. In its turn, the infl ation targeting period can be divided in two phases regarding the world GDP growth rate, that is, slow-growth, in 1999–06, when the average world annual GDP growth rate was 3.5 percent, and fast growth, in 2003–06, when the average annual growth rate was 4.9 percent.

7. Unless stated otherwise, all numbers are annual fi gures. 8. The infl ation rate of 1999 was high when compared to 1998, but modest when com-

pared to the magnitude of the exchange rate depreciation verifi ed in 1999. As we will see later in this section, most of the small pass-through of the exchange rate to domestic prices in 1999 can be credited to the high real interest rate and the abrupt fi scal crunch practiced by the Brazilian government during that year.

9. As we will also see later in this section, the high real interest rate practiced by the BCB and the very own recessive impact of the two supply shocks on the level of economic activity helped to reduce the magnitude of infl ation acceleration.

10. One of the fi rst measures of the Lula administration was to increase interest rates, cut public spending and increase the infl ation target for 2003, from the 4.0 percent set by the previous administration to 8.5 percent because of the adverse ‘electoral’ shock of 2002. A complete analysis of Lula’s economic policy is beyond the scope of this chapter. The inter-ested reader can fi nd more information in Barbosa-Filho (2007) and Arestis et al. (2007).

11. Note that most of the real exchange rate appreciation was concentrated in the fi rst semester of 2003 and in 2005, which coincided with the periods of increases in the Brazilian base interest rate, as we will see later.

12. These are geometric averages and the real interest rate was calculated ex post.13. At the time there was an intense debate in the media on whether or not the economy

was really overheating, since the increase in infl ation at the end of 2004 also refl ected an increase in an important indirect tax rate, the so-known PIS-COFINS rate.

14. Indexation to the interest rate increased the impact of exchange rate depreciation on domestic debt because the BCB usually increased its base rate abruptly after a devalu-ation of the domestic currency.

15. The nominal defi cit is the diff erence between the total government expenditures and revenues, that is, it is the net interest payments minus the primary surplus.

16. Only in 2006 the primary surplus fell in relation to GDP, but it was still higher than the level verifi ed in 2002.

17. Nevertheless, the combined reduction in public revenues and spending had a negative impact on economic growth in 2003 through the balanced-budget multiplier.

18. In 2003–06 all of the increase in government expenditures in percentage of GDP was directed to an increase in income transfers through social security, social assistance and unemployment benefi ts (Barbosa-Filho, 2007).

19. The expansion, started in 2006, gained force in 2007, when the Brazilian GDP grew 5.4 percent.

20. For a formal model of this case, see Barbosa-Filho (2006).


21. For an analysis of the link between real exchange rates and development, see Frenkel and Taylor (2006).

22. This idea has been fi rst proposed by Ros (1995) for Mexico. For a more recent analysis of infl ation targeting in Mexico, see Galindo and Ros (2006).

23. For an analysis of the methods of estimation of potential output applied to Brazil, see Barbosa-Filho (2005).

24. So far most of the mainstream studies on the uncertainty associated with estimates of potential output have been focused on the reliability of forecasts based on real-time data, rather than on the implications of a self-fulfi lling monetary policy. For an example of the mainstream approach, see Orphanides and van Norden (2005).

25. Field of Dreams is a 1989 movie in which a farmer becomes convinced by a mysterious voice that, if he constructs a baseball diamond in his corn fi eld, the ghosts of deceased star baseball players will come to play. The voice repeatedly says to the farmer: ‘If you build, he will come’, and the Hollywood story obviously ends up confi rming the voice’s prediction.

REFERENCES

Arestis, P., L.F. de Paula and F. Ferrari (2007), ‘Assessing the economic policies of President Lula da Silva in Brazil: has fear defeated hope?’, Oxford Centre for Brazilian Studies working paper CBS-81-07.

Barbosa-Filho, N.H. (2005), ‘Estimating potential output: a survey of the alterna-tive methods and their applications to Brazil’, IPEA working paper no. 1092.

Barbosa-Filho, N.H. (2006), ‘Exchange rate, growth and infl ation’, paper presented at the Annual Conference on Development and Change, Campos do Jordão, Brazil.

Barbosa-Filho, N.H. (2007), ‘An unusual economic arrangement: the Brazilian economy during the fi rst Lula administration, 2003–2006’, International Journal of Politics, Culture and Society, 19, 193–215.

Belaisch, A. (2003), ‘Exchange rate pass-through in Brazil’, International Monetary Fund working paper no. 03-141.

Bevilaqua, A.S., M. Mesquita and A. Minella (2007), ‘Brazil: taming infl ation expectations’, Central Bank of Brazil working paper no. 129.

Bogdanski, J., A.A. Tombini and S.R.C. Werlang (2000), ‘Implementing infl ation targeting in Brazil’, Central Bank of Brazil working paper no. 1.

Frenkel, R. and L. Taylor (2006), ‘Real exchange rate, monetary policy and employment’, United Nations, Department of Economic and Social Studies working paper no. 19.

Galindo, L.M. and J. Ros (2006), ‘Alternatives to infl ation targeting in Mexico’, University of Massachusetts Political Economy Research Institute, alternatives to infl ation targeting: central bank policy for employment creation, poverty reduction and sustainable growth working paper no. 7.

Minella, A., P.S. de Freitas, I. Goldfajn and M.K. Muinhos (2003), ‘Infl ation tar-geting in Brazil: constructing credibility under exchange rate volatility’, Central Bank of Brazil working paper no. 77.

Orphanides, A. and S. van Norden (2005), ‘The reliability of infl ation forecasts based on output gap estimates in real time’, Journal of Money, Credit and Banking, 37 (3), 583–601.

Ros, J. (1995), ‘La crisis mexicana: causas, perspectivas, lecciones’, Nexos, (May), 46.Tombini, A.A. and S.A.L. Alves (2006), ‘The recent Brazilian disinfl ation process

and costs’, Central Bank of Brazil working paper no. 109.

158

8. Alternatives to infl ation targeting in MexicoLuis Miguel Galindo and Jaime Ros1

8.1 INTRODUCTION

Infl ation targeting has become increasingly popular over the past decade. As a nominal anchor for monetary policy with a public and explicit com-mitment to maintain economic discipline, infl ation targeting is being pro-moted as a general framework in order to reduce and control the infl ation rate, improve infl ation predictability, accountability and transparency, reduce the expected infl ation variability (Sheridan, 2001), improve the output-infl ation trade-off (Clifton et al., 2001), help solve the dynamic consistency problem and even reduce output variability (Svensson, 1997, 1998). Several Latin American countries such as Chile (1990), Peru (1994), Mexico (1999), Brazil (1999) and Colombia (1999) have moved their monetary regimes to an infl ation targeting framework (Corbo et al., 2002; Schmidt-Hebbel and Werner, 2002).

There are also a number of concerns regarding the adoption of an infl a-tion targeting regime. There is an important concern about the ability of the central bank to control the infl ation rate, in particular under the presence of fi scal or external shocks. There are also fundamental doubts about the economic consequences of an infl ation targeting regime. For example, Ball and Sheridan (2003) argue that the reduction of infl ation and the reduction in output variability in the OECD countries is a general trend that is not necessarily related to the instrumentation of infl ation targeting and that this kind of regime does not aff ect output variability. Furthermore, Newman and von Hagen (2002) claim that the evaluation of infl ation targeting has been misleading because it has not properly con-sidered the problem of regression to the mean given that some infl ation targeting countries were worse off before their implementation than the countries without an infl ation targeting framework.

In the case of emerging market economies there is a special concern about the relationship of infl ation targeting with exchange rate move-ments, imperfect and poorly regulated fi nancial markets and weak

Alternatives to infl ation targeting in Mexico 159

monetary institutions. In eff ect, there are important concerns about the eff ectiveness of an infl ation targeting regime under weak fi scal conditions, poorly regulated fi nancial systems, large potential external shocks, low institutional credibility and currency substitution phenomena (Fraga et. al., 2003; Calvo and Mishkin, 2003). There is also a major concern that important shifts in exchange rates will aff ect the general competitiveness of the economy, the current account defi cit and generate external shocks on the infl ation rate. For example, abrupt changes in the exchange rate might lead to infl ation paths that are inconsistent with the original infl a-tion target and therefore the central bank might try to use the nominal exchange rate as a nominal anchor (Svensson, 1998). In this case exchange rate shocks to the infl ation rate are followed by a general appreciation of the real exchange rate aff ecting the general performance of the economy. Goldfajn and Gupta (2003) argue also that an appreciation of the real exchange rate is related with higher nominal interest rates or a tight mon-etary policy that makes economic recovery more diffi cult.

This chapter addresses Mexico’s experience with infl ation targeting. Section 8.2 presents some background on the operation of monetary policy in Mexico since the 1994–95 peso crisis. Section 8.3 assesses empiri-cally a number of important issues in the evaluation of infl ation targeting: the eff ects of the real exchange rate on output, the question of the pass-through of exchange rate movements into infl ation and the asymmetric response of monetary policy in the face of exchange rate shocks. Section 8.4 considers alternatives to infl ation targeting as currently implemented. Section 8.5 concludes.

8.2 INFLATION TARGETING IN MEXICO: SOME BACKGROUND

In the 1990s Mexico experienced a variety of monetary and exchange rate regimes. More precisely, the monetary regime went through three stages: nominal exchange rate targeting with a crawling band regime before the 1994 crisis, monetary targeting for a short period after the crisis, followed by a transition to infl ation targeting that by now has been largely com-pleted. With the 1994 crisis the exchange rate regime shifted from a crawl-ing band to fl oating.

The experience with the crawling band regime and its crisis has been analysed widely (see, for example, Lustig and Ros, 1998; Ros, 2001). It was during this period, in 1993, that the central bank was given indepen-dence and a mandate to preserve price stability. After the 1994–95 peso crisis, monetary policy focused on the growth of monetary aggregates and


limits to credit growth as a means to control the infl ationary eff ects of the sharp peso devaluations and rebuild the damaged credibility of the Banco de México. More precisely, the central bank established as its nominal anchor an intermediate target for the growth of the monetary base. As the country moved to a fl exible exchange rate regime, an assumption of no international reserves accumulation was made. The ceiling on the growth of the monetary base was thus in essence a growth ceiling on net dom-estic credit. This monetary policy framework was soon abandoned as the policy failed to stabilize infl ationary expectations, the exchange rate and infl ation itself. The main reasons for this failure were the instability of the relationship between the monetary base and infl ation and the fact that the central bank has no control of the monetary base in the short run. The demand for bills and coins in circulation largely determines the monetary base and this demand is very interest inelastic in the short term (Carstens and Werner, 1999).

The peso crisis had strong infl ationary eff ects, taking the infl ation rate from single digits to over 50 percent. These infl ationary eff ects were largely brought under control by means of a tight fi scal policy and an incomes policy negotiated with unions and business confederations. After this, the main objective of monetary policy became the reduction of infl ation in a gradual and sustainable way. At the same time, the monetary policy regime moved towards infl uencing the level of interest rates, establishing bor-rowed reserves as its key instrument with the monetary base becoming less relevant and the infl ation target more important in the conduct of policy.

Under this framework the Banco de México establishes that banks should seek a null balance in their accounts at the central bank. A penalty rate (double the government treasury bill rate, the CETES rate) is applied to overdrafts and positive balances are not remunerated. Thus, the banks seek this balance in particular because if the daily average balance were negative they would have to pay the penalty rate. By providing more or less liquidity through daily monetary auctions, the overall net daily average balance of all current accounts held by banks at the Banco de México may close the day at a predetermined amount. If negative, the central bank puts the banking system in ‘short’ (‘corto’) in which case at least one credit institution has to pay the penalty interest rate. This is the way in which the central bank exerts pressure on interest rates. Although the corto represents only a small amount of the total liquidity (normally between 0.09 and 0.068 of the monetarty base), increasing it induces banks to bid up interest rates to avoid paying overdraft charges and puts upward pressure on market interest rates. The reverse mechanism applies when the central bank targets a positive balance (‘largo’) (Banco de México, 1996, Appendix 4; Carstens and Werner, 1999, pp. 15–16; OECD, 2002, 2004).


The transition to infl ation targeting was accelerated in January 1999 when the Banco de México announced a medium-term infl ation objective, and since 2000 the central bank publishes quarterly infl ation reports to monitor the infl ationary process, analyse infl ation prospects and discuss the conduct of monetary policy. By now Mexico is considered to have in place the main components of an infl ation targeting framework: an independent monetary authority (since 1993) that has infl ation as its only policy objective, a fl exible exchange rate regime, the absence of other nominal anchors and a ‘transparent’ framework for the implementation of monetary policy (Schmidt-Hebbel and Werner, 2002).

The implementation involves the choice of an infl ation index to be adopted as target, the target range and the time horizon. The national consumer price index (CPI) is used to determine the infl ation objec-tive. Since the beginning of 2000 the bank also tracks a core CPI which excludes volatile items – such as agricultural and livestock products, and also education (tuition fees) – and prices controlled by or agreed with the public sector. In the transition from high to moderate infl ation rates, the infl ation objective was specifi ed in terms of a value that should not be exceeded. Today policy specifi es a target range of plus or minus a percentage point. From January 1999 onwards the target of monetary policy has been framed in terms of a medium-term infl ation objective of bringing down infl ation to the rate prevailing in Mexico’s main trading partners by the end of 2003, interpreted as an annual growth of the CPI of 3 percent.

Table 8.1 shows the infl ation targets since the 1994–95 crisis, actual infl ation performance, the growth projections and performance as well as the evolution of the real exchange rate. After a rough start, when in 1995 the infl ation objective was missed by over 30 percentage points, the central bank’s record has been improving over time with its infl ation target being met in fi ve years since 1999 and infl ation tending to converge towards its medium-term target range of 3 percent plus or minus one percentage point.

Along with the success in bringing down infl ation, the growth per-formance has been disappointing. More precisely, in the second half of the 1990s, under the stimulus of a very competitive exchange rate, the economy rapidly recovered from the 1995 recession and grew at rates that often surpassed the growth rate projected by the central bank and the government. However, from 2001 to 2003, growth sharply decelerated to rates that for three years in a row were below or barely above the rate of population growth. That is, the economy recorded a decline in per capita incomes from 2000 to 2003 before recovering in 2004. Overall perform-ance since 1994 to 2004 has been unsatisfactory with GDP growth below 3

162

Tab

le 8

.1

Infl a

tion,

gro

wth

and

the

real

exc

hang

e ra

te

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Infl a

tion

targ

et (%

)19

1015

1213

106.

54.

53

± 1

3 ±

13

± 1

3 ±

1A

ctua

l infl

atio

n (%

)52

.027

.715

.718

.612

.39.

04.

45.

74.

05.

23.

34.

1G

DP

grow

th p

roje

ctio

nn.

a..

3.

45

34.

5n.

a.1.

53.

03.

0–3.

53.

83.

6A

ctua

l GD

P gr

owth

–6.2

5.1

6.8

4.9

3.7

6.6

–0.1

0.7

1.3

4.2

3.0

4.5

Rea

l exc

hang

e ra

te

(e

nd o

f the

yea

r) (i

ndex

)14

1.4

127.

811

4.1

116.

110

5.8

100.

092

.510

2.3

106.

210

3.2

98.5

98.7

Sour

ce:

Ban

co d

e M

éxic

o, w

ww

.ban

xico

.org

.mx/

tipo/

esta

dist

icas

/inde

x.ht

ml.

The

real

exc

hang

e ra

te is

defi

ned

as t

he n

omin

al e

xcha

nge

rate

to

the

US

dolla

r, m

ultip

lied

by th

e ra

tio o

f for

eign

(CPI

of U

SA) t

o do

mes

tic p

rice

leve

ls (C

PI o

f Mex

ico)

.


percent per year, well below the historical rates of the period 1940 to 1980 (6 to 6.5 percent).

8.3 THE EMPIRICAL EVIDENCE

In this section we address three relevant questions for the evaluation of the infl ation targeting regime. First, what are the short-term and long-run eff ects of the real exchange rate on output? Have these eff ects changed with trade liberalization and integration with the US economy? Second, how important is the pass-through of the exchange rate to prices? Has infl ation targeting modifi ed the pass-through? Third, does infl ation tar-geting have a bias towards exchange rate appreciation? If so, and the real exchange rate has long-run eff ects on output, the regime is not neutral with respect to economic growth.

However, infl ation targeting monetary policies are relatively recent making it diffi cult to use several econometric techniques. In this sense, these results must be considered only preliminary. The database consists of quarterly data for the period 1980 to 2003 and 1999 to 2003. The selec-tion period was due to data limitations. A defi nition of the variables is included in the Appendix 8.1A.

8.3.1 The Real Exchange Rate and Output

The net eff ect of the real exchange rate on output is not clear-cut because there are alternative transmission channels with opposite eff ects in the short and long runs.2 In order to assess the long-run impact of the real exchange rate on output we proceed to estimate a vector autoregressive model (VAR) model including output (Yt), investment (INVt), US output (YUSt) and the real exchange rate (SRt).3 The unit roots tests of these vari-ables are summarized in Table 8A.1 in Appendix 8.1A. These tests indicate that these variables are all I (1) and therefore it is necessary to consider the option of possible cointegration4 among them. The statistics of the Johansen (1988) procedure5 (Table 8.2) indicates the presence of at least one cointegrating vector (lower case letters refer to the natural logarithm of the series) already correcting for certain instability in the cointegrating space (Johansen et al., 2000).

Normalizing this vector as an output equation (Equation 8.1) indicates the presence of a positive relationship between output, investment, US output and the real exchange rate. Therefore, a devaluation of the real exchange rate has a positive impact on Mexican economic growth in the long run.


yt 5 0.478*invt 1 0.320*yust 1 0.019*srt (8.1)

The series included in the estimated VAR are all considered initially as endogenous variables. Under these circumstances, the causal relations among these variables are not clear while the normalization of the cointe-grating vector already implies a particular classifi cation between endog-enous and exogenous variables. Thus, it is important to take the results with some caution due to the potential endogeneity problem (Lütkepohl and Reimers, 1992). Therefore, we perform an impulse response exercise in the VAR including, in the following order: yust, yt, invt and srt.

The impulse response function incorporates information on contem-poraneous eff ects since it ignores the contemporaneous correlations of residuals. In order to avoid this pitfall it is common to orthogonalize inno-vations. The orthogonalized innovation impulse response function should be interpreted with caution since orthogonalization imposes a pre-order which implies a semi-structural interpretation of the model. The variable which enters fi rst in the system acts as the most exogenous. The move-ments in these variables at one point in time precede the movements of the rest of the variables which come afterwards in the system. The impulse response analysis, using the Cholesky decomposition and the new order of the variables in the system, confi rms the positive eff ect of investment and US output on Mexico’s economic growth and indicates that the real exchange rate has an initial negative eff ect that tends to disappear in the long run (Figure 8.1).

The fi nding that an appreciation of the real exchange rate has long-run contractionary eff ects on economic growth in Mexico contradicts

Table 8.2 Statistics of the Johansen procedure including output, investment, US output and the real exchange rate, period 1981:01–2003:04

yt 5 b1*invt 1 b2*yust 1 b3*srt**

H0 Constant Trend Trace 95%

r 5 0 m0 0 63.09* 47.21r ≤ 1 m0 0 20.57 29.68r ≤ 2 m0 0 6.10 15.41r ≤ 3 m0 0 0.01 3.76

Notes: * Signifi cant at the 5 percent level; Trace 5 Trace test; r 5 number of cointegrating vectors. Number of lags in the VAR 5 4; VAR includes constant unrestricted.** y: output; inv: investment; yus: US output; sr: real exchange rate.


previous results of impulse response analysis that the long-run relation-ship between output and the real exchange rate is negative (Kamin and Rogers, 1997). It is also worth noting that the eff ect of the real exchange rate on output has signifi cantly increased after the beginning of the North American Free Trade Agreement (Table 8.3). Therefore, the Mexican economy seems to be more sensitive to exchange rate changes today than in the past.

8.3.2 The Nominal Exchange Rate and Infl ation (Pass-through)

The positive impact of the nominal exchange rate on the infl ation rate (pass-through) is one of the main concerns of the monetary authorities. Under these conditions, an infl ation targeting regime is prone to suff er from external dominance. That is, the high sensitivity of infl ation to any

–0.02

–0.01

0.00

0.01

0.02

0.03

1 2 3 4 5 6 7 8 9 10

Response of output (y) toUS output (yus)

–0.02

–0.01

0.00

0.01

0.02

0.03

1 2 3 4 5 6 7 8 9 10

Response of output (y) tooutput (y)

–0.02

–0.01

0.00

0.01

0.02

0.03

1 2 3 4 5 6 7 8 9 10

Response of output (y) toinvestment (inv)

–0.02

–0.01

0.00

0.01

0.02

0.03

1 2 3 4 5 6 7 8 9 10

Response of output (y) toreal exchange rate (sr)

Note: The horizontal axis refers to the ten periods or quarters while the vertical axis shows the response of the infl ation rate to a one standard deviation shock in each variable in the system.

Figure 8.1 Impulse – response analysis for yust, yt, invt and srt


exchange rate depreciation could cause external infl ation shocks that will make it diffi cult for the central bank to achieve its infl ation target. It has also been argued that an infl ation targeting regime reduces the pass-through problem due to its ability to increase the credibility of the monetary authorities and therefore the importance of forward-looking variables (Fraga, et al., 2003; Schmidt-Hebbel and Werner, 2002).

We evaluate the pass-through eff ect in Mexico. First, we estimate a VAR including the infl ation rate, the output gap and changes in the nominal exchange rate. The sample period covers quarterly data from 1986:01 to 2003:04. This specifi cation is relatively similar to a tradit-ional Phillips curve and includes some elements of the VAR specifi ca-tion used in Schmidt-Hebbel and Werner (2002) or Fraga et al. (2003). That is, the output gap should have a positive impact on the infl ation rate due to cost pressures through the labor market and input costs while a devaluation increases import costs and tradable goods prices and has a positive impact on the infl ation rate. Additionally, it is pos-sible to argue that this VAR includes variables that are all I (0) (Table 8A.1 in Appendix 8.1A). Therefore, this VAR, including the infl ation rate, the output gap and changes in the nominal exchange rate, was used to generate the impulse response (Figure 8.2) and indicates that the output gap and the change in the exchange rate have both a positive impact on the infl ation rate also refl ecting the relevance of the pass-through eff ect.

Second, in order to evaluate the relevance of the pass-through in the Mexican economy we can consider the ‘rolling’ correlation coeffi cient between the infl ation rate and exchange rate depreciation. This correlation coeffi cient is estimated by adding one observation sequentially since 1989 (1). Figure 8.3 indicates the presence of a strong relationship between these two variables. The initial reduction of the pass-through, arguably related with the instrumentation of the infl ation targeting regime, is not continu-ous and there is a persistence of a positive relationship between infl ation and exchange rate movements.

Table 8.3 Cointegration vectors of the Johansen procedure including output, investment, US output and the real exchange rate

yt 5 b1*invt 1 b2*yust 1 b3*srt

Period invt yust srt

1982(1)–1993(4) 0.852 0.328 0.4751994(1)–2003(4) 0.761 0.420 0.855


8.3.3 The Response of Monetary Policy to Exchange Rate Shocks

In this section we look for asymmetric eff ects of the exchange rate on monetary policy. The hypothesis is that monetary policy shows an asym-metric response to movements in the exchange rate. That is, the central bank raises the interest rate in response to a depreciating exchange rate but does not modify the interest rate in response to an appreciating exchange rate. The fi nal result of this monetary policy is an appreciation of the real exchange rate.

Table 8.4 shows the evolution of the ‘corto’. This table shows the dates of the changes and the quantities (with negative sign) of the monetary base to which the penalty rate is applied. The dates without changes in the ‘corto’ are not reported in Table 8.4. Over the period between January 1996 and December 2004, the Banco de México has increased the ‘corto’

–0.02

–0.01

0.00

0.01

0.02

0.03

0.04

1 2 3 4 5 6 7 8 9 10–0.02

–0.01

0.00

0.01

0.02

0.03

0.04

1 2 3 4 5 6 7 8 9 10

–0.02

–0.01

0.00

0.01

0.02

0.03

0.04

1 2 3 4 5 6 7 8 9 10

Response inflation rate toinflation rate

Response inflation rate tooutput gap

Response inflation rate tonominal exchange rate

Note: Period 1986:01–2003:04. The horizontal axis refers to the ten periods or quarters while the vertical axis shows the response of the infl ation rate to a one standard deviation shock in each variable in the system.

Figure 8.2 Impulse – response analysis for the infl ation rate, the output gap and changes in the nominal exchange rate


on 34 occasions and reduced it six times. Never over the period has the banking system been put in a ‘largo’.

Increases in the ‘corto’ were preceded by or were simultaneous to depreciations of the peso on 18 occasions (53 percent of the number of increases) and the increase in the corto was followed by appreciations of the peso on 14 occasions. Only on four occasions (August and November 1996 and May and July 2001) were appreciations followed by a reduction of the ‘corto’ and only on one occasion did the central bank allow a sus-tained depreciation of the peso to take place without increasing the ‘corto’ (from June through December of 2003).

We also explore the possibility of an asymmetrical response of mon-etary policy to exchange rate movements with an econometric exercise. In order to assess the relevance of this asymmetric eff ect we use a two-step procedure similar to Cover (1992), Karras (1996) and Kim, Ni and Ratti (1998). This procedure contains initially an equation in order to obtain an equilibrium exchange rate trough the use of the purchasing power parity (PPP) hypothesis (Hallwood and MacDonald, 2000). We consider that a value above the equilibrium position is an undervaluation while a value under the equilibrium position is an overvaluation. These values are, after-wards, used to test for a possible asymmetric response of monetary policy to movements in the exchange rate.

0.58

0.60

0.62

0.64

0.66

0.68

0.70

0.72

1990 1992 1994 1996 1998 2000 2002

Cor

rela

tion

coef

ficie

nt

Figure 8.3 Rolling correlation coeffi cient between the infl ation rate and exchange rate depreciation, quarterly data 1989: 1–2003:4


Hence, Equation (8.2) describes the exchange rate using the purchasing power parity condition (ibid.):

St 5 b0 1 b1Pt/P*t 1 ut (8.2)

where St denotes the nominal exchange rate, Pt denotes the national price index and P*t denotes the foreign price index.

The database includes quarterly information from 1995(1) to 2004(4) on the nominal exchange rate and the price indexes of Mexico and the USA. The estimation of Equation (8.2) using ordinary least squares (OLS)

Table 8.4 Changes in the corto, 1996–2004

Date of change Corto (millions of pesos)

Date of change Corto (millions of pesos)

1996 200123 January −5 12 January −40025 January −20 18 May −3507 June −30 31 July −30021 June −40 20025 August −30 8 February −36019 August 0 12 April −30014 October −20 End of September −40026 November 0 6 December −4751998 200311 March −20 10 January −55025 June −30 7 February −62510 August −50 28 March −70017 August −70 200410 September −100 20 February −29 (daily)30 November −130 12 March −331999 27 April −3713 January −160 23 July −412000 27 August −45January −180 24 September −51May −200 22 October −57June −230 26 November −63July −280 10 December −69October −310November −350

Source: Banco de México, www.banxico.org.mx. Informe Anual and Informe sobre Politica Monetaria, various issues.


indicates that the nominal exchange rate has a long-term relationship with the price index diff erentials.

The estimated long-run equation is as follows:

St 5 3.36 1 1.202 * Pt/P*t (8.3)

ADF(1) 5 23.07

Then using Equation (8.3) we defi ne two shock functions based on the diff erences between the fi tted values and the actual values of the exchange rate (U).

Pos U1

t 5 max (shock, zero)Neg U2

t 5 min (shock, zero)

The fi rst function is associated with an undervaluation and the second one refers to an overvaluation. The second step of the procedure consists in estimating an interest rate equation including the positive and negative shocks given by the previous functions.

Rt 5 b0 1 b1U1t 1 b2U2

t 1 b3Rt21 1 et (8.4)

Equation (8.4) describes the impact of exchange rate deviations on the interest rate. This equation is estimated to fi nd evidence of asymmet-ric eff ects of the exchange rate on the interest rate. The interest rate is the three months nominal interest rate on government bonds (CETES) (Mexico’s central bank database). Making use of such a procedure we get the following estimates:

Rt 5 20.087*U2t21 1 0.145*U1

t21 1 0.76*Rt21 2 23.6*D98(3) 220.7*D95(4)

(t) (21.01) (3.16) (25.70) (27.95) (26.72) ( p) (0.26) (0.00) (0.00) (0.00) (0.00)

R2 5 0.82Normality test:6 Jarque-Bera: X2(2) 5 2.11(0.35)Autocorrelation: LM(4) 5 0.169(0.952)Heteroskedasticity: ARCH(4) 5 1.211(0.326)Period: 1995:1–2004:4


The long-run solution is:

Rt 5 20.371*U2t21 1 0.615*U1

t21 (8.5)

The main results indicate that only in the case of an undervalued exchange rate do we fi nd a statistically signifi cant coeffi cient. According to such estimates a tight monetary policy is carried out by the central bank whenever there is an undervalued exchange rate; however an overvalued exchange rate is not followed by an easy money monetary policy. The previous estimates thus confi rm that the exchange rate has an asymmetric impact on monetary policy.

8.4 PROBLEMS WITH INFLATION TARGETING AND ALTERNATIVES

As discussed in Section 8.1, the recent past has been characterized by success on the infl ation front and a poor growth performance. Two of our fi ndings in Section 8.2, the positive eff ect of the real exchange rate on output and the signifi cant pass-through of the exchange rate move-ments on prices, suggest that those two aspects, success on the infl ation front and poor growth performance, are linked through the evolution of the real exchange rate (see Table 8.1). After the sharp depreciation of 1995, the real exchange rate has been appreciating over time, a trend that was only interrupted in 2002–03. From the end of 1995 to 2002, the real exchange rate fell by 35 percent (and by 27 percent from 1995 to 2004). With a relatively stable nominal exchange rate, in an increas-ingly open economy, government policy has provided a strong disinfl a-tionary pressure. At the same time, real appreciation, through the loss of competitiveness of the economy, has contributed to a poor growth performance.

Our third fi nding in Section 8.2, on the asymmetric response of mon-etary policy to exchange rate shocks, indicates that real appreciation may be the result of a built-in bias in monetary policy towards real exchange rate appreciation. This bias has to do with the high pass-through. Over the past few years the central bank has been trying to break the link between exchange rate and prices by increasing the ‘short’ in response to ‘sharp’ depreciations. In doing so, it has reversed the depreciation itself. Because the economy has been in a process of disinfl ation in which the central bank has barely met its infl ation targets, the process is not symmetrical, that is, there is no similar incentive to reverse the appreciations that may take place as a result of shocks to the exchange rate.


8.4.1 Alternatives

Infl ation targeting is certainly a more fl exible framework for monetary policy than the previous monetary frameworks such as the use of some monetary aggregate to control the infl ation rate. That is, infl ation target-ing considers additional factors in order to control infl ation such as the exchange rate. Additionally, the hypotheses that prices and the monetary aggregate have a stable relationship and that money can be considered as the exogenous variable do not appear to be valid. For example, Carstens and Werner (1999) fi nd that base money is essentially accommodating to exogenous shocks in the Mexican case. Therefore, infl ation targeting rep-resents a new framework with could include additional possible options.

A fi rst option is to move monetary policy towards a more neutral stance (that is, a more symmetric response to exchange rate shocks). In this sense, monetary authorities should try, at least, to use a reduction in the amount of ‘corto’ in the case of an appreciation of the real exchange rate. This is more likely and feasible than in the past as infl ation tends to converge towards the long-run target and, as a result of either an enhanced credibil-ity of the central bank or a less infl ationary environment, the pass-through of the exchange rate on infl ation tends to fall. In this case, monetary policy has more degrees of freedom because the authorities have established a reputation against infl ation. In this context, a neutral monetary policy does not imply that the central bank will lose credibility in its commit-ment to control infl ation. A neutral monetary policy will have at least two main impacts. First, it will allow a monetary policy with less bias against economic growth. That is, if the side eff ect of a contractionary monetary policy is a reduction, at least in the short run, of economic growth, a neutral monetary policy will contribute, in the margin, to a more dynamic economic growth. Second, a neutral monetary policy will not contribute to an overvaluation of the real exchange rate and, therefore, will not have an additional bias against economic growth through the real exchange rate channel.

A second option is to shift from a CPI target to a domestic infl ation target (that is, a measure of infl ation purged from the direct eff ects of the exchange rate on imported goods in the CPI). Such a move would further reduce the pass-through eff ect of the exchange rate on the targeted price index and contribute to eliminate the asymmetric response of monetary policy to exchange rate shocks. In this case, a transitory shock on the nominal exchange rate will not generate a direct response in the interest rate. Under these circumstances monetary policy will be more focused on the long-term path of the infl ation rate and therefore it will not overreact to transitory shocks.


These two options preserve the infl ation targeting framework. A third, more radical departure from the current framework, would be to combine infl ation targeting with real exchange rate targeting. More precisely, the central bank would promote a competitive exchange rate by establishing a sliding fl oor to the exchange rate in order to prevent excessive appreciation (an ‘asymmetric band’ with a fl oor and no ceiling as in Ros 1995). This would imply intervening in the foreign exchange market at times when the exchange rate hits the fl oor but allow the exchange rate to fl oat freely otherwise.

Such a proposal is free from some of the orthodox objections that have been made to real exchange rate targeting. In particular, it does not require knowledge of the adequate or equilibrium real exchange rate but only of the danger zone in which overvaluation severely hurts the growth process. This is because under our proposal the central bank does not target a par-ticular real exchange rate but only establishes a fl oor on its value, leaving the real exchange rate to move freely above this threshold. Moreover, there is no problem with the amount of reserves required to defend the exchange rate since the central bank only defends the fl oor (which requires to accu-mulate rather than deplete reserves as would be the case if the central bank defended a ceiling). Of course, there is an additional orthodox objection when it comes to defending a fl oor to the exchange rate and this is that the central bank may lose control of the money supply and this could imply giving up the achievement of the infl ation target (for a fuller discussion, see Frenkel and Rapetti, 2004). The problem arises at times of excess supply of foreign currency as a result, in particular, of massive capital infl ows. It is worth noting, however, that speculative capital infl ows will tend to be deterred to the extent that the central bank clearly signals that it will prevent the appreciation of the domestic currency, thus stabilizing exchange rate expectations. If necessary, however, the central bank can impose capital account regulations on short-term capital fl ows in order to recover control over the money supply.

8.5 CONCLUSIONS

The main conclusions that emerge from our analysis can be summarized as follows. Infl ation has declined, in part during the infl ation targeting regime, from levels over 50 percent in 1995 to around 5 percent in 2004. There has also been a reduction of the pass-through of exchange rate movements into infl ation possibly as a result of an enhanced central bank credibility (with its stabilizing infl uence on infl ation expectations) or of lower infl ation itself. These achievements have, however, come with a cost,


an almost continuous appreciation of the real exchange rate that has had contractionary eff ects on output in the long run. Our empirical analysis clearly supports these assertions (the reduction of the pass-through and the negative eff ect of real appreciation on output). It also suggests that real appreciation has been fed by an asymmetrical response of monetary policy to exchange rate movements: depreciations are followed by a tightening of monetary policy while appreciations are not reversed by a relaxation of monetary conditions. The alternative to the current framework is to give a more prominent place to the achievement of a competitive and stable real exchange rate in the design of monetary and exchange rate policy.

NOTES

1. We are indebted for comments to two anonymous referees, Jerry Epstein, Roberto Frenkel, Lance Taylor, Erinc Yeldan and other participants at the Amherst/CEDES Conference on Infl ation Targeting, Buenos Aires, 13-14 May 2005. The usual caveat applies.

2. The literature on the contractionary eff ects of devaluation is very large and includes, among others, Diaz-Alejandro (1963), Cooper (1971), Krugman and Taylor (1978), Kamin (1988), Edwards (1989), Edwards and Montiel (1989), Morley (1992) and Razmi (2006). On the empirical evidence for Mexico, see Kamin and Rogers (1997), López and Guerrero (1998) and Kamin and Klau (1998).

3. The real exchange rate is defi ned as: SRt 5 St(Pus/P)t, where S is the nominal exchange rate; Pus is the consumer price index of the USA; and P is the consumer price index of Mexico.

4. The time span of the data is certainly not enough to consider the results of the cointegrat-ing vector as a long-run solution. Therefore, it is not possible to make long-term infer-ences on the basis of these results but it represents a valid approximation for the period.

5. The cointegration tests without any dummies are reported in Appendix 8.1A in Table 8A.2.

6. The non-normality can be attributed to an outlying value in 1995:2 when debt and fi nan-cial crisis took place in Mexico.

REFERENCES

Ball, L. and N. Sheridan (2003), ‘Does infl ation targeting matter?’, National Bureau for Economic Research working paper no. 9577, March.

Banco de México (1996), Informe Anual, México: Banco de México.Calvo, G.A. and F.S. Mishkin (2003), ‘The mirage of exchange rate regimes for

emerging market countries’, Journal of Economic Perspectives, 17 (4), 98–118.Carstens, A. and A. Werner (1999), ‘Mexico’s monetary policy framework under

a fl oating exchange rate regime’, Banco de México, Documento de Investigación No. 90-05, May.

Clifton, E.V., H. Leon and C.H. Wong (2001), ‘Infl ation targeting and the unem-ployment-infl ation trade off ’, International Monetary Fund IMF working paper WP/01/166.


Cooper, Richard N. (1971), ‘Currency devaluation in developing countries,’ in G. Ranis (ed.), Government and Economic Development, New Haven, CT: Yale University Press, pp. 472–513.

Corbo, V., O. Landerretche and K. Schmidt-Hebbel (2002), ‘Does infl ation target-ing make a diff erence?’, in N. Loayza and R. Soto (eds), Infl ation Targeting: Design, Performance, Challenges, Santiago: Banco Central de Chile, document 106, pp. 1–65.

Cover, J.P. (1992), ‘Asymmetric eff ects of positive and negative money-supply shocks’, Quarterly Journal of Economics, 107 (4): 1261–82.

Diaz-Alejandro, Carlos F. (1963), ‘A note on the impact of devaluation and the redistributative eff ects’, Journal of Political Economy, 71, 577–80.

Dickey, D. and W.A. Fuller (1981), ‘Likelihood ratio statistics for autoregressive time series with unit root’, Econometrica, 49 (4), 1057–977.

Edwards, Sebastian (1989), Real Exchange Rates, Devaluation, and Adjustment, Cambridge, MA: MIT Press.

Edwards, Sebastian and Peter J. Montiel (1989), ‘Devaluation crises and macro-economic consequences of postponed adjustment in developing countries’, IMF Staff Papers, 36, 875–903.

Fraga, A., I. Goldfajn and A. Minella (2003), ‘Infl ation targeting in emerging market economics’, National Bureau for Economic Research working paper no. 100019, pp. 1–50.

Frenkel, R. and M. Rapetti (2004), ‘Políticas macroeconómicas para el crecimiento y el empleo’, Paper prepared for the International Labor Organization.

Goldfajn I. and P. Gupta (2003), ‘Does monetary policy stabilize the exchange rate following a currency crisis?’, IMF Staff Papers, 50 (1), 90–114.

Hallwood C. and R. MacDonald (2000), International Money and Finance, Malden, MA: Blackwell Publishers.

Johansen, S. (1988), ‘Statistical analysis of cointegrating vectors’, Journal of Economic Dynamics and Control, 12 (2–3), 231–54.

Johansen, S., R. Mosconi and B. Nielsen (2000), ‘Cointegration analysis in the pres-ence of structural breaks in deterministic trends’, Econometrics Journal, 3, 216–49.

Kamin, S.B. (1988), ‘Devaluation, external balance, and macroeconomic per-formance in developing countries: a look at the numbers’, Princeton Essays in International Finance no. 62.

Kamin, S.B. and J.H. Rogers (1997), ‘Output and the real exchange rate in devel-oping countries: an application to Mexico’, Board of Governors of the Federal Reserve Bank international fi nance discussion paper no. 580, May.

Kamin, S.B. and M. Klau (1998), ‘Some multicountry evidence on the eff ects of real exchange rates on output’, Board of Governors of the Federal Reserve Bank international fi nance discussion paper no. 611, May.

Karras, G. (1996), ‘Are the output eff ects of monetary policy asymmetric? Evidence from a sample of European countries’, Oxford Bulletin of Economics and Statistics, 58 (2), 267–78.

Kim, J., S. Ni and R. Ratti (1998), ‘Monetary policy and asymmetric response in default risk’, Economics Letters, 60 (1), 83–90.

Krugman, P. and L. Taylor (1978), ‘Contractionary eff ects of devaluation’, Journal of International Economics, 8 (3), 445–56.

Kwiatkowski, D., P.C.B. Phillips, P. Schmidt and Y. Shin (1992), ‘Testing the null hypothesis of stationary against the alternative of a unit root’, Journal of Econometrics, 54 (1–3), 159–78.


López, J. and C. Guerrero (1998), ‘Crisis externa y competitividad de la economía mexicana’, El Trimestre Económico, October–December, LXV (260), 582–98.

Lütkepohl, H. and H.E. Reimers (1992), ‘Impulse response analysis of co-inte-grated systems’, Journal of Economic Dynamics and Control, 16, 53–78.

Lustig N. and J. Ros (1998), ‘Reforma estructural, estabilización económica y el síndrome mexicano’, Desarrollo Económico, 37 (148), 503–31.

Maddala, G.S. and I. Kim (1998), Unit Roots, Cointegration and Structural Change, Cambridge: Cambridge University Press.

Morley, Samuel A. (1992), ‘On the eff ect of devaluation during stabilization pro-grams in LDCs’, Review of Economics and Statistics, LXXIV, 21–7.

Neumann, M.J.M. and J. von Hagen (2002), ‘Does infl ation targeting matter?’, Federal Reserve Bank of St Louis Review, (July–August).

OECD (2002), Economic Surveys, Mexico City: OECD.OECD (2004), Economic Surveys, Mexico City: OECD.Phillips, P.C.P. and P. Perron (1988), ‘Testing for unit root in the time series

regression’, Biometrica, 75 (2), 335–46.Razmi Arslan (2006), ‘The contractionary short-run eff ects of nominal devaluation

in developing countries: some neglected nuances’, University of Massachusetts Amherst, Department of Economics, working paper 2005-09.

Ros, J. (1995), ‘Después de la crisis: la política económica’, Nexos, accessed October 1996 at www.nexos.com.mx/articulosEspeciales.php.

Ros, J. (2001), ‘Del auge de capitales a la crisis fi nanciera y más allá: México en los noventa’, in Ricardo Ffrench-Davis (ed.), Crisis fi nancieras en países ‘exitosos’, Santiago: CEPAL-McGraw Hill, pp. 119–57.

Schmidt-Hebbel, K. and A. Werner (2002), ‘Infl ation targeting in Brazil, Chile and Mexico: performance, credibility, and the exchange rate’, Central Bank of Chile working paper no. 171, July.

Sheridan, N. (2001), ‘Infl ation dynamics’, PhD dissertation, Johns Hopkins University.

Svensson, L.E. (1997), ‘Infl ation forecast targeting: implementing and monitoring infl ation targets’, European Economic Review, 41 (6), 1111–46.

Svensson, L.E. (1998), ‘Open economy infl ation targeting’, National Bureau for Economic Research working paper no. 6545, May.


APPENDIX 8.1A

Description of Variables

y 5 Real Gross Domestic Product (GDP) in millions of Mexican pesos of 1993, Instituto Nacional de Geografía, Estadística e Informática (INEGI), http://www.inegi.gob.mx.yus 5 Real Gross Domestic Product (GDP) in billions of US dollars at

Table 8A.1 Unit root tests

Variable ADF PP(4) KPSS(10)

A B C A B C hm ht

yt −3.19(8) 0.54(8) 1.99(8) −4.85 −0.33 2.32 0.946 0.171Δyt −3.91(8) −3.64(8) −2.54(8) −21.52 −21.54 −18.91 0.096 0.047yust −2.68(2) −0.64(2) 4.27(2) −2.69 0.12 6.98 0.967 0.065Δyust −3.45(8) −4.04(2) −2.67(1) −7.57 −7.63 −4.51 0.072 0.065invt −3.69(4) −0.97(5) 0.34(2) −2.44 −1.12 0.51 0.663 0.148Δinvt −4.11(8) −3.88(8) −3.83(8) −8.82 −8.81 −8.82 0.156 0.066srt −4.05(4) −3.26(4) −0.02(0) −2.77 −2.49 −0.03 0.262 0.101Δsrt −10.8(0) −10.90(0) −10.96(0) −10.81 −10.82 −10.87 0.108 0.071pt −1.86 (3) −3.50 (3) −0.55 (3) −0.37 −3.91 0.99 0.903 0.238Δpt −4.01 (4) −2.01 (2) −1.30 (2) −3.59 −2.60 −1.79 0.613 0.074ΔΔpt −9.09 (1) −9.08 (1) −9.14 (1) −9.80 −9.85 −9.91 0.068 0.063p*t −1.39 (3) −1.84 (3) 2.55 (3) −3.48 −4.60 7.87 0.981 0.242Dp*t −7.64 (0) −3.66 (2) −2.63 (2) −7.75 −6.50 −3.42 0.627 0.087rt −2.64 (3) −0.96 (3) −0.82 (3) −2.70 −1.10 −0.72 0.615 0.097Δrt −5.28 (2) −5.14 (2) −5.12 (2) −9.82 −9.69 −9.72 0.236 0.077r*t −2.62 (3) −1.24 (3) −1.82 (3) −1.66 −0.32 −1.71 0.769 0.076Dr*t −2.67 (8) −2.49 (8) −2.17 (8) −10.63 −10.44 −10.31 0.140 0.101st −2.10 (3) −3.35 (3) −1.71 (4) −1.02 −3.21 −1.62 0.868 0.229Δst −4.18 (2) −3.17 (2) −2.42 (2) −7.73 −6.89 −5.57 0.512 0.068gapt −3.99 (8) −4.00 (8) −4.05 (8) −8.67 −8.71 −8.75 0.043 0.043Δgapt −4.75 (8) −4.79 (8) −4.81 (8) −23.38 −23.42 −23.54 0.052 0.048

Notes: Test statistics in bold indicate a rejection of the null hypothesis. Critical values at 5 percent signifi cance level for the Augmented Dickey-Fuller and Phillips-Perron tests for a size T5100 are −3.45 including constant and trend (model A), −2.89 including constant (model B) and –1.95 without constant and trend (model C) (Maddala and Kim, 1998, p. 64). hμ and ht are the KPSS tests for the null hypothesis of stationarity around a level and deterministic linear trend, respectively. Both tests are calculated with a lag window size equal to ten. The 5 percent critical values for the two tests are 0.463 and 0.146, respectively (Kwiatkowski et al. 1992, p. 166). Small letters represent the values in logarithms.

Sources: Dickey and Fuller (1981) and Phillips and Perron (1988).


prices of 2000, (the series is already seasonally adjusted), US Department of Commerce: Bureau of Economic Analysis.p 5 Mexican consumer price index (base 2002 5 100), Bank of Mexico, http://www.banxico.org.mx.p* 5 US consumer price index (base 1982–84 5 100), US Department of Commerce, Bureau of Economic Analysis.GAPY 5 Deviation of the real Gross Domestic Product (GDP) from potential output obtained using the Hodrick-Prescott fi lter.S 5 Exchange rate (pesos per US dollar), 48-hour interbank exchange rate. The data is taken from the last day of each quarter, Bank of Mexico, http://www.banxico.org.mx.Sr 5 Real exchange rate defi ned as S (P*/P) where S is the nominal exchange rate, P refers to domestic prices and P* to foreign prices.r 5 Nominal interest rate on three-month treasury bills (CETES 91 days). Average of the last month of the quarter, Bank of Mexico, http://www.banxico.org.mx.r* 5 Nominal interest rate three-month treasury bill, Board of Governors of the Federal Reserve System

Table 8A.2 Mis-specifi cations tests of the Vector Autoregressive Model: st, (pt 2 p*t ) and (rt 2 r*t )

Variable Autocorrelation: LM(4)

Heteroskedasticity ARCH(4)

NormalityJ-B

st F(3,14) 5 0.40 [0.75] F(3,11) 5 0.095 [0.96] c2(2) 5 8.39[0.02]*pt 2 p*t F(3,14) 5 1.59 [0.24] F(3,11) 5 0.31 [0.82] c2(2) 5 3.57[0.17]rt 2 r*t F(3,14) 5 2.05 [0.15] F(3,11) 5 0.09 [0.97] c2(2) 5 9.38[0.01]**

Note: *, ** indicate a rejection of the null hypothesis to 5 percent and 1 percent signifi cance level. Period 1995(1)–2004(4).

179

9. Five years of competitive and stable real exchange rate in Argentina, 2002–07Roberto Frenkel and Martín Rapetti1

9.1 INTRODUCTION

In 1991 Argentine authorities established the convertibility regime, which implied the pegging of the peso (AR$) to the US dollar ($) by law and the validation of contracts in foreign currencies. The new monetary arrange-ment also stipulated that the central bank must fully back the monetary base with foreign reserves,2 what in practice turned the central bank into a currency board. The convertibility regime was the pillar of a broader stabilization program, intended to take the economy away from the high infl ation regime settled since the mid 1970s, which had led to two brief hyperinfl ationary episodes in 1989 and 1990. The program also included an almost complete liberalization of trade fl ows and the full deregulation of the capital account of the balance of payments. It was jointly applied with an impressive process of market-friendly reforms, targeting the priva-tization of a large proportion of state-owned fi rms.

The program successfully managed to stop infl ation and initially spurred rapid growth. However, as happened with many other stabiliz-ation programs in the region, it led to the appreciation of the real exchange rate, which made economic growth highly dependent on external debt accumulation.3 Since the Asian and Russian crises, and especially after the Brazilian devaluation in 1999, the deceleration of capital infl ows put the economy into a defl ationary trend that ended up in a fi nancial and exter-nal crisis in 2001–02. Between the last days of 2001 and the beginning of 2002, Argentina declared the default of its international debt and devalued the peso. The collapse of the convertibility regime implied a 21 percent contraction in GDP with respect to the peak of mid-1998 and a rise in the unemployment rate up to 21.5 percent, taking half of the population below the poverty line.

However, only one quarter after the devaluation and default economic


activity gradually started to recover. By the end of 2002, once the govern-ment managed to stabilize domestic fi nancial markets, the recovery gained momentum and since then the economy has shown an impressive perform-ance. In fi ve years – from the fi rst quarter of 2002 to the same period in 2007 – GDP has been growing at 8.1 percent annual rate, reaching a peak 18.1 percent higher than the one in mid-1998. The investment rate rose to 22 percent of GDP (on seasonally adjusted basis), which is the maximum range of the offi cial time series beginning in 1993 and continues to grow at a higher pace than GDP. During these fi ve years exports have expanded at a slightly higher rate than GDP, but its rate of growth has substantially increased from mid 2004 onwards.

Current economic evolution contrasts with Argentina’s economic per-formance of the last 60 years. Since the Second World War economic growth has been low and very volatile, especially in the second fi nancial globalization period beginning in the mid 1970s. For the fi rst time in 30 years Argentina has grown fi ve years in a row. More importantly, current expansion is based on solid macroeconomic fundamentals. The volatility of Argentine growth has been typically associated with current account and fi scal defi cits. Between the mid 1940s and the mid 1970s, macro-economic evolution was characterized by stop-and-go cycles related to external imbalances. During the second fi nancial globalization period, the availability of external funds momentarily relaxed the external constraint to growth, but it led to two episodes of explosive fi scal and external debt accumulation,4 one in the late 1970s and the beginning of the 1980s and the other during the convertibility regime period.

In contrast to those traditional fi scal and external imbalances, the current macroeconomic confi guration stands out with the existence of external and fi scal surpluses. Certainly, the debt restructuring in 2005 – implying a $67 billion reduction in the nominal stock – softened both external and fi scal requirements, releasing resources for private sector spending. Similarly, favorable external conditions – especially the high prices of some commodities – have also played a role. However, in our view the main factor behind the current success is the offi cial policy aiming at preserving a stable and competitive real exchange rate (SCRER). The SCRER has been a key factor explaining the current account adjustment, which passed from a $14.5 billion defi cit in 1998 to $7.6 billion surplus in 2006. From this $22 billion adjustment, $20 billion came from the improvement in the trade balance, which is mainly attributable to the eff ects of the exchange rate depreciation.

The infl uence of the SCRER on the fi scal accounts performance has also been important. After devaluation, the government introduced taxes on traditional exports, mainly agricultural products and oil. In practice,

Five years of competitive and stable real exchange rate in Argentina 181

this measure implied the introduction of multiple exchange rates that contributed to reduce the pass-through of devaluation to wage-goods prices, but also to capture part of the rent that these sectors obtained from the competitive real exchange rate. In 2006 the federal administration recorded a primary surplus of 3.5 percent of GDP and a total surplus of 1.8 percent of GDP, from which taxes on exports accounted for 63 percent of the former and 122 percent of the latter.

In our view, the positive eff ects of the SCRER policy are the princi-pal factors explaining the rapid growth experienced so far. This policy promotes economic growth not only by preserving external and fi scal accounts sustainability, but also by providing incentives to the tradable sector and thus encouraging the expansion of its production, employment and investment. Although the success of the SCRER strategy throughout these fi ve years has undoubtedly had a persuasive impact among ana-lysts, skepticism remains.5 The SCRER policy collides with conventional wisdom, particularly with the trilemma paradigm.

In this chapter, we argue that a macroeconomic regime based on a SCRER is both desirable and manageable for a developing open economy. The next section describes the evolution of monetary and exchange rate policies in Argentina in the post-convertibility period. Section 9.3 discusses the usual criticisms against the SCRER policy and shows the conditions in which this policy is sustainable. Section 9.4 presents some concluding remarks regarding the management of a macroeconomic regime with a SCRER as an intermediate target.

9.2 MONETARY AND EXCHANGE RATE POLICIES IN THE POST-CONVERTIBILITY PERIOD

The deceleration of capital infl ows that led to the 2001–02 crisis began in mid 1998. This process took place simultaneously with a persistent rise of the sovereign risk premium. However, the divergent trends in the domestic fi nancial market that triggered the collapse of the convert-ibility regime only started in October 2000, associated with the political turmoil caused by the Vice-President’s resignation. The process followed a simple dynamics. Devaluation expectations and the perception of a higher risk of default led the private sector to withdraw deposits and run against the central bank’s international reserves. There were no bankruptcy reports of failing banks because the central bank supported the liquidity of the banking system. Despite several signals issued by the government aiming at changing the expectations, the intensifi cation of this process could not be stopped. In December 2001 restrictions on


capital outfl ows and on the withdrawal of deposits (the so-called ‘cor-ralito’) were established.

After the abandonment of the convertibility regime, the government aimed to restrain the capital outfl ow and stabilize the foreign exchange (FX) market by introducing a dual exchange rate regime. The idea was to use this scheme only temporarily, in order to stabilize the nominal exchange rate while the domestic prices absorbed the impact of the devalu-ation, and then pass to a fl oating rate regime. The authorities also decided to convert to pesos most of the domestic debts contracted in dollars (bank credits, rents and so on) at a AR$/$ 1 rate (plus indexation to consumer price index (CPI) infl ation), thus neutralizing most of the eff ects of relative price change on the debtors’ balance sheets. In contrast, banks’ deposits originally denominated in dollars were ‘pesoifi cated’ at a AR$/$ 1.40 rate (plus indexation to the evolution of CPI infl ation).6 Together with the ‘pesoifi cation’, the authorities unilaterally decided to extend the maturity and duration of all deposits, including those originally contracted in pesos. In exchange, private depositors received certifi cates for the reprogrammed deposits.

In February 2002 the FX market was unifi ed and the peso started to fl oat freely. Given the political and economic uncertainty, the exchange rate skyrocketed fed by self-fulfi lling expectations. Interestingly, this process developed in an illiquid environment because of the restrictions on the withdrawal of cash from banks. The erratic monetary policy that followed in the fi rst quarter of 2002 also failed to stabilize the exchange rate. The authorities delayed the launching of a domestic asset that could perform as a potential substitute for foreign currency. Given the distrust in banks and in the Treasury, the economic depression and the growing infl ation, the international currency appeared as the only asset available to allocate fi nancial savings. Only two and half months after the devalua-tion the central bank started to issue notes (the Lebac) in order to supply a fi nancial instrument that could compete with the dollar.

All these factors contributed to deepen the perverse dynamics of the fi nancial variables during the fi rst semester of 2002. The capital fl ight from domestic assets between March 2001 and mid 2002 is illustrated in Figure 9.1, which shows the large fall in private bank deposits7 and international reserves, while the nominal demand for cash remains stagnant. These developments provide evidence for the substitution of local assets (cash and deposits) in exchange for external assets (international reserves).

The result of the asset substitution aff ected the FX market. The nominal exchange rate (NER) and real exchange rate (RER)8 rose continu-ously in the fi rst semester of 2002 (around 260 percent and 180 percent, respectively). Their paths are shown in Figure 9.2. Real exchange rate


overshooting was so pronounced that in June 2002 its value was almost 193 percent higher than the 1980–2001 period average value, and 309 percent higher than the convertibility decade average.

These disruptive trends began to revert in July 2002. The turning point was the exchange rate stabilization. Several factors contributed to this outcome. Controls on FX transactions9 had been introduced in November 2001 – before the convertibility collapse – and they were further tightened in March 2002. Since June 2002 controls and interventions in the FX market were strengthened in order to conduct a systematic policy to sta-bilize the exchange rate. The decision that export revenues surpassing $1 million had to be sold directly to the central bank was especially important in this regard. This became the main source of international reserves ac-cumulation for the monetary authority, which in turn agreed to increase the volume of its interventions in the FX market.

Financial market behavior itself also contributed to stop the bubble in the exchange rate. On the one hand, local interest rates skyrocketed (Figure 9.3). In July 2002 the average time deposits annual interest rate reached a

0

10000

20000

30000

40000

50000

30000

40000

50000

60000

70000

80000

90000

100000

2001 2002 2003 2004 2005

Pesoification Exchange ratestabilization

Demand for cashCentral bank reservesPrivate bank depositsLebac

Source: Central Bank of Argentina.

Figure 9.1 Demand for cash, central bank international reserves, Lebac and private bank deposits (right axis) (in millions of pesos and dollars)


76 percent peak, and the annual interest rate of the 14-day Lebacs reached almost 115 percent. Thus local fi nancial assets began to appear more attractive as substitutes for the dollar. On the other hand, as mentioned above, the real price of the dollar reached very high and ‘abnormal’ levels in historical terms (that is, the prices in dollars of domestic assets, non-tradable goods and salaries were perceived as abnormally low). In this context, once the authorities managed to stop the exchange rate bubble in July, the public rapidly changed expectations and the market started to show an appreciation trend.

Thus, in the second half of 2002 a phase of monetary and fi nancial variables normalization started. After reaching a peak of almost AR$/$ 4 during the last days of June, the exchange rate began to experience a smooth nominal appreciation trend. Although the infl ation rate was already low and decelerating, the rise in domestic prices contributed to the real appreciation. In that context, local assets became increasingly

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2001 2002 2003 2004 2005 2006

MRERNERRER

Notes:* Calculated using US and Argentina consumer price indexes.** Elaborated using monthly data of the exchange rates of the 18 principal trading partners

of Argentina, adjusted by the respective consumer price indexes.


Figure 9.2 Bilateral nominal (NER) and real exchange rate (RER) with the US and multilateral real exchange rate (MRER) (in pesos and indexes 1 5 December 2001)


attractive. Bank deposits began to grow, as did the demand for Lebac, local shares and the demand for cash (Figure 9.1). This portfolio substitu-tion in favor of local assets resulted in a persistent drop in the interest rates (Figure 9.3).

The normalization in fi nancial activity dissipated, disrupting expecta-tions and thus favored the recovery of private expenditure. Interestingly, this recovery took place without signifi cant contribution from bank credits. Even though private deposits gradually improved allowing the recuperation of banks’ liquidity, credit to the private sector continued shrinking until late 2003. The fi nancial crisis appeared to have persistent eff ects on the behavior of bank credit, which at the beginning of 2007 was still below the peak reached in 1998 (Figure 9.4).

Domestic expenditure was mainly fi nanced by private sector cash holdings. Figure 9.5 shows the increase in cash holdings since the fourth quarter of 2001. Both the monetary base/GDP ratio and the monetary base/total bank deposit ratio showed very high rates of growth and also relatively high levels in comparison to the convertibility period. Although the low interest rates on banks’ deposits (and the tax on fi nancial transac-tions) have contributed to that performance, this behavior seems to be another persistent consequence of the fi nancial crisis.

0

20

40

60

80

Mon

thly

avg

100

120

2001 2002 2003 2004

14d Lebac91d LebacTime deposits 30–59dPrime 30d

Exchange rate stabilization


Figure 9.3 Interest rates in pesos: Lebac (14 and 91 days), time deposits (30 to 59 days) and prime (30 days) (monthly average, in %)


8

12

16

20

Ban

k cr

edit/

GD

P in

per

cent

age 24

28

96 97 98 99 00 01 02 03 04 05 06

Source: Central Bank of Argentina; Ministry of Economy of Argentina.

Figure 9.4 Bank credit to private sector in relation to GDP seasonally adjusted

15

20

25

30

35

Mon

etar

y ba

se/to

tal d

epos

its

Mon

etar

y ba

se/G

DP40

45

50

4

5

6

7

8

9

10

11

12

1994 1996 1998 2000 2002 2004

MB/GDPMB/Deposits

Source: Central Bank of Argentina; Ministry of Economy of Argentina.

Figure 9.5 Monetary base in relation to total bank deposits and with GDP seasonally adjusted


The nominal and real appreciation process stopped around mid 2003, when the government decided to manage the fl otation of the exchange rate in order to preserve the SCRER. The impact of the SCRER on economic activity, employment and external and fi scal accounts was proving to be highly favorable. Thus, the government gradually started to recognize and make explicit reference to the importance of preserving the SCRER in the offi cial economic strategy. On the contrary, central bank authorities never made any explicit statement regarding the existence of any exchange rate target. According to their offi cial statements and documents, the per-manent intervention in the FX market has been oriented to accumulate international reserves for a precautionary purpose, namely to protect the economy from international capital markets volatility. Statements aside, the joint intervention of the central bank and the Treasury in the FX market actually controlled the price of the dollar in a narrow range between AR$2.8 and AR$3.1. The resulting fl uctuation of the exchange rate in this range made the multilateral real exchange rate remain stable around a level 129 percent higher than the one at the end of the converti-bility regime. The bilateral real exchange with the US dollar also remained stable for some years, but since early 2005 it has shown a soft appreciation trend (see Figure 9.2).

In 2002, when the Congress passed a law revoking the currency board, the government decided to keep the central bank’s independence with the mandate of pursuing low infl ation rates as its primary mission. Given that the economy was still absorbing the eff ects of the crisis and the devaluation and that the domestic fi nancial markets had shrunk signifi cantly, the central bank disregarded the option of following an infl ation targeting regime. The transmission mechanisms through the interest rate on aggregate demand were thought to be uncertain and weak.10 Instead, the authorities opted to follow a more pragmatic policy based on broad quantitative monetary targets. From 2003 on, targets have been announced at the beginning of every year throughout the central bank monetary programs, in which the authorities commit themselves to maintain monetary aggregates within a certain range. Given the uncertainty surrounding the eff ects of monetary policy, the central bank has tended to set these ranges suffi ciently broadly. However, their upper bounds ended up being systematically lower than the monetary expansion arising from the intervention in the FX market to preserve the SCRER. Thus, since 2003 the central bank has dealt with two ‘confl icting’ objectives: the preservation of a competitive exchange rate by intervening in the FX market and at the same time the attainment of the targets of monetary expansion announced in the monetary program.

The tension between these two policy objectives can be observed in Table 9.1, which shows the sources of variation of the monetary base. In the fi rst

188

Tab

le 9

.1

Sour

ces o

f var

iatio

n of

the

mon

etar

y ba

se (

mon

thly

ave

rage

var

iatio

n, in

mill

ions

of p

esos

or d

olla

rs)

Cen

tral

ban

k FX

in

terv

entio

nM

onet

ary

base

va

riat

ion

‘Exc

ess’

of

mon

etar

y ex

pans

ion

Cen

tral

ba

nk

ster

iliza

tion

Ass

ista

nce

to b

anks

Ass

ista

nce

to th

e T

reas

ury

Oth

ers

Tre

asur

y FX

in

terv

entio

na

2002

:01b

–1 45

0 4

06–1

856

–216

1 426

124

522

n.a.

2002

:02c

1 28

11 6

74–3

93–2

7086

250

327

n.a.

2003

1 37

4 8

0956

5–4

20–1

25–5

232

28

2004

1 93

1 4

831 4

47–3

23–6

01–5

4319

112

2005

2 35

2 1

862 1

66–8

36–9

39–3

53–3

934

320

06 3

584

2 113

1 471

–812

–204

–446

–8 4

0

Not

es:

a.

In m

illio

ns o

f dol

lors

.b.

C

alcu

late

d fo

r the

per

iod

Feb

ruar

y–Ju

ne 2

002.

c.

A ca

ncel

latio

n of

a B

anco

Nac

ión’

s red

iscou

nt b

y th

e Tre

asur

y w

ith a

ssist

ance

of t

he ce

ntra

l ban

k in

Sep

tem

ber 2

002

for a

bout

AR

$350

0 m

illio

n is

omitt

ed.

Sour

ce: C

entr

al B

ank

of A

rgen

tina.


semester of 2002 the central bank intervened in the FX market providing dollars to contain the depreciation pressures. Thus, the FX intervention operated as a source of monetary base contraction. Once the exchange rate was stabilized, the accumulation of international reserves resulted, on the contrary, in a source of monetary expansion. During the second semester of 2002 this source of monetary expansion was easily absorbed by the rapid growth in the demand for cash caused by the re-monetization of the economy. However, since 2003 the gradual deceleration of money creation established in the monetary programs in order to maintain infl a-tion expectation under control started to confl ict with the increasing expansion of monetary base generated by central bank’s intervention in the FX market aiming to preserve the SCRER. Since the amount of mon-etary base created to intervene in the FX market (fi rst column in Table 9.1) exceeded the actual expansion of the monetary base to accomplish the monetary targets (second column), an ‘excess’ of monetary expansion (third column) had to be absorbed.

This ‘excess’ of monetary expansion has been absorbed through several mechanisms. Throughout 2003 the sterilization operations implemented by the issuing of central bank notes were especially relevant. The need for sterilization increased during 2004 and 2005. However, the central bank could limit the issuing of Lebac because other compensatory mechanisms began to operate. In the fi rst place, as liquidity grew the banks started to service the debt incurred with the central bank during the fi nancial crisis. Hence, banks’ capital payments and especially the payment of interests operated as a source of monetary base contraction. In 2005 the central bank launched a program allowing the acceleration of banks’ debts amort-izations, reinforcing this contractionary mechanism. By early 2006 most banks had cancelled their debts with the monetary authority.

The Treasury also helped to absorb the ‘excess’ of monetary expansion. While in 2002 a net fl ow of fi nancing from the central bank to the Treasury was observed, in 2003, and especially since 2004, the transactions between the Treasury and the central bank operated as a source of contraction of the monetary base. The Treasury’s purchases of international reserves with the proceeds of the primary surplus gave place to a monthly average contraction of the monetary base of AR$543 million in 2004. The main purpose of these operations was to continue servicing of the debt with the multilateral fi nancial institutions. The Treasury and other offi cial agencies also accumulated part of the fi scal surplus in foreign currency and thus intervened directly in the FX market to alleviate central bank’s manage-ment of the ‘confl icting’ objectives. These operations started in late 2002 and gradually expanded afterwards, thus becoming an important policy instrument (see last column of Table 9.1).


In 2005 the sterilization needs through the issuing of Lebac increased again. Thus, in order to soften the appreciation pressures in the FX market, controls on the capital account were introduced in June. Basically, the new measures established that all capital infl ows – excluding the issuing of new private and public debt, international trade fi nancing and foreign direct investment – would be subject to a 30 percent unremunerated reserve requirement for at least 365 days. This strategy is inspired by that applied in Chile in the early 1990s and attempts to reduce short-term capital infl ows. However, controls left open ways to avoid the reserve require-ments. For instance, capital infl ows can easily circumvent the reserve requirement by operating through the stock exchange market (by buying domestic assets abroad and selling them in the local market). There has been no evidence of a reduction in the supply of dollars in the FX market after the measures were implemented. Local analysts believe that controls are ineff ective and even the authorities do not reject the idea that they were introduced more as a signal of the offi cial willingness to maintaining the SCRER strategy rather than as an eff ective control mechanism.

As from 2006 monetary policy stopped targeting the monetary base and started to focus on M2. The authorities argued that the change in the target was due to the increasing monetization of the economy and the gradual recovery of bank credit. In these conditions, it was argued, the use of a larger monetary aggregate represented a step forward toward the fi ne-tuning of monetary policy. In practice, the switch of the monetary aggregate target helped to relax the confl icting management of exchange rate and monetary policies. The central bank was facing increasing dif-fi culties to accomplish the monetary base targets. As Table 9.1 shows, the ‘excess’ of monetary expansion had risen substantially between 2003 and 2005. The use of M2 as a target gave the authorities greater fl exibility to conduct the two-target policy, allowing for greater intervention in the FX market and expansion of the monetary base.

In summary, during the post-convertibility period the central bank has been able to conduct the two-target policy successfully. Moreover, while doing so it has obtained quasi-fi scal surpluses every year. Some analysts have argued that the management of monetary policy focusing on two targets has had an infl ationary bias. Certainly, infl ation accelerated during 2004 and 2005 and has remained stable around an annual rate of 10 percent since 2006. In our view, the acceleration of infl ation is due not to inconsistencies in the management of monetary and exchange rate poli-cies, but to the lack of coordination between these and the fi scal policy. The expansion of public spending well above the increase of tax revenues since 2006 has implied an expansionary fi scal impulse to an already fast-growing aggregate demand. Given the fact that monetary and exchange


rate policies focus on preserving a SCRER – which is intended to put the economy in a high growth path – fi scal policy is the only macroeconomic instrument available to moderate aggregate demand when infl ationary pressures arise.

9.3 THE ECONOMICS OF THE SCRER

The notion that a SCRER favors economic development has a long tradit-ion in economic theory. Advocates of the outward orientation approach to development during the 1960s and 1970s pointed to the SCRER as a key element for that strategy (Balassa, 1971; Díaz Alejandro, 1979). According to this view, a competitive real exchange rate boosts economic growth because it softens the balance of payment constraint and favors the development of tradable activities, which tend to be more dynamic. The stability of the exchange rate is also important because low volatility reduces the risk and uncertainty of investment in tradable sectors. These arguments have been recently revitalized by modern scholars.11 Besides the traditional eff ects, modern advocates also emphasize that a development strategy based on a SCRER is market friendly (avoiding rent-seeking practices) and compatible with free trade agreements. In recent years many studies have documented a statistically and economically positive relationship between growth and real exchange rates.12

The preservation of a SCRER has also been invoked for other reasons. Maintaining a competitive real exchange rate typically involves interven-tion in the FX market and the accumulation of international reserves. It is a well documented fact that international fi nancial integration may lead to macroeconomic instability and increases the likelihood of external crises. Some scholars argue that international reserves accumulation serves as a shield against volatile capital fl ows, especially for developing countries (Feldstein, 1999). Empirical studies show a positive relationship between reserve accumulation and growth (Polterovich and Popov, 2002).

Another less studied motive is that competitive real exchange rates promote job creation. Besides the above-mentioned growth eff ects, a SCRER may impact on employment through a more intense use of labor (Frenkel, 2004). A competitive parity favors labor-intensive activities and sectors and also the substitution of expensive inputs (such as imports) in favor of labor across sectors.13

Probably because of all these reasons, the preservation of a SCRER does not attract much criticism by itself. Few scholars deny the benefi cial aspects of stable and predictable relative prices and the positive eff ects on growth. In some cases welfare arguments against public intervention in


the FX market are raised. But the optimality of the free market determi-nation of the exchange rate and the argument that the public sector has no informational advantage over the private sector are not very appeal-ing ideas in the specialized discussion about exchange rate regimes and policies. The apparent volatility of capital fl ows and the instability and unpredictability of free-fl oating exchange rates greatly lessen the relevance of those ideas (Frankel and Rose, 1995). Moreover, in some scholars’ view the free-fl oating exchange rate indeterminacy and unpredictability is pre-cisely the deeper foundation of the need for managing the exchange rate (Blecker, 2005). This is particularly emphasized in countries in which the real exchange rate plays a crucial role in the economic performance.

Skepticism towards the SCRER policy points to the ability of govern-ments to conduct it. The main objection is that the real exchange rate – as any real variable – is not under the government’s control, at least in the long run. However, given the weak empirical support of real exchange rate determination models,14 the objections relevant for economic policy for-mulations are based on the trilemma or impossible trinity argument. The trilemma says that it is impossible for a country to simultaneously main-tain free capital mobility, active monetary policy and the ability to manage the exchange rate. One of these features must necessarily be given up. In other words, the trilemma says that in an economy open to capital fl ows it is impossible for the authorities to simultaneously control the exchange rate and the interest rate (or the monetary base).

There are at least two ways to express the objection to the SCRER policy based on the trilemma. One of them argues that targeting the exchange rate implies a central bank intervention in the foreign exchange market. In doing so, the central bank loses its ability to control money supply. Targeting the exchange rate and controlling the money supply can be simultaneously pursued only if capital fl ows are regulated. However, the eff ectiveness of capital regulation tends to decrease, because the private sector innovative capacity is greater than the public sector regula-tory ability. The conclusion is that central banks have to choose between two poles (Fischer, 2001): active monetary and fl oating exchange rate or hard peg cum passive monetary policy.

The second way to express the objection focuses on the argument of controlling infl ation. If the interventions in the exchange market target the real exchange rate (instead of the nominal exchange rate), no nominal anchor remains for the public to confi gure infl ationary expectations. Since the central bank cannot control the money supply, the infl ation rate is completely out of control.

The trilemma is essentially a policy argument, logically derived from interest rate parity theorems for open economies. When forwards exchange


markets are not fully developed the relevant theorem is the uncovered interest parity (UIP) condition. The UIP states that the returns of two perfect substitutes assets nominated in diff erent currencies should be equal. This implies that the domestic interest rate (i) should equalize the sum of the foreign interest rate (i*) and the expected variation of the nominal exchange rate (E(S

#) 5 SE

t11 2 St/St). With the additional assumptions of small country (i*t 5 i*) and perfect foresight (SE

t11 5 S), the UIP condition formally implies:

it 5 i* 1SSt

2 1 (9.1)

Equation (9.1) is a simple model with two unknowns: the domestic interest rate and the exchange rate (St). Under a credibly fi xed exchange rate regime (St 5 S), the model is solved by determining the interest rate endogenously equal to the international rate. In other words, the govern-ment is able to set the exchange rate but loses control of the monetary policy. When the exchange rates fl oats freely, Equation (9.1) is solved by setting the domestic interest rate exogenously. This is the case in which governments have an active monetary policy at the cost of letting the exchange rate fl oat. If both the interest rate and the exchange rate are exogenous, Equation (9.1) is overdetermined. The only way to avoid this situation is to consider the imposition of capital controls, which prevent arbitrage forces to make the parity hold.

In any model conclusions critically depend on the assumptions. In the case of the trilemma one crucial assumption is that assets are perfect substitutes. If this assumption is relaxed the validity of the trilemma as a general theorem characterizing the performance of economies open to capital fl ows no longer holds.15 Moreover, it has been recognized for a long time in open economy macroeconomics that in the context of free capital mobility central banks have room to conduct active monetary policy and control the nominal exchange rate when assets are imperfect substitutes.16 The degrees of freedom of monetary policy vary inversely with the degree of assets substitutability.

The degrees of freedom of monetary policy also depend on the insti-tutional characteristics of the central bank, and the situation of the FX market. In a case of excess supply of foreign exchange at the targeted exchange rate, if the central bank is allowed to issue bonds to sterilize, it can control both the prevailing exchange and interest rates by purchasing all the excess supply of international currency in the FX market and ster-ilize the monetary eff ect of that intervention through the issuing of bonds in the monetary market. The central bank has two available instruments to perform its two targets: the intervention in the exchange market to control


the exchange rate and the intervention in the money market to control the interest rate. Tinbergen’s maxim is fulfi lled. The excess supply of internat-ional currency, at the exchange rate targeted by the central bank, implies an excess demand for domestic assets at the prevailing domestic interest rate. The fully sterilized intervention in the exchange market can be imag-ined as a policy implemented in two steps. In the fi rst one, before steriliz-ation, the central bank intervention generates a monetary base expansion. The resulting situation would show a higher amount of monetary base, the same amount of domestic bonds and an interest rate lower than the initial one. In the second step, the complete sterilization fully compensates for the change in the private portfolio that took place in the fi rst step. The central bank absorbs the increment in the monetary base and issues an amount of domestic assets equal to the initial excess demand for domestic assets (the excess supply of international currency) turning the domestic interest rate to its previous level (Bofi nger and Wollmerhäuser, 2003).

Therefore, if assets are imperfect substitutes and sterilization is allowed, the central bank’s ability to simultaneously manage the exchange rate and the interest rate critically depends on the existence of an excess supply of international currency at the targeted exchange rate. In this setting the trilemma is invalid. It seems that this conclusion is not generally acknowl-edged because the literature discussing monetary autonomy and exchange regimes rarely considers situations of excess supply of international cur-rency. It is mostly focused on balance of payments defi cit situations.17

Certainly, in excess demand contexts the predictions of the trilemma are generally valid. Even when assets are imperfect substitutes, in these situ-ations even powerful central banks have a limited capacity to intervene in the FX market. The limit is determined by the stock of international reserves. Consequently, it may be argued that even powerful central banks cannot simultaneously control the exchange rate and the interest rate in contexts of excess demand for international currency. But there is no symmetry between excess demand and excess supply situations. In the fi rst case the trilemma is valid while not necessarily in the second one. The asymmetry lies in the fact that in the fi rst case sterilization is constrained by a fi xed stock (that is, the international reserves), while in the second sterilization may be done indefi nitely because of a variable stock (that is, central bank’s bonds). The central bank’s ability to issue bonds but not international reserves is the key diff erence.

This ability raises the question whether it is possible to carry the fully sterilized intervention policy under excess supply of foreign currency situ-ations permanently. In order to do so, the central bank has to fulfi ll a sus-tainability condition: its net worth should not follow an explosive trend. Sustainability, therefore, depends on the magnitudes of the international


and the domestic interest rates and on the rate of variation of the nominal exchange rate. Taken as given the international interest rate and the trend of the nominal exchange rate, the sustainability condition depends on the domestic interest rate. The central bank enjoys autonomy to determine the domestic interest rate, but in order to be sustainable the policy must deter-mine domestic interest rates lower than a certain upper limit. This limit can be formally determined as follows.18 Assume a central bank that holds international reserves (R) as its unique asset and issues monetary base (H) and remunerated liabilities (L) yielding the domestic interest rate set by the monetary authority (it). Therefore, the central bank’s net worth (N) at any point in time would be:

Nt 5 StRt 2 (Ht 1 Lt) (9.2)

In each period the central bank earns the yielding of international reserves – which for simplicity we assume are invested at the international interest rate – and serves the interest payments of its remunerated liabilities. There is also a valuation eff ect on the international reserves due to the variation of the exchange rate (S

#). Since the changes in the stocks cancel out, central

bank’s quasi-fi scal result is equal to the variation of its net worth.

dN 5 SR(i* 1 S#) 2 iL (9.3)

A simple (although restrictive) condition for the central bank’s net worth not to follow an explosive trend is to assume that the quasi-fi scal result has to be non-negative (dN $ 0). Under this sustainability condition, we obtain the maximum domestic interest rate that makes the fully sterilized intervention policy sustainable:

i max t 5

i* 1 S#

Lt/StRt (9.4)

It follows that there is a range of interest rates from zero to i max that makes the fully sterilized intervention policy sustainable. Given that central banks typically benefi t from seigniorage and infl ation tax revenues, the case in which Lt , StRt does not seem unlikely. In these cases, the upper limit of this range would be greater than the sum of the international inter-est rate and the rate of variation of the nominal exchange rate.

It is important to notice that since i max depends on the behavior of Rt and Lt, the range of sustainable interest rates also evolves over time. Given a set of variables and parameters of the economy (such as the infl a-tion rate, the elasticity of money demand and the rate of variation of the exchange rate), i max would tend to decrease as the interest rate set by the


central bank increases. Thus, in order to keep the policy in a sustainable trend, the cumulative sterilization cost should be bounded and manage-able. A key point for sustainability is, therefore, that the domestic interest rates set by the central bank should be ‘moderate’ in the mentioned sense.

9.4 CONCLUDING REMARKS

In this chapter we show that monetary and exchange rate policies target-ing a SCRER are viable for developing open economies. We illustrated our argument with recent Argentine experience, which is just one of many other economies like China or India following this strategy. It is important to notice, however, that when a country is trying to preserve a SCRER, monetary, exchange rate and fi scal policies should be coor-dinated. Otherwise, potential confl icts between domestic goals – such as the exchange rate, infl ation rate and employment – might arise. Recent infl ationary pressures in Argentina could be an example of confl icts arising from the lack of coordination between economic policies.

An outline of a macroeconomic regime targeting a SCRER in which monetary, exchange rate and fi scal policies are coordinated is briefl y described as follows.19 First, it is important to mention that such a macro-economic regime does not imply segmentation between objectives and instruments. The preservation of a SCRER, the level of employment and the control of infl ation set the priorities and the restrictions that the economic policy must fulfi ll. Monetary, exchange rate and fi scal policies should be coordinated in order to guarantee the consistency between the multiple objectives.

The exchange rate policy should focus on signaling the stability of the real exchange rate in the medium and long term, in order to set in motion the positive feedbacks mentioned in Section 9.3. In particular, the emer-gence of appreciation trends should be avoided to prevent self-fulfi lling bubbles that increase the monetary ‘costs’ of buying FX interventions and also because real exchange rate appreciation may harm the profi tability of tradable activities, making many of them non-viable and forcing fi rms to close.

The preservation of a SCRER does not mean short-run indexation of the nominal exchange rate to domestic prices. The fl exibility and advan-tages of fl oating the nominal exchange rate in the short run should be pre-served. Central bank interventions in the FX market have to achieve two confl icting goals: they have to prevent expectations of real exchange rate appreciation and allow the nominal exchange rate to fl oat in order to dis-courage short-term speculative capital fl ows. The interval of interventions


has to be narrow enough to perform the fi rst function and wide enough to perform the second.

The FX market behaves like an asset market. Buying and selling decis-ions are mostly based on expectations. If central bank interventions and signals stabilize expectations around the SCRER – a necessary condition for that is the consistency of monetary, exchange rate and fi scal policies and the robustness of the external sector accounts – the market forces by themselves will tend to stabilize the exchange rate. The monetary ‘costs’ of central bank interventions will be lower and fewer interventions will be required. For this reason, interventions should be fi rm, in order to clearly show to the market the willingness and strength of the monetary authority.

It is implicit in the above presentation of the exchange rate policy that the buying and selling fl ows of international currency are manageable. This means that the central bank manage to keep the policy in a sustain-able path. If capital infl ows are massive the cost of sterilization may turn the monetary policy unsustainable. It is important to notice, however, that such a situation might arise as an endogenous consequence of the exchange rate policy itself. Massive capital infl ows may result from an excessively stable nominal exchange rate in the short run, which turns speculative investments in one-way bets. Short-run volatility in the exchange rate increases the uncertainty of speculative investments and thus may reduce capital infl ows, diminishing the amount and cost of interventions. This is the main reason why the exchange rate policy under a SCRER regime should preserve short-run volatility.

However, massive capital infl ows that make the policy unsustainable may occur even when the central bank induces short-run volatility in the exchange rate. This could happen in the context of high liquidity in international capital markets. In this kind of situations, it would make little sense to risk macroeconomic stability in order to preserve the capital account full openness principle. The preservation of the macroeconomic regime requires in this case capital account regulations, intended to restrict capital infl ows and facilitate the management of exchange and monetary policies. There is a menu of measures able to accomplish this function that, even when they do not work perfectly well, evidence suggests that they contribute to soften capital infl ows during booms.20 The need for controls is not permanent; they have to do their job only in a booming phase, and we now know well that booming phases do not last forever.

When there is an excess demand for international currency that turns exchange and monetary policies unmanageable, FX interventions would cause an excessive monetary contraction and the consequent rise in the interest rate would trigger a recession. The defense of some nominal


exchange rate may risk a speculative attack on the central bank reserves. The situation has similarities with a fi xed exchange rate regime crisis. But there is an important diff erence. If there are no fundamental reasons for depreciation – generated, for instance, by expectation of balance of payments defi cit – fi scal and monetary policies are consistent with the targeted real exchange rate, and infl ation is under control, then the macro-economic regime should be preserved. This would only be possible if exchange controls and restrictions on capital outfl ows were imposed. If there are no fundamental reasons inducing the excess demand for inter-national currency, there is no need for the controls and regulations to last for long. As described in Section 9.2, Argentina successfully imposed exchange controls and capital outfl ow regulations in mid 2002, when the run into foreign currency was mainly caused by a self-fulfi lling bubble in the exchange rate. The measures were gradually softened when the buying pressure in the FX market diminished.

In a SCRER macroeconomic regime, monetary policy should not be exclusively focused on infl ation. It is important to emphasize, however, that this regime performs a preventative role with respect to infl ation acceleration. In most developing economies the exchange rate is the main transmission mechanism of monetary impulses to the infl ation rate. The SCRER precisely encourages the central bank to implement monetary policies that avoid excessive fl uctuations in both the nominal and real exchange rates. In contrast, for the same reason, an exclusive infl ation focus of monetary policy generates incentives towards real exchange rate appreciation.

In coordination with the other policies, monetary policy should be managed in order to attain multiple objectives. To manage monetary aggregates or set the interest rate to accomplish this goal, the central bank may have to compensate for the interventions in the FX market. Out of the extreme situations discussed above, this can be done through diff erent instruments. The most common is the sterilization operations. In Section 9.3 we showed that fully sterilized interventions can be carried out in the context of excess supply of foreign exchange and we derived the range of interest rates that make the policy sustainable in time.

Apart from sterilization, there are other instruments at hand to conduct the monetary policy under the SCRER regime. For instance, a central bank in possession of a signifi cant amount of bank debt can manage it as an instrument for monetary control (Lavoie, 2001). Public sector deposits in the central bank can be used in an analogous way. Some prudential regulations can be oriented to the same target, particularly when the problem is to constraint money expansion. The central bank, for instance, can raise the cash requirements of the banking system and thus lower the


banking multiplier. Other prudential regulations can be directly focused on smoothing the selling pressure in the exchange market. For instance, if local banks are not allowed to back credits in domestic currency with liabilities in international currency and credits in international currency are limited, there are fewer incentives to the banks procuring international funding. The existence of public banks with a signifi cant share of the fi nan-cial market can facilitate the monetary management. Public banks can be coordinated in order to help the central bank in both the management of the liquidity and the FX interventions. Through the management of these instruments the central bank should be able to keep money expansion under control.

Finally, fi scal policy should complement monetary policy in attaining infl ation and employment targets in the short run. It should focus on the management of nominal aggregate demand: moderating it in cases of infl ationary pressures and expanding it in the opposite situations. However, it is important to notice that since a SCRER regime is meant to promote development and growth, infl ationary pressures may be more likely than defl ationary ones. Since in this regime monetary policy typi-cally has an expansionary bias through the buying interventions in the FX market, the role of moderating the aggregate demand to avoid infl ation and real exchange rate appreciation would tend to rely on the fi scal policy. Conservative fi scal policy could also be necessary in the SCRER regime to ease the central bank needs to intervene in the FX market. Public sector surpluses may be used to buy part of the excess supply in the FX market. An anti-cyclical fi scal fund intended to perform this role could be a good institution to develop in a SCRER regime.

NOTES

1. The authors would like to thank Nelson Barbosa-Filho, Erinc Yeldan, Jan Kregel and the participants in the workshop on ‘Alternatives to Infl ation Targeting Monetary Policy for Stable and Egalitarian Growth in Developing Countries’ held at CEDES 13-14 May 2005 for their comments to previous versions of this chapter. We also thank Martín Fiszbein for his collaboration. The usual caveats apply.

2. In 1992 the new central bank law slightly relaxed this constraint by setting narrow margins to the possibilities of purchasing public bonds and lending to the commercial banks.

3. For an analysis of the macroeconomic performance during the convertibility period, see Damill and Frenkel (2007).

4. An analysis of the evolution of Argentine debt, default and restructuring can be found in Damill et al. (2005).

5. By mid 2007 energy supply shortages raised concerns about the sustainability of high rates of economic growth. These concerns, however, are not related to the SCRER policy.


6. Later on, the government issued new debt to compensate the banks for the balance sheet eff ect of the asymmetric ‘pesoifi cation’.

7. Figure 9.1 shows a ‘jump’ in the private bank deposit series in January 2002. It refl ects the accounting eff ect of the ‘pesoifi cation’ at 1.40 pesos per dollar of deposits issued in foreign currencies, previously valued at a AR$/$ 1 rate. If we put this mere accounting eff ect aside, it is easy to see the drop in deposits.

8. Exchange rates are defi ned so that a rise in this variable implies a nominal or real depreciation.

9. They included the obligation to surrender the proceeds from exports in the local FX market.

10. For instance, the eff ects of the interest rate through the credit channel are very weak in an economy where the bank credit to private sector remains below 13 percent of GDP as in Argentina.

11. The literature emphasizing the positive eff ects of the SCRER on development increases day by day. See, for instance, Williamson (2003), Rodrik (2005), Dooley et al. (2004), Frenkel (2004), Frenkel and Taylor (2005) and Subramanian (2007).

12. See Hausman et al. (2005) and Prasad et al. (2007).13. See Frenkel and Ros (2006) for an analytical and empirical study of the positive rela-

tionship between real exchange rate and employment in Latin America.14. Evidence regarding short-run indeterminacy of the real exchange rate seems to be

conclusive. Its behavior is almost completely determined by the nominal exchange rate. Although most scholars agree about the existence of an equilibrium real exchange rate in the long run, there is no consensus regarding the factors aff ecting its determination. The purchasing power parity (PPP) is the most accepted hypoth-esis (Taylor and Taylor, 2004). However, evidence regarding the PPP shows time series reverting to their means in very long periods (that is, half life of 3–5 years) and results are highly sensitive to data sets and estimation techniques. On the other hand, mean-reverting time series is no suffi cient condition for the validity of the PPP hypothesis.

15. Another key assumption is that exchange rate expectations are formed with perfect foresight. If departures from the perfect foresight-rational expectation paradigm are considered, the predictions of the trilemma may no longer hold. For a critique of the trilemma in this vein, see Frenkel and Rapetti (2007).

16. See, for instance, Chapter 10 of Dornbusch (1980).17. See, as an example, Canales-Kriljenko (2003).18. The complete model is in Frenkel (2007).19. For a detailed description of a macroeconomic regime proposal with a SCRER as an

intermediate target, see Frenkel (2006).20. See Epstein et al. (2003).

REFERENCES

Balassa, B. (1971), ‘Trade policies in developing countries’, American Economic Review, 61 (2), 178–87.

Blecker, R. (2005), ‘Financial globalization, exchange rates and international trade’, in G. Epstein (ed.), Financialization in the World Economy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp.183–209.

Bofi nger, P. and T. Wollmershäuser (2003), ‘Managed fl oating as a monetary policy strategy’, Economics of Planning, 36 (2), 81–109.

Canales-Kriljenko, J.I. (2003), ‘Foreign exchange intervention in developing and transition economies: results of a survey’, International Monetary Fund working paper no. 03/95.


Damill, M. and R. Frenkel (2007), ‘Argentina: macroeconomic performance and crisis’, in R. Ffrench-Davis, D. Nayyar and J. Stiglitz. (eds), Stabilization Policies for Growth and Development, Initiative for Policy Dialogue, Macroeconomics Policy Task Force, New York, forthcoming.

Damill, M., R. Frenkel and M. Rapetti (2005), ‘The Argentinean debt: history, default and restructuring’, Columbia University Initiative for Policy Dialogue working paper series, New York.

Díaz Alejandro, C. (1979), ‘Algunas vicisitudes históricas de las economías abier-tas en América Latina’, Desarrollo Económico, 19 (74), 147–59.

Dooley, M., D. Folkerts-Landau and P. Garber (2003), ‘An essay on the revived Bretton Woods system’, International Journal of Finance and Economics, 9, 307–13.

Dooley, M., D. Folkerts-Landau and P. Garber (2004), ‘The US current account defi cit and economic development: collateral for a total return swap’, National Bureau Economic Research working paper no. 10727.

Dornbusch, D. (1980), Open Economy Macroeconomics, New York: Basic Books.Epstein, G., I. Grabel and K.S. Jomo (2003), ‘Capital management techniques in

developing countries: an assessment of experiences from the 1990s and lessons for the future’, Political Economy Research Institute, working paper no. 56, University of Massachusetts Amherst.

Feldstein, M. (1999), ‘Self-protection for emerging markets’, National Bureau for Economic Research, working paper no. 6907.

Fischer, S. (2001), ‘Exchange rates regimes: is the bipolar view correct?’, Journal of Economic Perspectives, 15 (2), 3–24.

Frankel, J. and A.K. Rose (1995), ‘Empirical research on nominal exchange rates’, in: G.M. Grossman and K. Rogoff (eds), Handbook of International Economics, vol. III, Amsterdam: Elsevier, pp. 1689–709.

Frenkel, R. (2004), ‘Real exchange rate and employment in Argentina, Brazil, Chile and Mexico’, paper prepared for the G-24, Washington, DC, 24 August.

Frenkel, R. (2006), ‘An alternative to infl ation targeting in Latin America: macroeconomic policies focused on employment’, Journal of Post Keynesian Economics, 28 (4), 573–91.

Frenkel, R. (2007), ‘La sostenibilidad de la política de esterilización’, CEPAL Review, 93 (December), 29–36.

Frenkel, R. and J. Ros (2006), ‘Unemployment and the real exchange rate in Latin America’, World Development, 34 (4), 631–46.

Frenkel, R. and L. Taylor (2005), ‘Real exchange rate, monetary policy, and employment’, paper prepared for the High-Level United Nations Development Conference, 14-15 March, New York.

Frenkel, R. and M. Rapetti (2007), ‘Política cambiaria y monetaria después del colapso de la Convertibilidad’, Ensayos Económicos, 46, 137–66.

Hausman, R., L. Pritchett and D. Rodrik (2005), ‘Growth accelerations’, Journal of Economic Growth, 10 (4), 303–29.

Lavoie, M. (2001), ‘The refl ux mechanism in the open economy’, in L.P. Rochon and M. Vernengo (eds), Credit, Interest Rates and the Open Economy: Essays on Horizontalism, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 215–42.

Polterovich, V. and V. Popov (2002), ‘Accumulation of foreign exchange reserves and long term growth’, New Economic School, working paper, Moscow.


Prasad, E., R. Rajan and A. Subramanian (2007), ‘Foreign capital and economic growth’, Brooking Papers on Economic Activity, (1), 153–20.

Rodrik, D. (forthcoming), ‘Growth strategies’, Brookings Papers on Economic Activity.

Subramanian, A. (2007), ‘Capital account convertibility: a neglected consider-ation’, Economic and Political Weekly, 42 (25), 2413–18.

Taylor, A. and M. Taylor (2004), ‘The purchasing power parity debate’, Journal of Economics Perspectives, 18 (4), 135–58.

Williamson, J. (2003), ‘Exchange rate policy and development’, paper prepared for the project Capital Market Liberalization Task Force Initiative for Policy Dialogue, Barcelona, 2 June, 2003.

203

10. A general equilibrium assessment of twin-targeting in TurkeyCagatay Telli, Ebru Voyvoda and A. Erinç Yeldan1

10.1 INTRODUCTION: MACROECONOMICS OF TWIN-TARGETING IN TURKEY

After a decade of failed reforms and deteriorated macroeconomic per-formance, Turkey entered the millennium under a Staff Monitoring Program signed with the International Monetary Fund (IMF) in 1998, and put into eff ect in December 1999. The program currently sets the macro-economic policy agenda in Turkey and relies mainly on two pillars: (1) fi scal austerity that targets a 6.5 percent surplus for the public sector in its primary budget2 as a ratio to the gross domestic product (GDP); and (2) a contractionary monetary policy (through an independent central bank) that exclusively aims at price stability (via infl ation targeting). Thus, in a nutshell the Turkish government is charged to maintain dual targets: a primary surplus target in fi scal balances (at 6.5 percent to the GDP); and an infl ation targeting central bank whose sole mandate is to maintain price stability and is divorced from all other concerns of macroeconomic aggregates – hence the terms in the title: macroeconomics under twin-targeting.

According to the logic of the program, successful achievement of the fi scal and monetary targets would enhance ‘credibility’ of the Turkish government ensuring reduction in the country risk perception. This would enable reductions in the rate of interest that would then stimulate private consumption and fi xed investments, paving the way to sustained growth. Thus, it is alleged that what is being implemented is actually an expansion-ary program of fi scal contraction.

On the monetary policy front, the Central Bank of Turkey (CBRT) was granted its independence from political authority in October 2001. In what follows, the central bank announced that its sole mandate is to restore and maintain price stability in the domestic markets and that it will follow a disguised infl ation targeting until conditions are ready for full targeting.


Thus, over 2002 and 2003 the CBRT targeted net domestic asset position of the central bank as a prelude to full infl ation targeting. Finally on 1 January 2006 the CBRT announced that it will adopt full-fl edged infl ation targeting.

The purpose of this chapter is to provide an assessment of the key macro-economic developments in Turkey over the post-2001 crisis period and to provide a general equilibrium analysis of the macroeconomic policy alternatives of the twin-targeters. We focus on three sets of issues: fi rst, we study the macroeconomics of the expanded foreign capital infl ows in resolving (temporarily) the macroeconomic impasse between the disinfl a-tion motives of the CBRT and imperatives of debt sustainability and fi scal credibility of the Ministry of Finance. Second, we study the reduction of the central bank’s interest rates. Third, we implement a labor market reform and study the implications of reducing/eliminating payroll taxes (paid by the employers). To these ends we construct a macroeconomic general equilibrium model with a full-fl edged fi nancial sector in tandem with a real sector. Across all policy simulations we exclusively focus on both the fi scal and fi nancial adjustments and study the possible dilemmas of gains in effi ciency in the labor markets versus the loss of fi scal revenues to the state.

Our fi nding is that the current monetary strategy followed by the CBRT that involves a heavy reliance on foreign capital infl ows along with a rela-tively high real rate of interest is eff ective in bringing infl ation down; yet it suff ers from increased cost of interest burden to the public sector and strains fi scal credibility. In contrast, our simulation results suggest that, given the ex ante constraints of the domestic economy in the short run, an alternative heterodox policy of reduction of the central bank interest rate and lowering of the payroll tax burden in labor markets may have strong employment and growth eff ects. The policy also achieves signifi cant gains in fi scal credibility in the short run. Yet it suff ers from increased infl ation-ary pressures in the commodity and the fi nancial markets. Even though observed to prevail at a modest scale in our simulation experiments, the ex ante constraints of maintaining infl ationary expectations may lead to intolerance of the CBRT and render the policy ineff ective. Thus, maintain-ing an integrated and coherent policy framework between the monetary and fi scal authorities is seen of prime importance for the success of the policy formulation at the macro scale.

Our premise in this chapter is that a proper modeling of the general equi-librium linkages between the production-income generation and aggregate demand components across individual sectors as well as responses of the real macro aggregates to fi nancial decisions are essential steps to understand the impact of the current austerity program on the evolution of output, fi scal, fi nancial and external balances, and on employment.

A general equilibrium assessment of twin-targeting in Turkey 205

Accordingly, we develop a computable general equilibrium (CGE) model with a relatively aggregated productive sector, a segmented labor market and a full-blown public sector with a detailed treatment of fi scal balances and fi nancial fl ows.

The current model shares many of the analytical structure of the Agénor et al. (2006) design in the dynamics of fi nancial transactions, especially with respect to formation of expectations and fragility. It is explicitly designed to capture the relevant linkages between the fi scal policy decis-ions, fi nancialization constraints and external balances that we believe are essential to analyse the impact of disinfl ation and fi scal reforms on labor market adjustment and public debt sustainability. We pay particular attention to fi scal issues such as a high degree of debt overhang and fi scal dominance; the link between real and fi nancial sector interactions, and interactions between external (current account) defi cits private saving-investment defi cits and the public (primary balance) surpluses.

We organize the chapter under four sections. First, we provide a broad overview of the recent macroeconomic developments in Turkey in Section 10.2. Here we study, exclusively, the evolution of the key macroeconomic prices such as the exchange rate, the interest rate and price infl ation. Here we also comment on the external balances, the dynamics of external debt, fi scal policy issues and the labor market. In Section 10.3, we introduce and implement our CGE modeling analysis of the alternative policy scen-arios to depict the short-run macroeconomic adjustments of the Turkish economy under the conditionalities of the IMF program targets on primary surplus to GDP ratio and on infl ation rate. Finally, we provide a brief summary with concluding comments in Section 10.4.

10.2 MACROECONOMIC DEVELOPMENTS UNDER IMF’S STAFF MONITORING

The growth path of the Turkish economy over the post-1998 period had been erratic and volatile, mostly subject to the fl ows of hot money. Following the contagion eff ects of the Asian, Russian and Brazilian turmoil, the economy fi rst decelerated in 1998 with a growth rate of 3.1 percent, and then con-tracted in 1999 at the rate of –5.0 percent. The boom of 2000 was followed by the 2001 crisis. The recovery was sharp as the economy has grown at an average rate of 7.1 percent over the 2002–06 period. Price movements were also brought under control through the year and the 12-month average infl ation rate in consumer prices has receded from 45 percent in 2002 to 7.7 percent in 2005, and from 50.1 percent to 5.9 percent in producer prices.

The post-2003 period has also meant a period of acceleration of exports,


where export revenues reached $91.7 billion in 2006. Nevertheless, with the rapid rise of the import bill over the same period, the defi cit in the current account reached $31.7 billion (or about 7.9 percent of the gross national product (GNP) in 2006). The current account defi cit continued to widen in 2007 and reached $34 billion over 12 months cumulative period in the fi rst quarter. On the public sector front one witnesses a very strong fi scal discipline eff ort. The ratio of central government budget defi cit to the GNP was reduced from its peak of 16.2 percent in 2001 to 0.8 percent by 2006. Consequently, the public sector borrowing requirement (PSBR) as a ratio to the GNP fell from 16.1 percent to negative 3 percent, indicat-ing a surplus, in 2006. Table 10.1 documents the main macro indicators of the post-1998 Turkish economy under close IMF supervision.

10.2.1 Macroeconomic Prices and the Monetary Policy

The CBRT initiated an open infl ation targeting framework starting 1 January 2006. The CBRT’s current mandate is to set a ‘point’ target of 5 percent infl ation of the consumer prices. Given internal and external shocks, the CBRT has recognized an internal (of 1 percent) and an external (of 2 percent) ‘uncertainty’ band around the point target. Thus, the CBRT will try to keep the infl ation rate at its point target; however, recognizing a band of maximum 2 percentage points below or above the 5 percent target rate. The CBRT has announced that it will continue to use the overnight interest rates as its main policy tool to reach its target. It is stated explicitly that the ‘sole objective of the CBRT is to provide price stability’, and that all other possible objectives are out of its policy realm.3

Despite the positive achievements on the disinfl ation front, rates of inter-est remained slow to adjust. The real rate of interest remained above 10 percent for much of the post-2001 crisis era, and generated heavy pressures against the fi scal authority in meeting its debt obligations (Table 10.1). The persistence of the real interest rates, on the other hand, had also been condu-cive in attracting heavy fl ows of short-term speculative fi nance capital over 2003 and 2006. This pattern continued into 2007 at an even stronger rate.

Inertia of the real rate of interest is enigmatic from the successful macro economic performance achieved thus far on the fi scal front. The credit interest rate, in particular, had been constrained by a lower bound of 16 percent despite the deceleration of price infl ation. Consequent to the fall in the rate of infl ation, the inertia of credit interest rates translates into increasing real costs of credit.

High rates of interest were conducive in generating a high infl ow of hot money fi nance to the Turkish fi nancial markets. The most direct eff ect of the surge in foreign fi nance capital over this period was felt in the foreign

207

Tab

le 1

0.1

Bas

ic c

hara

cter

istic

s of t

he T

urki

sh e

cono

my

unde

r the

IMF

surv

eilla

nce,

199

8–20

06

Staff

m

oni-

tori

ng

prog

ram

in

itiat

ed

Con

ta-

gion

of

emer

ging

m

arke

t fi n

anci

al

cris

es

IMF-

dire

cted

di

sin-

fl atio

n pr

ogra

m

Fina

ncia

l cr

isis

IMF-

Dir

ecte

d P

ost-

Cri

sis A

djus

tmen

ts

Und

er th

e th

ree-

part

y co

aliti

on

gove

rn-

men

t

Und

er th

e pr

agm

atic

and

w

este

rn-f

rien

dly

Isla

mis

m

of th

e A

KP

1998

1999

2000

2001

2002

2003

2004

2005

2006

Rea

l Rat

e of

Gro

wth

GD

P3.

1−

5.0

7.4

−7.

47.

65.

88.

97.

46.

1C

onsu

mpt

ion

Exp

endi

ture

s

Priv

ate

0.6

−2.

66.

2−

9.2

2.0

6.6

10.1

8.8

5.2

Pu

blic

7.8

6.5

7.1

−8.

65.

4−

2.4

0.5

2.4

9.6

Inve

stm

ent E

xpen

ditu

res

Pr

ivat

e−

8.3

−17

.816

.0−

34.9

−7.

220

.345

.523

.617

.4

Publ

ic13

.9−

8.7

19.6

−22

.014

.5−

11.5

−4.

725

.9−

0.2

Exp

orts

12.0

−7.

119

.27.

411

.016

.012

.58.

58.

5Im

port

s2.

3−

3.7

25.4

−24

.815

.727

.124

.711

.57.

1M

acro

econ

omic

Bal

ance

s (as

ratio

to th

e G

NP

, %)

Agg

rega

te D

omes

tic S

avin

gs22

.721

.218

.217

.519

.219

.320

.217

.116

.6A

ggre

gate

Fix

ed In

vest

men

ts24

.322

.122

.819

.017

.316

.118

.420

.323

.1B

udge

t Bal

ance

−7.

0−

11.6

−10

.9−

16.2

−14

.3−

11.2

−7.

1−

2.0

−0.

8Pu

blic

Sec

tor B

orro

win

g

Req

uire

men

t9.

315

.511

.816

.412

.79.

34.

7−

0.4

−3.

0C

urre

nt A

ccou

nt B

alan

ce1.

0−

0.7

−4.

82.

4−

0.8

−3.

4−

5.2

−6.

2−

7.9

Stoc

k of

For

eign

Deb

t55

.471

.063

.492

.777

.557

.150

.446

.950

.4

208

Tab

le 1

0.1

(con

tinue

d)

Staff

m

oni-

tori

ng

prog

ram

in

itiat

ed

Con

ta-

gion

of

emer

ging

m

arke

t fi n

anci

al

cris

es

IMF-

Dir

ecte

d di

sin-

fl atio

n pr

ogra

m

Fina

ncia

l cr

isis

IMF-

Dir

ecte

d P

ost-

Cri

sis A

djus

tmen

ts

Und

er th

e th

ree-

part

y co

aliti

on

gove

rn-

men

t

Und

er th

e pr

agm

atic

and

w

este

rn-f

rien

dly

Isla

mis

m

of th

e A

KP

1998

1999

2000

2001

2002

2003

2004

2005

2006

Mac

roec

onom

ic P

rice

sR

ate

of C

hang

e of

the

N

omin

al E

xcha

nge

Rat

e (T

L/S

)71

.760

.628

.611

4.2

23.0

−0.

6−

4.9

−5.

76.

9In

fl atio

n (P

PI)

71.8

53.1

51.4

61.6

50.1

25.6

14.6

5.9

9.4

Infl a

tion

(CPI

)84

.664

.854

.954

.444

.925

.310

.67.

79.

6R

eal I

nter

est R

ate

on G

DIs

a29

.536

.84.

531

.89.

115

.413

.110

.47.

9R

eal W

age

Gro

wth

Rat

esb

Pr

ivat

e Se

ctor

−0.

98.

6−

2.6

−14

.4−

5.0

0.5

4.8

1.6

1.9

Pu

blic

Sec

tor

5.5

18.3

15.6

−11

.50.

5−

5.3

4.7

7.9

−3.

0

Not

es:

a.

Defl

ate

d by

the

Prod

ucer

Pric

e In

dex.

b.

Bas

ed o

n re

al w

age

inde

xes (

1997

5 1

00) i

n m

anuf

actu

ring

per h

our e

mpl

oyed

, Tur

ksta

t dat

a.

Sour

ce:

SPO

Mai

n E

cono

mic

Indi

cato

rs; U

nder

secr

eter

iat o

f Tre

asur

y, M

ain

Eco

nom

ic In

dica

tors

; CB

RT

dat

a di

ssem

inat

ion

syst

em.


exchange market. The overabundance of foreign exchange supplied by the foreign fi nancial arbiters seeking positive yields led signifi cant pressures for the Turkish lira to appreciate. As the CBRT has restricted its mon-etary policies only to the control of price infl ation, and left the value of the domestic currency to be determined by the speculative decisions of the market forces, the lira appreciated by as much as 40 percent in real terms against the US dollar and by 25 percent against the euro (in producer price parity conditions, over 2002–06).

The overvaluation of the lira was the most important contributor in reducing the burden of an ever-expanding foreign indebtedness. While the aggregate foreign debt stock has increased from US$113.6 billion in 2001 to US$206.5 billion by the end of 2006, as a ratio to the GNP it has created an illusionary tendency to fall when measured in the overvalued lira units.

10.2.2 Fiscal Policy and Debt Management

The current fi scal policy stance in Turkey relies primarily on expenditure restraint. On the revenue side one witnesses a signifi cant eff ort in raising tax revenues, both in real terms and also as a ratio to the GNP. Much of this eff ort can be explained by the rise in the share of indirect/excise taxes on goods and services (to 21 percent as a ratio to the GNP, or about 70 percent of total tax revenues), while the contribution of direct income taxes to the budgetary revenues are observed to fall especially after 2000.

Data reveal a secular fall in the budget defi cit through the post-2001 crisis adjustments and is now reduced to less than 1 percent to the GNP. As discussed above, much of the aggregate budget expenditures can be explained by the high costs of debt servicing, and the main logic of the current austerity program rested on maintaining the debt turnover via only primary surpluses. As a result, the boundaries of the public space are severely restricted, and all fi scal policies are directed to securing debt ser-vicing at the cost of extraordinary cuts in public consumption and invest-ments. Within total expenditures, public investments’ share has fallen from 12.9 percent in 1990 to 5.1 percent in 2003. As a ratio to the GNP, public investments stand at less than 2 percent currently.

All of these painful adjustments on the fi scal front can be contrasted against the ‘gains’ over the existing debt burden of the public sector. Data from the Ministry of Finance4 reveal that, as a ratio to the GNP, gross public debt of the aggregate public sector has fallen from 68.1 percent in 2000 to 63.1 percent by the end of 2006, a decline of only 5 percentage points. This could have been achieved despite the very rapid rise in the rate of growth of GNP (7.2 percent per annum over the whole period), and the very strict fi scal austerity measures of primary surplus targets (of


6.5 percent to the GNP for 2002 and beyond). Furthermore much of this decline has come only after 2005, and all of it is due to the decline in the ratio of foreign debt to the GNP. As a ratio to the GNP, public external debt declined from 25.2 percent in 2000 to 16.9 percent in 2006; while the domestic debt burden increased from 43.1 percent to 46.2 percent over the same period. It is a clear fact that the illusion of falling foreign indebted-ness is a direct outcome of the real appreciation of the Turkish lira. As the increased external indebtedness of the public sector from $47.6 billion in 2000 to $69.6 billion in 2006, its ratio to the GNP had the eff ect of a fall when denominated in appreciated liras.

In fact, appreciation of the lira disguises much of the fragility associated with both the level and the external debt induced fi nancing of the current account defi cits. A simple purchasing power parity (PPP) ‘correction’ of the real exchange rate, for instance, would increase the burden of external debt to 76.8 percent as a ratio to the GNP in 2005.5 This would bring the debt burden ratio to the 2001 pre-crisis level. Under conditions of the fl oating foreign exchange regime, this observation reveals a persistent fragility for the Turkish external markets, as a possible depreciation of the lira may severely worsen the current account fi nancing possibilities.

10.2.3 Persistent Unemployment and Jobless Growth

Yet the most striking observation on the Turkish labor markets over the post-2001 crisis era is the sluggishly slow performance of employment gen-eration capacity of the economy. Despite the very rapid growth perform-ance across industry and services, employment growth has been meager. This observation, which actually is attributed to many developing econ-omies as well,6 is characterized by the phrase ‘jobless growth’ in the litera-ture. The rate of open unemployment was 6.5 percent in 2000, increased to 10.3 percent in 2002, and remained at that plateau despite the rapid surges in GNP and exports. Open unemployment is a severe problem, in particu-lar, among the young urban labor force reaching 24.5 percent in 2005.

On the other hand, the participation rate fl uctuates around 48 percent to 50 percent, due mostly to the seasonal eff ects. It is known, in general, that the participation rate is less than the EU averages. This low rate is principally due to women choosing to remain outside the labor force, a common feature of Islamic societies, but its recent debacle depends as much on the size of the discouraged workers who had lost their hopes for fi nding jobs. According to Turkish Statistical Institute (Turkstat) data, the excess labor supply (unemployed plus underemployed) is observed to reach 13.6 percent of the labor force by the end of 2006 (Figure 10.1).

Thus, to conclude, two important characteristics of the post-crisis


adjustment path stand out. First, the post-2001 expansion is observed to be concomitant with a deteriorating external disequilibrium, which in turn is the end result of excessive infl ows of speculative fi nance capital. Second, the output growth contrasts with persistent unemployment, warranting the term ‘jobless growth’.

The foregoing facts bring the following tasks to our agenda: (1) What are the viable policy choices in combating unemployment in the short run, and under the conditionalities of the ‘twin targets’? (2) Given our assess-ments of fragility conditions currently prevailing in Turkey, what are the short-run eff ects of a reduction in the interest cost of the central bank credit in terms of output, employment, foreign indebtedness and other macro aggregates?

We now turn to the analytics of general equilibrium with the aid of our CGE model to study these questions.

10.3 COMPUTABLE GENERAL EQUILIBRIUM MODELING ANALYSIS

Given the overview of the recent macroeconomic developments, we now develop a real-fi nancial CGE model for Turkey. In what follows, we provide a bird’s-eye overview of the model, and invite the interested reader to the Political Economy Research Institute website for a full algebraic description.7

0

2

4

6

8

10

12

14

16

18

20

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Une

mpl

oym

ent (

%)

45

46

47

48

49

50

51

52

53

54

Par

ticip

atio

n ra

te (

%)

Total unemployment rate (%)Labor participation rate

Source: Author’s calculation using Turkish Statistical Institute (Turkstat), Household Labor Force Surveys database.

Figure 10.1 Labor participation rate and total unemployment


10.3.1 The Algebraic Structure of the Model and Adjustment Mechanisms

Product marketsThe model is fairly aggregate over its microeconomic structure but accom-modates a relatively detailed treatment of the public accounts, and of real-fi nancial sector linkages. There are four production sectors as agriculture, industry, private services and public services. There is a fi nancial sector with a full-fl edged banking segment, a central bank, enterprises, govern-ment and household portfolio instruments.

Sectoral production is modeled via multilevel functions. At the top level total output is given as a Leontieff specifi cation of value-added and inter-mediate inputs. The value-added in each sector is generated by combining labor, as well as public and private physical capital. At the last stage of this multilevel production a sector-specifi c public capital combines with the composite input under a Cobb-Douglas specifi cation. The composite primary input, in turn, is defi ned to be a combination of private capital and labor aggregate Li through a constant elasticity of substitution (CES) type of production function.

Public capital is assumed to be fi xed and sector-specifi c. Private capital is mobile across sectors and the movement is directed by the diff erence in the diff erentiated private profi t rates. Labor’s wage rate is fi xed in the short run and the labor market clears through quantity adjustments on employment.

Households save a fraction 0 , sP , 1 of their disposable income. The saving rate is considered to be a positive function of the expected real interest rate in domestic currency denominated deposits:

sP 5 sP0 a 1 1 intD

1 1 E [Inf ]bsH

SAV (10.1)

with E[Inf ], the expected infl ation rate and sp0 is a scaling parameter. The

portion of income that is not saved is allocated to consumption and that total fl ow of savings of the household is channeled to the accumulation of household fi nancial wealth.

Private capital investment is assumed to depend on a number of factors. The fi rst is the growth rate of real GDP, which captures the regular accel-erator eff ect. This eff ect is positive. The next one is the negative eff ect of the expected real cost of borrowing from the domestic banks. Specifi cally, private investment demand is represented by:

PK # PINV

NomGDP5 a1 1

DRealGDP21

DRealGDP21bsACCa 1 1 intLD

1 1 E [INF ] b2sINTL (10.2)


where NomGDP and RealGDP are the nominal and real values of the gross domestic product, respectively, valued at market prices.

Financial markets, asset allocation and risk premiaHousehold’s fi nancial wealth is typically allocated to fi ve diff erent cate-gories of assets: domestic money, HD, domestic currency denominated bank deposits held at home, DDH, foreign currency denominated deposits held domestically,8 FDDomH, holdings of government bonds, GDIH and port folio investments abroad, PFIH.9

The household demand function for currency is positively related to consumption and negatively related to expected infl ation and interest on domestic currency denominated deposits, intD. It also depends negatively on the interest on foreign currency denominated deposits, intDF, adjusted for the expected rate of depreciation (1 1 Δeexp):

HD 5H

0PRIVCON qH

CON (1 1 E [Inf ])2qHInf (1 1 intD)2qH

DD

[ (1 1 Deexp) (1 1 intDF ) ]qHDF

(10.3)

Household allocation on domestic versus foreign currency deposits is a function of the interest rate on domestic currency denominated deposits as a ratio to the rate of return on foreign currency denominated deposits held at home. Total portfolio investments of households abroad is taken to be a fi xed fraction of total household fi nancial wealth, and the demand for government bonds by households is regarded as a function of the expected bond interest rate, E[intB].

A crucial decision of the enterprise sector is how to allocate their profi ts between funds to private investment and funds to government bonds. This decision depends on the average profi t rate expected from production activities and, on the other hand, expected returns on government debt instruments:

PK # PINV

DGDIE 5 mEGDI c (1 1 E [intB ])

(1 1 avgRPR21)d2sE

GDI (10.4)

Banks set both deposit and lending interest rates. The deposit rate on domestic currency denominated deposits, intD, is set equal to the borrow-ing rate from the central bank, intR. The deposit rate on foreign currency deposits at home, on the other hand, is set on the basis of the (premium inclusive) marginal cost of borrowing on world capital markets:

(1 1 intDF ) 5 (1 1 intFW ) (10.5)

Following Agénor et al. (2006), the risk-premium inclusive foreign


interest rate is formulated as a function of the (risk-free) world interest rate, intFWRF, and an external risk premium:

(1 1 intFW ) 5 (1 1 intFWRF ) (1 1 riskpr) (10.6)

in which the risk premium is assumed to be a function of total foreign debt to exports ratio:

riskpr 5 contag 1k

2aaForDebt

aEib2

(10.7)

In Equation (10.7) contag is an exogenous coeffi cient used to capture the characteristic changes in the ‘sentiments’ in world capital markets. The bank lending rate, intLD, in the last analysis is defi ned as a weighted average of the cost of borrowing from the central bank and the cost of borrowing from foreign capital markets. It also takes into account the (implicit) cost of holding required reserves.

Public sector, credibility and expectationsSince the government debt instruments constitute a relatively signifi cant share of the assets in the domestic fi nancial markets in Turkey, modeling the interactions between the public sector and the central bank (the so-called ‘fi scal dominance’) is one of the crucial concerns of this study. With a mandated target of the ‘primary surplus-GNP ratio’, the government’s fi scal policy is basically centered around the primary balance. A fi scal defi cit is still realized if interest costs on the outstanding public debt exceeds the primary surplus. The public sector borrowing requirement, PSBR, is fi nanced by either an increase in foreign loans or by issuing bonds. Of crucial importance is the realization of the interest rate on government bonds. The expected rate of return on this instrument is defi ned as:

E [intB ] 5 (1 2 PRdefault) intB (10.8)

where PRdefault denotes the ‘subjective’ probability of default on the current stock of public debt as perceived by the ‘markets’. This variable is set to depend on, among various alternative measures, the current debt stock to tax revenues ratio with a one-period lag:

PRdefault 5 1 2 e2g0 (DomDebtG 1 ForDebtG )

GTaxRev (10.9)

The probability of default, PRdefault, has also a further eff ect on infl ation expectations in such a way that the less the probability of default that


is perceived, the higher the chances for the ‘declared’ infl ation target to materialize. Following Agénor et al. (2006) the expected infl ation rate is formulated as a function of the government’s ‘credibility infi cator’, that is, the inverse of the probability of default, PRdefault, and the targeted rate of infl ation in the previous period:

E [Inf ] 5 (1 2 PRdefault)Inf trgt 1 PRdefaultInf21 (10.10)

Note that, under such a setting, the demand for government bonds is aff ected by the probability of default. Private investors assign a non-zero probability of default in the current period. The expected rate of return will refl ect the probability and will demand compensation in the form of higher nominal interest rates on government bonds. On the other hand, the larger the stock of debt, the higher the probability of default, and the higher the interest rate.

For a given probability of default, a continued increase in the supply of bonds will require an increase in interest rates to evoke investors’ demand. Next an increase in the stock of debt will lead to a rise in the probability of default, which will also rise the prevailing interest rate on government bonds. Such a mechanism in the model tries to capture the structure of government trying to provide a signal of confi dence to the markets under the current measures of the program.

10.3.2 General Equilibrium Analysis of Alternative Policy Environments

Now we utilize our CGE apparatus to provide a general equilibrium analysis of the macroeconomic policy alternatives under twin-targeting. In what follows we will focus on three sets of issues to depict three alternative policy environments: fi rst, we highlight the important role of the expanded foreign capital infl ows in resolving (temporarily) the macroeconomic impasse between the disinfl ation motives of the CBRT and imperatives of debt sustainability and fi scal credibility of the Ministry of Finance. Second, we implement a ‘fi scally benign’ monetary policy of reducing the interest rate charged by the CBRT. Third, we complement the interest rate reduction policy with a labor market reform and study the implications of reducing/eliminating payroll taxes (paid by the employers). In all of the policy simulations we exclusively focus on both the fi scal and fi nancial adjustments, and study the possible dilemmas of gains in effi ciency in the labor markets versus the loss of fi scal revenues to the state. Our simulation experiments are implemented as one-shot, comparative-static exercises. The results are tabulated in Table 10.2. As valid for all types of modeling exercises of the current genre, the simulation results should not be taken as


Table 10.2 Experiment results

Base year (2003) data

EXP-1: Eff ects of increased foreign capital infl ows

EXP-2A: Reduce central bank

interest rate

EXP3: Reduce central

bank interest rate and reduce

payroll taxes

Macroeconomic AggregatesReal GDP (Bill 2003 TL) 369.700 369.765 370.283 375.631Real Private Consumption (Bill

2003 TL) 255.022 263.259 250.255 256.549Real Private Investment (Bill 2003 TL) 66.212 75.138 72.931 74.535Merchandise Imports (Bill US$) 69.378 76.307 70.240 70.910Merchandise Exports (Bill US$) 47.215 43.475 47.866 48.354Current Account Balance (Bill US$) −9.201 −22.491 −9.234 −9.511Unemployment Rate (%) 10.55 10.35 10.69 7.63Average Profi t Rate (%) 16.15 16.20 16.20 16.80As Ratios to the GDPPrivate Consumption 68.98 71.10 67.60 68.80Private Investment 17.91 20.30 19.70 20.00Imports 28.06 29.70 28.39 28.35Exports 19.09 16.92 19.34 19.33Current Account Balance −3.72 −8.75 −3.73 −3.80Financial Rates and PricesInfl ation Rate (CPI) 25.30 18.72 27.72 28.23Expected Infl ation Rate 17.65 18.83 16.58 16.79Expected Depreciation Rate 41.66 41.78 41.56 41.58Realized Depreciation Rate −1.01 −9.52 0.98 0.97Central Bank Interest Rate (intR) 40.27 40.27 20.00 20.00Interest Rate on Domestic Deposits

(intD) 40.27 40.27 20.00 20.00



Interest Rate on Private Domestic Debt

(intLD) 46.50 47.16 37.65 37.65Interest Rate on Government Bonds

(intB) 36.56 60.37 5.49 5.92Expected Interest Rate on Gov. Bonds

(E[intB]) 18.28 25.63 3.13 3.29Risk Premium Incl. Foreign Int. Rate

(intFW) 3.41 4.09 3.24 3.23Fragility IndicatorsRatio of Gov. Dom. Debt to Tax

Revenues 156.98 196.26 122.34 127.76Government’s Fiscal Credibility Index 0.50 0.42 0.57 0.56Perceived Probability of Default on Gov.

Debt 0.50 0.58 0.43 0.44Ratio of Foreign Debt to Central

Bank Foreign Reserves 235.87 264.88 223.08 223.77Ratio of Foreign Debt to GDP 48.93 52.89 46.25 45.92Risk Premium on Private Foreign

Borrowing 1.39 2.10 1.20 1.20Currency Substitution (FX deposits/Tot.

Deposits) 90.86 90.99 93.85 93.85Monetary Aggregates (as ratio to the GDP)Money Demand by HH 2.64 2.75 2.68 2.68Domestic Deposits of HH 21.27 20.76 20.75 20.60FX Deposits of HH 19.32 18.89 19.47 19.33Central Bank Foreign Reserves 0.21 0.20 0.21 0.21


a ‘forecast’ of the future, but rather ought to be regarded as quantitative insights on the relevant macroeconomic outcomes of alternative policy environments.

EXP-1: Macroeconomics of foreign capital infl owsThe post-2001 Turkish economy has benefi ted from the recent surge of fi nancial fl ows quite extensively. The increased buoyancy in the global fi nancial markets led both to a fall in the rates of interest in the global markets and also served for provision of expanded liquidity, propelling consumption and investment expenditures. Mostly driven by the private portfolio fl ows, the net annual infl ow of fi nance capital into the ‘new emerg-ing market economies’ totaled $456 billion in 2005, before receding to $406 billion in 2006.10 These magnitudes exceeded the previous peaks hit in the global fi nancial markets before the eruption of the 1997 Asian crisis.

As outlined in Section 10.2, Turkey too had been one of the major benefi ciaries of this fi nancial glut. Balance of payments data indicate that the fi nance account has depicted a net surplus of $103.3 billion over the ‘AKP (Justice and Development Party) period’, 2003 through 2006.


Base year (2003) data

EXP-1: Eff ects of increased foreign capital infl ows

EXP-2A: Reduce central bank

interest rate

EXP3: Reduce central

bank interest rate and reduce

payroll taxes

Fiscal Results (as ratios to the GDP)Government Aggregate Revenues 39.13 39.44 38.76 37.03Government Tax Revenues 33.35 33.79 33.15 31.48Government Consumption Exp. 11.95 12.05 11.84 11.31Government Investment Exp. 4.36 4.53 4.16 3.20Government Interest Exp. 16.60 42.40 3.99 4.13PSBR 14.28 28.01 1.00 1.09Primary Balance 6.50 6.50 6.50 6.50


About half of this sum ($151.2 billion) was due to credit fi nancing of the banking sector and the non-bank enterprises, while a third ($32.8 billion) originated from non-residents’ portfolio investments in Turkey. It is also observed that 64 percent of the total infl ows (net fi nancial fl ows plus errors and omissions) was used for fi nancing the current account defi cit which had totaled $71.8 billion over the same period; while 36 percent had been used for reserve accumulation of the CBRT.

In this fi rst policy experiment we fi rst study the macroeconomic adjust-ment mechanisms against this continued infl ow of fi nance capital into the Turkish economy. To this end, we exogenously increase the total infl ow of portfolio investments from abroad, PFIROW, by a factor of $30 billion (roughly the realized net cumulative fl ow over 2003–06). No change in the CBRT’s current monetary policy stance is envisaged with respect to the level of interest rates and/or exchange rate administration.11 The exchange rate was left to full fl oat to be determined by the free play of foreign exchange market transactors. Our results are tabulated under the column EXP-1 in Table 10.2.

The immediate eff ects of the increased infl ow of foreign capital are felt in the currency markets. The exchange rate appreciates by 9.5 percent and cost savings on the import side leads to a fall in the infl ation rate to 18.7 percent, from 25.3 percent. Appreciation of the exchange rate leads to a rise in imports and the current account defi cit widens to increase by about four-fold to reach $22.5 billion in 2003 prices. As a ratio to the GDP, it increases to 8.7 percent from its base value of 3.7 percent. The domestic counterpart of the widening current account defi cits is the expansion of private investment (by 2.4 percentage points as a ratio to GDP) and of private consumption (by 2.1 percentage points as a ratio to GDP). The monetary base expands by 20 percent and serves for the liquidity require-ments of this expansion.

The aforementioned expansion of the economy is limited, however, only to the private sector. Given the fi scal constraint on the primary surplus target, the government’s room for maneuver is limited on the expenditure side. This constraint becomes even more binding as the domestic economy continues to operate with a signifi cantly high real interest burden. It has to be remembered that a critical feature of the simulated policy environ-ment is that the central bank continues to maintain its interest rate at the already high level. As the economy disinfl ates, however, the real cost of credit increases even further. The interest cost on the government’s debt instruments, in particular, expands to 60 percent from 36.3 percent. The government’s interest expenditures as a ratio to the GDP rises to 25 percent and that of the public sector borrowing requirement increases to 28 percent. Increased interest expenditures lead to a widening of the


fi scal defi cit. Consequently there is a worsening of the fi scal credibility of the government. The credibility index falls to 0.43 from its value of 0.50. The loss in fi scal credibility leads to a rise in the subjective probability of default as perceived by the private markets.

Thus, the main result of the scenario unveils an important dilemma for the post-2001 Turkish economy: a policy of maintaining high real rates of interest along with heavy reliance on foreign fi nance proves to be disinfl a-tionary, and it also has expansionary eff ects on the private sector. The net result is that the CBRT achieves relative success in controlling infl ationary pressures. In the meantime, however, the increased debt burden strains the already fragile fi scal balances and results in further loss of fi scal credibility. The predicament of controlling price infl ation via high real interest rates and enhanced foreign capital infl ows, on the one hand, and the imperatives of debt turnover and fi scal credibility, on the other, remains unresolved. This impasse is further accentuated with the rise of foreign indebtedness and consequent external fragility. As the results of EXP-1 suggest, stock of external debt increases both as a ratio to the GDP (from 48.9 percent to 52.9 percent) and to the foreign reserves of the central bank (from 235 percent to 264 percent). As a result of these adverse developments on external fragil-ity, the risk premium on the Turkish liabilities in the world markets increase by 6 percentage points. Clearly, the realized quandary is not to be resolved by reliance on foreign capital and tight macro management alone, and postponing the necessary adjustments on the domestic front simply lead to culminated pressures on the fi scal side as well as on the external balances.

EXP-2: Reduce the central bank interest rateGiven the rather high costs of disinfl ation in terms of high fi scal and external fragility in the previous experiment, the natural policy question is to study the eff ects of a reduction in the interest rate charged by the central bank. In general, the burden of the interest rates has a signifi cant contractionary eff ect on the Turkish fi nancial sector. The cost of central bank liquidity is held responsible by many scholars for the external debt cycle and intensi-fi ed infl ows of speculative short-term fi nance into the Turkish economy. There is a general call for a reduction of the central bank’s rate of interest to escape the trap of speculative infl ows of fi nance leading to appreciation and more infl ows, with the consequent widening of the current account defi cit and the rise of external indebtedness. Thus, in this experiment we reduce the central bank interest rate by half. Again, as above, no further change is envisaged, when experimenting with this simulation, in the policy instru-ments or in the parameterization of the exogenously set variables.

Note that given the algebraic characterization of the loanable funds market, the central bank interest rate has a direct eff ect on the determination


of the deposit interest rate of the domestic banks. Thus, intD is reduced by the same magnitude (50 percent) immediately. With declining rates of interest on deposits the banks fi nd it possible to lower their credit interest rate charged to the enterprises (intLD falls by 8 percentage points). Private investment expenditures increase by 1.9 percentage points as a ratio to the GDP.

The distinguishing adjustment mechanism at work is the expenditure switching of private consumption with investment. As private consump-tion expenditures are reduced by 2 percentage points as a ratio to the GDP, domestic savings could be generated to sustain the expansion in investments; thus, the net eff ect on the external balances remains modest. In other words, with a given level of domestic disposable income, an expanded level of investment demand could have been sustained. Thus, the current account balance is aff ected only marginally and the (nominal) adjustments in the exchange rate are revealed to be modest as well, with a realized depreciation of less than 1 percent.

What brings forth this adjustment in the private expenditure patterns is the infl ation tax. Price infl ation accelerates by 2.7 percentage points, and causes a downward shift of aggregate private consumption. The accelera-tion of infl ation further strengthens the decline of the real interest cost for all agents of the economy, private and public. In fact, the relative buoyancy of the economy, along with the decline of the public interest expenditures, leads to an improvement in the fi scal balances. Of particular interest is the decline in the public sector borrowing requirement to less than 1 percent of the GDP, and the increase of the credibility index by 7 percentage points.

It is not clear, however, whether the central bank would be willing to tolerate the resultant increase of the infl ation rate, which turns out to be the crucial adjusting variable to bring forth the warranted adjustments in the real economy. Yet a further issue is that even though the fi scal results of the policy are observed to be benign, the employment gains remain quite meager. Unemployment rate persists at above the 10.5 percent level, and it is this problem we aim to tackle in the next experiment.

EXP-3: Complement central bank interest rate reduction with labor tax reformIn this experiment we continue on the policy environment of the previous experiment and complement the central bank’s interest reduction strategy with a labor tax reform. Keeping the central bank’s interest rate at its reduced level (at half of the base run value), we now implement a further reduction on taxes paid by employers of labor.

Turkey has one of the highest tax burdens on the labor markets. Employer-paid social security contributions averaged about 36 percent


of total labor costs during 1996–2000; it has been argued that these high social security taxes create strong disincentives to job creation. More gen-erally, many observers have called for a thorough overhaul of Turkey’s social insurance system. Ercan and Tansel (2006) too state that both the red tape and non-wage labor costs are higher in Turkey relative to, for instance, OECD averages. The authors consider the high tax burden on employment and high social security contributions among the institutional factors that contribute to the high level of unemployment and high level of undeclared work. Tunali (2003) indicates that employee contribution to the social security system can be as high as 15 percent while an employer in a typical risk occupation contributes as much as 22.5 percent.

Thus in this experiment we study the implications of lowering the payroll tax paid by the employers on employment, production and fi scal balances. Maintaining the central bank rate of interest at half of its base run value, we reduce the payroll tax by half, from its base rate of 19 percent. The lower tax revenues are not compensated by any other taxes. The results of the experiment are depicted under column EXP-3 in Table 10.2.

Clearly, the most important variable of this experiment is its eff ects on unemployment rate and the fi scal balances. Unemployment rate falls by around 3 percentage points, and the real GDP expands by 1.7 percent upon impact. We fi nd, however, that the main adjustment falls on public investments and then on the price infl ation. The fi rst outcome is the direct result of the fi scal administration under the current austerity program. The logic of the fi scal balances is that, given the tax revenues and interest costs, the public sector is to maintain a primary surplus (of 6.5 percent) as a ratio to the GDP. Once this constraint is met the rest of the public expenditures are calculated. Thus, within the context of our experiment, as tax revenues are curtailed, the government fi nds it necessary to adjust public investments downwards. As a percentage of GDP, public invest-ments are observed to fall to 3.2 percent from its base value of 4.4 percent (a signifi cantly low rate itself).

The eff ect on fi scal accounts is also emphasized in the ratio of govern-ment tax revenues against public debt stock. The aggregate tax revenues fall by almost 2 percentage points; yet, given the cost savings on the interest expenditures, the overall solvency of the public sector remains improved. Thus, the lower interest rate policy is enacted here as an import-ant component of the labor tax reform policy. With a reduced interest burden over the public sector, the CBRT facilitates the fi scal authority to alleviate pressures on the fi scal balances that would have emerged as a result of reduced labor tax revenues. With accelerated growth in GDP and lower interest costs, the fi scal balances improve, with consequent gains in fi scal credibility.12


At the outset, the trade-off s as suggested by the simulation exercise under EXP-3 seem modest and not severely binding: at a loss of 2.9 percentage point increase of the infl ation rate (from 25.3 percent to 28.2 percent), the gains in fi scal credibility and employment are found to be relatively robust. Given that under the macroeconomic adjustments of the experiment the external balances were not strained any further, equilib-rium in the foreign exchange market seems to be maintained, as well.

10.4 CONCLUDING COMMENTS AND POLICY DISCUSSION

In this chapter, we reported on the current state of the macroeconomic policy environment in the Turkish economy throughout the 2000s and studied the general equilibrium eff ects of two widely discussed policy changes in the current context: reduce payroll taxes and reduce the central bank interest rate. The current IMF-led austerity program operates with a dual targeting regime: a primary surplus target in fi scal balances (at 6.5 percent to the GDP); and an infl ation targeting central bank whose sole mandate is to maintain price stability. Accordingly both policy questions are analysed within the constraints of the aforementioned dual targets set as outer conditionalities of Turkish macroeconomic decision making.

Our policy experiments reveal that the current monetary strategy fol-lowed by the CBRT that involves heavy reliance on foreign capital infl ows along with a relatively high real rate of interest, is eff ective in bringing infl ation down, yet it suff ers from increased cost of interest burden to the public sector and strains fi scal credibility. It also leads to excessive foreign indebtedness with increased external fragility. In the medium to long run the increased fi scal and external fragilities along with a persistent and high unemployment manifest a severe impasse, whose resolution will likely lead to onerous adjustments in the labor markets and the real sector. Against this background, we utilized the CGE model to search for applicable alter-native policy regimes starting from the immediate short run. Our results indicate that a heterodox policy of (1) reducing the central bank’s interest rates along with (2) lowering the (payroll) tax burden in the labor markets off ers a viable environment in the short run, with accelerated growth and improved employment outcomes. The fi rst arm of the policy, viz. reduc-tion of the CBRT interest rate, is important to facilitate the improvement in fi scal balance (and fi scal credibility) at a time when tax monies from labor taxes are expected to be reduced. The second critical element is low-ering of the tax burden on employers. With lower payroll taxes levied on employment in production, the employers are led to increase employment


demand (and also most probably be more willing to employ ‘formal’ labor, and reduce the unrecorded activities along with informalization of the labor markets; issues that our model is not well-equipped to address).

With increased credibility of the public sector and lower rates of unemployment, the returns to the heterodox policy reform agenda are quite benign. However, all these come at visible opportunity costs; in particular on the infl ation side. As lower interest rates boost domestic investment expenditures and the domestic economic activity is revived due to expanded employment, infl ationary pressures accumulate in the commodity and fi nancial markets. It is not so clear at the outset how toler-ant would the CBRT be to the acceleration in infl ation. Even though our results are quite modest on the pace of both realized and expected rates of infl ation, it is clearly an important constraint that merits close observation in the Turkish macroeconomic environment. It is, in fact, mainly for this reason that we maintain some of the key features of the current austerity program with respect to expectations management in the short run. Of particular importance among these is the signaling eff ect of the primary surplus target. The policy environment of EXP-3 sets the fi scal balances with the programmed target of 6.5 percent primary surplus ratio to the GNP, rather than proposing a drastic break away from it. In fact, with a proper emphasis on dynamics, a direct case can clearly be proposed to stimulate domestic investment expenditures with a policy of lower interest rates, and advocating a fi scal policy of high public investments towards enhancing human capital formation and social infrastructure. Rather than cutting public investments on health, education and social infrastructure, a case can be forwarded to disregard the rising public sector borrowing requirement to GDP ratio in the short run, and implement a fi scal policy to maintain a level of household income capable of addressing the tasks of accumulating human capital. Yet, given the short-run framework of our current modeling framework, we choose to abstain from making ad hoc statements regarding the dynamic consequences of such a policy environ-ment, an issue that had been dealt with elsewhere more eff ectively (see, for example, Gibson, 2005; ISSA, 2006; Voyvoda, 2003; Voyvoda and Yeldan, 2005).

Above all, our simulation experiments clearly underscore the importance of maintaining an integrated and coherent policy framework between the monetary and fi scal authorities. Given the acuteness of the perceived dilemmas on disinfl ation and fi scal credibility, the resolution of the current impasse will surely necessitate a more tolerant view over the programmed targets (on both infl ation and the primary surplus ratio) as well as a coher-ent and a mutually supportive macro policy design. Furthermore, there is a clear case for a acute need to design viable policies to diminish the


exposure of the domestic economy (in particular of the fi nancial markets) to short-term, speculative foreign capital. This, in turn, may necessitate implementation of capital management techniques to gear infl ows towards longer maturities.

NOTES

1. Author names are in alphabetical order and do not necessarily refl ect authorship seniority. We are indebted to Korkut Boratav, Yilmaz Akyüz, Jerry Epstein, Bill Gibson and to the members of the Independent Social Scientists’ Alliance for their valu-able comments and suggestions on previous versions of the chapter. Previous versions of the chapter were presented at the Istanbul Conference of the EcoMod (June 2005); the 9th Congress of the Turkish Social Sciences Association (December 2005, Ankara); the Ankara congress of the Turkish Economics Association (September 2006); and seminars at Bilkent, METU, Bogazici, Utah, Massachusetts-Amherst, Connecticut and the central bank of Turkey. Research for this chapter was completed when Yeldan was a visiting Fulbright scholar at the University of Massachusetts-Amherst for which he acknowledges the generous support of the J. William Fulbright Foreign Scholarship Board and the hospitality of the Political Economy Research Institute at the University of Massachusetts-Amherst. Needless to mention, the views expressed in the chapter are solely those of the authors and do not implicate in any way the institutions mentioned above.

2. That is, balance on non-interest expenditures and aggregate public revenues. The primary surplus target of the central administration budget was set 5 percent to the gross national product.

3. Further institutional details of the central bank’s infl ation targeting framework can be found in the December 2005 document, ‘General framework of infl ation targeting regime and monetary and exchange rate policy for 2006’, accessed December 2006 at: www.tcmb.gov.tr/yeni/announce/2005/ANO2005-45.pdf.

4. http://www.maliye.gov.tr. 5. Measured in 2002 producer prices. If the PPP-correction is calculated in 2000 prices, the

revised debt to GNP ratio reaches to 82.3 percent. 6. See, for example, UNCTAD, Trade and Development Report (2002 and 2003), New

York and Geneva. 7. http://www.peri.umass.edu/Alternatives-to.382.0.html. 8. By allowing households to hold foreign currency denominated deposits in the domestic

banking system, we try to represent the high level of dollarized liabilities in the Turkish fi nancial system (see Table 10.1).

9. Both residents’ portfolio investments abroad, PFIH, and non-residents’ portfolio investments at home, PFIROW, are incorporated in the model in order to capture any real-economy eff ects of these ‘speculative’ means, which we believe are important in understanding the growth pattern of the Turkish economy in the last decade.

10. See, for example, Institute for International Economics, http://www.iie.com.11. It has to be noted, as a reminder, that the current rate of interest set by the CBRT is

already signifi cantly high in real terms. Maintaining high real rates of interest was but one of the discretionary measures of the CBRT in an attempt to reduce infl ation by curtailing domestic demand expansion, as well as to sustain the infl ow of foreign capital to cover the widening current account defi cit.

12. Note, however, that this comparison is valid against the base run. One witnesses a slight loss in fi scal credibility relative to EXP-2.


REFERENCES

Agénor, P.R., H.T. Jensen, M. Verghis and E. Yeldan (2006), ‘Disinfl ation, fi scal sustainability, and labor market adjustment in Turkey’, in Richard Agénor, A. Izquierdo and H.T. Jensen (eds), Adjustment Policies, Poverty and Unemployment: The IMMPA Framework, Oxford: Blackwell Publishing, Chapter 7, pp. 383–456.

Ercan, H. and A. Tansel (2006), ‘How to approach the challenge of reconciling labor fl exibility with job security and social cohesion in Turkey’, in Reconciling Labour Flexibility with Social Cohesion: Facing the Challenge, Strasbourg: Council of Europe Publishing.

Gibson, B. (2005), ‘The transition to a globalized economy: poverty, human capital and the informal sector in a structuralist CGE model’, Journal of Development Economics, 78, 60–94.

ISSA (Independent Social Scientists Alliance) (2006), Turkey and the IMF: Macroeconomic Policy, Patterns of Growth and Persistent Fragilities, Penang, Malaysia: Third World Development Network.

Tunali, İ. (2003), ‘Background study on the labour market and employment in Turkey’, paper prepared for the European Training Foundation, Turin, Italy, June.

Voyvoda E. (2003), ‘Alternatives in debt management: investigation of Turkish debt in an overlapping generations general equilibrium framework’, unpub-lished PhD thesis, Bilkent University.

Voyvoda E. and E. Yeldan (2005), ‘IMF programs, fi scal policy and growth: investigation of macroeconomic alternatives in an OLG model of growth for Turkey’, Comparative Economic Studies, 47, 41–79.

World Bank (2000), ‘Turkey – country economic memorandum – structural reforms for sustainable growth’, vols. I and II, report no. 20657TU, September, Washington, DC.

227

11. Employment targeting central bank policy: a policy proposal for South AfricaGerald Epstein1

11.1 INTRODUCTION

The story of South Africa’s struggle with apartheid is a dramatic and heroic one, a story that has culminated in a hugely successful politi-cal transition from a minority dominated, authoritarian and repressive government organized along racial lines before 1994, to an inclusive, democratic government, in which the majority rules, subject to important protections for minority rights.

The story of South Africa’s economic transition, on the other hand, is not such a rosy one. Income and wealth are still highly unequally distrib-uted, largely along ethnic lines with the white population still command-ing the bulk of the national income, and still controlling the bulk of the national wealth. Economic growth has been moderate during most of the period since the creation of the new South Africa in 1994, running at an average of 3.4 percent and unemployment rates are hovering somewhere between 25 percent and 36 percent depending on exactly how one counts. Among the black population, in 2006 unemployment was over 30 percent, while among the white population it was less than 5 percent.

Such problematic economic trends are in no small part due to the highly unequal economy inherited from the past. But some of the diffi culties with the economic transition have been due to the ideologies and the policies, including macroeconomic policies, chosen by the new government in the last decade or so. The government has adopted many pages from the neoliberal play book, including fi nancial liberalization, capital account liberalization, austere fi scal policy and, most relevant to this chapter, formal infl ation targeting. These policies have achieved certain gains, but have done little to reduce unemployment and generate more economic equality.

Many in South Africa are looking for some alternatives to these policies.


While the government is still strongly committed to infl ation targeting, recognition of its limitations are becoming more widespread.

In this chapter I present employment targeting (ET) as an alternative to infl ation targeting monetary policy framework for South Africa. It is a framework for monetary policy which attempts to incorporate some of the advantages normally claimed for a targeting framework – namely enhancing transparency and accountability – while focusing the goals of monetary policy more directly on critical macroeconomic problems facing the South African economy, namely, employment. Of course, no govern-ment, including that of South Africa, can ignore the infl ationary impacts of policy, and any ET approach will have to include a goal of stabilizing infl ation at sustainable levels.2

As will be seen as the argument develops, reorienting monetary policy toward generating employment will require that monetary policy rede-velop and utilize a multiplicity of monetary policy and credit tools, including tools for credit allocation and capital management. While these policies had been in the toolkit of many central banks for decades, they have recently fallen out of favor, or have been made impotent by fi nancial liberalization or legal changes (Epstein, 2007). So one will fi nd a bit of ‘going back to the future’ in some of the proposals found in this chapter. But, as there is no simple return to the past, the ET monetary policy frame-work for South Africa takes into account important changes at both the national and international levels, especially in the operations of fi nancial markets, which must be accommodated in any policy framework.

11.2 INFLATION TARGETING IN SOUTH AFRICA IN MACROECONOMIC CONTEXT

At the time of the 1994 victory in the struggle against apartheid a major debate ensued over the future direction of economic policy, which culmi-nated in the adoption of the Department of Finance’s macroeconomic strategy, the ‘Growth, Employment and Redistribution’ (GEAR) policy. GEAR was similar to the neoliberal macroeconomic frameworks devel-oped by the International Monetary Fund (IMF), and it is based on similar premises: the key to achieving employment growth and a rapid increase in living standards is to win the confi dence of both domestic and foreign investors, by engaging in fi scal austerity and fi nancial liberaliz-ation, among other policies.

With respect to monetary policy, more specifi cally, soon after being elected in 1994, the South African government decided that price stabil-ity should be the central concern of monetary policy. In 1998 the Reserve

A policy proposal for South Africa 229

Bank adopted, for the fi rst time, an informal infl ation target range of 1–5 percent. Then in February 2002 the Reserve Bank adopted a formal infl a-tion targeting approach. Under infl ation targeting the Ministry of Finance establishes the target range and the reserve bank chooses the instruments to achieve the target. It chose a target of 3–6 percent rate of increase of CPI, which is a consumer price index excluding interest costs on mortgage rates and the Reserve Bank chose the repo rate as the main monetary policy instrument.

Looking back over the initial years, however, one can see that the GEAR policies were implemented with mixed results (Table 11.1).

The GEAR policy ‘delivered on its promises’ in the early years with respect to fi scal defi cits, infl ation and the current account, but failed in terms of economic growth and employment generation.

Real GDP growth has been higher since that time but still not large enough to substantially reduce the rate of unemployment. Figure 11.1 shows the rate of real GDP growth before and after the transition to demo cratic rule. While GDP growth has gone up since the confl ict-ridden and sanctions-laden period between 1984 and 1994, the 3.45 percent per year average growth rate is still relatively modest, though in the most recent years, largely due to the global commodity boom, growth rates have been above 4.5 percent.

Given this relatively low per-capita growth in real GDP, it should not be surprising that employment performance has been fairly weak.

Table 11.2 shows the offi cial unemployment rate broken down by popu-lation group in 2001 and 2007. The contrast among the population groups is striking with African unemployment, by the offi cial measure showing 26.8 percent unemployment in 2007, in contrast to the overall white rate of 5.8 percent. Both rates have come down somewhat since 2001 due again to the higher than average rate of economic growth during the latter years

Table 11.1 Comparison of GEAR projections and actual outcomes (1996–2000 averages)

GEAR projections Actual outcomes

GDP Growth (%) 4.2 2.8Infl ation (CPI) (%) 8.2 6.7Current Account Defi cit (% GDP) 2.4 1.1Employment Growth (%) 2.9 0.7Non-Gold Exports 8.4 7.1Fiscal Defi cit (% GDP) 3.7 3.2

Source: Du Plessis and Smit (2005).


of that period. This shows that in all likelihood higher economic growth does, in fact, lead to lower unemployment rates among all population groups.

Table 11.3 shows the offi cial unemployment rate in September 2001 and September 2007, broken down by gender and population group. These

Table 11.2 Offi cial unemployment, September 2001 and September 2007 (percentage of labor force)

September 2001 September 2007

African 35.7 26.8Coloured 21.2 20.6Indian 18.8 8.2White 5.8 3.8Average 29.4 22.7

Source: Statistics South Africa.

–3

–2

–1

0

1

2

Per

cent

3

4

5

6

84 86 88 90 92 94 96 98 00 02 04 06

After democratic transition

Average growth: 1.03 Average growth: 3.45

Source: Reserve Bank of South Africa.

Figure 11.1 Real GDP growth


show that unemployment rates among women are higher than among men for all population groups, but they too have gone down somewhat in response to higher economic growth rates.

These data refl ect the offi cial unemployment statistics. Table 11.4 presents data that include ‘discouraged workers’ and therefore may be a more accurate measure of unemployment. They indicate that the ‘unof-fi cial’ unemployment rate might be at 34 percent overall and over 35 percent for women in 2007.

As mentioned earlier, since 2002, if not before, the Reserve Bank of South Africa has followed an infl ation targeting regime for monetary policy; the policy has been directed, most recently toward keep CPI

Table 11.3 Offi cial unemployment by gender and ethnic group, September 2001 and September 2007 (percentage of labor force)

September 2001 September 2007

MaleBlack African 31.5 23.3Coloured 19.5 20.0Indian/Asian 15.7 7.4White 4.7 3.5Average 25.8 19.8FemaleBlack African 40.7 30.9Coloured 23.1 21.3Indian/Asian 23.5 10.2White 7.4 4.2Average 33.8 26.1

Source: Labour Force Survey, March 2006, March 2007 (Pretoria: Statistics South Africa, September 2006 and 2007).

Table 11.4 Offi cial unemployment plus discouraged workers, September 2001 and 2007 (percentage of labor force)

2001 2007

Male 33.8 31.8Female 47.0 34.9Total 40.0 34.0

Source: Author’s calculations from Labor Force Survey, March 2006 and March 2007 (Pretoria: Statistics South Africa).


infl ation within a target range of 3–6 percent. Indeed, infl ation has been in that range since mid 2003 or so (Figure 11.2).

Whether it has caused the decline is more diffi cult to answer. As Ball and Sheridan (2003) point out, however, infl ation fell in many countries over the last decade, independently of whether the countries were adopt-ing infl ation targeting or not. What is clear is that this infl ation targeting regime has led for most of the period to very high rates of real interest rates which, in turn, have contributed to relatively modest economic growth and, until the recent period of the global commodity boom, to anemic growth in employment.

11.3 AN EMPLOYMENT TARGETING FRAMEWORK FOR SOUTH AFRICA

In light of the problems with the infl ation targeting regime in South Africa, and especially given South Africa’s major problems with unemployment and underemployment, an alternative framework for monetary policy is

2

4

6

8

Per

cent

10

12

2001 2002 2003 2004 2005 2006

CPI inflation(Monthly, on an annual basis)

Inflation targetrange

Source: Reserve Bank of South Africa.

Figure 11.2 CPI infl ation


badly needed. Here I propose a ‘real targeting’ framework for monetary policy, and then describe a specifi c type of real targeting framework that applies to South Africa, namely an ET framework for monetary policy.

11.3.1 A Real Targeting Framework for Monetary Policy

A real targeting framework for monetary policy should adhere to the fol-lowing principles:

Context-appropriate monetary policy. Central bank policy goals ●

and operating procedures must be based on the structure and needs of South Africa: no generic, one-size fi ts all approach is likely to be appropriate to every situation.Real economy-oriented monetary policy. Policy makers should recog- ●

nize that very high rates of infl ation can have signifi cant costs, but that short of that, policy must also be oriented toward promoting invest-ment, raising employment growth and reducing unemployment.Transparency and accountability. Taking a leaf from the targeting ●

approach, central banks should be made more accountable to the public by making their objectives and approaches more transparent. They should tell the public what their targets for monetary policy are, describe the economic assumptions underlying their plans to reach those targets and, if they do not reach them, explain why while describing their plans for achieving them in the next period. And, most importantly, the goals of the central bank should be deter-mined by a democratic process.Policy fl exibility. A fundamental fact is that there is a great deal of ●

uncertainty concerning the underlying structure of the economy and about the nature of national and international shocks at any particular time. Hence, adherence to any target has to be somewhat fl exible.Suffi cient tools to reach the targets. Monetary policy has been ●

stripped by fi nancial liberalization and neoliberal conceptions of policy tools needed to achieve a multiplicity of targets: credit allo-cation techniques, which used to be very important components of central bank policy in many countries, have been systematically eliminated, capital controls (capital management techniques) like-wise, which in the past have been part of the arsenal of monetary policy, have been dramatically liberalized. These techniques need to be revitalized and modernized so that, where and when appropriate, they can be used as instruments to help central banks reach their goals.


Supporting institutions. Central bank policy is no panacea. Other ●

important supporting institutions are also required, including strong tax institutions to enable the government to raise the revenue it needs to fund important public investments, and public fi nancial institutions to channel credit in support of productive investment.

In addition to the direct advantages of a real targeting framework, there are several extremely important indirect advantages of the ET framework which themselves will contribute in a crucial way to the framework’s success. These include the accumulation of new knowledge about the con-nections between monetary policy and employment, and the implementa-tion of new policies to generate more employment and economic growth. Both of these positive outcomes will be facilitated by focusing the atten-tion of the central bank, with its enormous human and fi nancial resources, on the key issue of employment growth.

Making employment growth a target of monetary policy might also induce a profound shift in the attitude and the activities of the central bank. It might begin to assign its economists to study the relationships between monetary policy and employment growth; to study what mon-etary policy instruments are best used to achieve employment growth; and to organize conferences on employment growth and monetary policy. It might even give promotions and more resources to members of its staff that make breakthroughs in the understanding of these connections. In the case of South Africa, it will lead the Reserve Bank to link up with others outside the bank who have knowledge and experience with respect to employment generation and its relationship to fi nancial variables. It could also lead the bank to design new programs to help it reach its targets.

11.4 EMPLOYMENT TARGETING MONETARY POLICY IN SOUTH AFRICA

In this section I describe the outlines of an ET policy for the Reserve Bank of South Africa. This policy has been developed in the context of an overall ET program for the South African economy developed by economists at the Political Economy Research Institute.3 This larger ET program includes fi scal stimulus, public credit allocation and develop-ment banking, capital management techniques, mechanisms of infl ation control, tax reforms, including mechanisms such as an enhanced securities transactions tax to raise more revenue to fi nance employment policies, and


other sectoral policies, for example, anti-trust and competition policy to correct infrastructure bottlenecks preventing more growth and employ-ment generation.

In this chapter I focus on the role of the Reserve Bank and monetary policy within this ET framework.

11.4.1 The Role of Monetary Policy and the Reserve Bank in this ET Program

The basic structure of the proposed central policy is as follows: the Reserve Bank would set its interest rates to achieve an overall real growth rate con-sistent with the plan which has an employment target at its core. As part of its mandate, the Reserve Bank would try to reach an infl ation constraint that is mutually decided upon as part of the overall program. In addition, the Reserve Bank would manage some of the credit allocation programs that are part of the targeted components of the overall ET macroeconomic policy. Finally, the Reserve Bank would manage the capital account as needed to maintain the exchange rate and exchange rate stability needed to implement the program.

If the ET policy is to be fully successful, in the long run it will help con-siderably if the Reserve Bank makes certain institutional commitments. These might include some of the following. The Reserve Bank will launch a set of research programs both within the bank and outside to improve understanding of the relationship between monetary policy tools and employment growth, both in the formal sector and informal sector. The Reserve Bank will work with fi nancial institutions, including development banks and major commercial banks, as well as smaller fi nancial institu-tions, to develop instruments and programs to facilitate the allocation of credit for eff ective employment generating activities. If the rates of economic growth achieved through more expansionary monetary policy do not generate as much employment as projected, the Reserve Bank is committed to fi nding new policies and mechanisms, along with the rest of the government, to achieve these targets.

11.4.2 Targets and Instruments and All That

In the terminology of targets and instruments used in the monetary policy literature, in this framework the ultimate target of monetary policy is employment growth, while the intermediate targets are real GDP growth, infl ation and exchange rates. The instruments are short-term interest rates (the repo rate in the case of current practice in South Africa), capital management techniques and credit allocation policies.


11.5 SIMULATING THE IMPACT OF MONETARY POLICY: A VAR-BASED SIMULATION MODEL

In this section I present a simple exercise to show the impact of an expan-sionary monetary policy as part of an ET framework. Following this section, I consider the other instruments and intermediate targets to be utilized by the Reserve Bank in its ET policy.

11.5.1 Monetary Policy Experiments

In this section I present some simulations based on a series of simple struc-tural vector auto-regression (VAR) models I built of the South African economy. These simulations are designed to estimate the impacts of employment targeted monetary policy on real GDP growth – a key deter-minant of employment generation – and on infl ation and exchange rate stability which are crucial other variables of likely concern to the central bank.

VAR’s-based simulation models are widely used in monetary policy analysis (see, for example, Aron and Muellbauer, 2002; Bernanke et al., 1997; Gallardo and Ros, 2008). This approach uses a minimum number of assumptions about the structure of the economy to estimate a small, sim-plifi ed model of the macroeconomy. Based on that estimation, we simulate the model with changes in various policy tools and estimate the impact of those changes in policy on the economy. I do this by estimating the diff er-ence between the ‘baseline’ path of economic variables, such as infl ation and economic growth, and the path these variables take under the new (hypothetical) settings of the monetary policy tools. The diff erence between these is taken to be the ‘impact’ of the policy change on the economy.

It is important to point out that, despite their widespread use, these VAR models have many drawbacks. They are highly simplifi ed models of very complex economies; and in situations where there are either short data series, or a large amount of structural change, these models may not be able to form good estimates of economic relationships or forecast with a great deal of accuracy. Still, for our purposes, we hope that these results can give us some ball-park idea of the impact of the approach I am proposing.

This analysis addresses the impact of changes in the South African Reserve Bank monetary policy’s rule on exchange rates, infl ation and economic growth. The goal is to estimate the quantitative impact of dif-ferent interest rate settings on economic growth, infl ation and exchange rate variability, and indirectly, then, on employment. It incorporates some important exogenous variables as well.


Before describing the estimates and the simulations, I fi rst discuss our data variables and sources.

11.5.2 Data and Preliminary Statistical Tests

All data are either originally quarterly data, or have been transformed into quarterly data from monthly data.

Prime rate: the ‘prime rate’ is the ‘prime lending rate’. It tracks ●

closely the changes in the repurchase (or ‘repo’) rate, which is the main tool of monetary policy. The reason we use the prime rate rather than the repo rate is that it has a longer data series.4

Exchange rate: the nominal rate relative to the US dollar. Our ●

measure is the four quarter rate of change of the rand, measured so that an increase means an increase in the rate of depreciation of the rand.5

Infl ation rate: we use the CPI variable, as calculated by Aron and ●

Mullbauer (2002). This is the consumer price level, excluding mort-gage interest. We use the four quarter rate of infl ation.6

GDP growth: the four quarter rate of growth of real GDP. Source: ●

South African Reserve Bank.7

US Tbill: the three-month US Treasury bill rate. ● 8

Private credit: the four quarter rate of growth of private credit ●

created by all monetary institutions; monthly data transformed to quarterly data. Source: South African Reserve Bank.9

Terms of trade: four quarter change in the terms of trade. Source: ●

South African Reserve Bank.10

11.5.3 Stationarity Tests on Variables

All variables were tested for stationarity using the Augmented Dickey-Fuller tests. The results of these tests are presented in Table 11.5.

For the estimates that follow, we report on results using the growth variable (of real GDP). We have done the estimates with de-trended GDP growth as well and the results are roughly the same. We report on the non-de-trended variable because of greater ease of interpretation.

11.5.4 VAR Estimates of Monetary Policy

I estimate a VAR model with four endogenous variables (prime rate, exchange rate change, infl ation and growth) and one exogenous variable, the US Treasury bill rate (US Tbill), over the period 1989, fi rst quarter


(1989q1) to 2004, fourth quarter (2004q4), with four quarter lags. I choose 1989q1 as a starting date because of concerns of signifi cant structural breaks at that time. I then estimated the impulse response function using the Choleski de-composition, with the following ordering (prime, change in exchange rate, infl ation, growth). I tried the results with diff erent order-ings and the results did not seem very sensitive to these changes. A positive shock in the prime reduces economic growth, has a modest, cyclical impact on infl ation and is associated with increased variability in the exchange rate.

Figure 11.3 presents the impulse response functions from this model.Note that in this VAR estimate, an increase in the prime rate has a

stable impact on infl ation, but it generates an increase in the infl ation rate. This probably refl ects the so-called ‘cost channel’ or ‘Patman Eff ect’, named after US Congressman Wright Patman, who warned that interest rate increases were infl ationary.

5.5 Simulation Results

The next step is to transform this VAR model into a simulation model to be used for monetary policy experiments. The goal of the simulation exercise is to estimate the impacts of a medium- to long-term decline in the prime lending rate on exchange rate variability, infl ation and economic growth. The basic idea is to estimate the likely impacts of a looser monetary policy by the Reserve Bank which attempts to expand real GDP growth in order to generate more employment, as part of an ET policy.

In order to estimate these impacts, I undertook the following steps.I fi rst solved the estimated VAR model using a dynamic simulation,

over the period 1994q1–2004q4 and call the results of this dynamic simulation the ‘baseline’ results. These baseline estimates are then used in combination with data generated by ‘policy experiments’ to estimate

Table 11.5 Stationarity tests

Variable Stationary (signifi cance level)

Intercept Intercept and Trend

Exchange Rate Change Yes (5%) Yes NoCredit Change Yes (5%) Yes NoGrowth Yes (1%) No YesInfl ation Yes I (1) (1%) Yes NoPrime Yes I (1) (1%) Yes NoUS Tbill Yes I (1) (1%) Yes


the impacts of changing the prime lending rate in the steps described below.

As seen in Figure 11.4, the baseline estimates, when compared to the actual data, capture basic trends in the data, for the most part, but they do not do a very good job of tracking all the turning points in the data. Unfortunately, this is not an uncommon problem in such models which have a limited number of degrees of freedom.

The next step is to create alternative scenarios in which the prime lending rate is adjusted to refl ect an employment-oriented monetary policy. Here I report on one scenario: lowering the prime rate so that it remains 4 percentage points below the baseline and keeping it there for fi ve years. Table 11.6 presents the numerical summary of the main eff ects of this policy experiment.

Figure 11.5 describes the impacts by showing the diff erence from the baseline. The results show that GDP growth goes up on average by about

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10

Accumulated Response ofPrime to Prime

–30

–20

–10

0

10

20

30

1 2 3 4 5 6 7 8 9 10

Accumulated Response of Exchange Rate to Prime

–4

–3

–2

–1

0

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10

Accumulated Response ofInflation to Prime

–5

–4

–3

–2

–1

0

1

1 2 3 4 5 6 7 8 9 10

Accumulated Response ofGrowth to Prime

Figure 11.3 Accumulated response to Cholesky one SD innovations ± 2 SE


6 percent for the period 2001–04 (about 0.5 on average for the whole fi ve-year period), infl ation increases by about 1 percentage point on average, while the exchange rate changes become more variable.

Table 11.6 summarizes these results. In this simulation on average economic growth is raised by 0.6 of 1 percent, with only modest increases in infl ation and exchange rate instability. The exchange rate instability

–40

–30

–20

–10

0

10

20

30

40

94 95 96 97 98 99 00 01 02 03 04

Exchange rate changeExchange rate (baseline)

Exchange rate change

0

1

2

3

4

5

6

94 95 96 97 98 99 00 01 02 03 04

Growth

ActualGrowth (baseline)

3

4

5

6

7

8

9

10

11

94 95 96 97 98 99 00 01 02 03 04

Inflation

8

12

16

20

24

28

94 95 96 97 98 99 00 01 02 03 04

Prime rate

Prime ratePrime rate (baseline)

ActualInflation (baseline)

0

1

2

3

4

5

6

7

94 95 96 97 98 99 00 01 02 03 04

US Tbill

ActualBaseline

Figure 11.4 Baseline estimate versus actuals 1994Q1–2004q4


created is relatively small given the swings in exchange rates that South Africa has experienced in the last decade. And the increase in infl ation stays well within the single digits. This increase in the rate of economic growth is itself modest, but accumulated over time, it could have a signifi -cant impact on employment growth.

It should be obvious that, by itself, lowering interest rates will not ‘solve’ South Africa’s unemployment problem. The Reserve Bank, along with other economic institutions, will need to participate in more aggressive and creative initiatives to enhance employment opportunities.

11.6 CREDIT ALLOCATION AND CAPITAL MANAGEMENT POLICIES TO SUPPORT EMPLOYMENT GENERATION

11.6.1 Credit Allocation Policies

As the empirical exercise in the previous section suggests, manipulation of interest rates will certainly have some positive impact on employment in South Africa, but ultimately, other tools will be needed to encourage employment at suffi cient levels. Credit allocation tools can be an important set of policies to contribute to employment growth in South Africa if they are properly implemented and monitored. While some of this operation might occur outside of the purview of the South African Reserve Bank, the bank can and should play a signifi cant role in this area. (See Pollin et al. (2006) for much more detail on credit allocation tools in South Africa.)

In doing so, the Reserve Bank has many examples to draw on both from its own history and from the central banks of many highly success-ful developed and developing countries, including South Korea, Japan, France and the Taiwan Province of China and the People’s Republic of China (Amsden, 2001; Epstein, 2007). Still, experiences are not simply transferable in a ‘turnkey’ way, and even the earlier experience of directed

Table 11.6 Impact of 4 percent point decline in prime rate

2000 2001 2002 2003 2004

Exchange Rate Change 6.16 5.74 −1.91 0.35 1.29Growth 0.22 0.79 0.44 0.61 0.55Infl ation 0.12 1.29 0.96 0.76 0.85Prime Rate −4.00 −4.00 −4.00 −4.00 −4.00US Tbill 0.00 0.00 0.00 0.00 0.00


credited under the apartheid system is not something the current regime would want to replicate. So, the Reserve Bank, while learning from past experience, will need to develop credit allocation mechanisms to fi t the current situation. These could include concessionary loans and increased capitalization for development banks, loan guarantees and asset-based reserve requirements.

–4

–2

0

2

4

6

8

10

12

2000 2001 2002 2003 20040.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

2000 2001 2002 2003 2004

Exchange rate change Growth

–0.4

0.0

0.4

0.8

1.2

1.6

2000 2001 2002 2003 2004–4.2

–4.1

–4.0

–3.9

–3.8

–3.7

2000 2001 2002 2003 2004

Inflation Prime rate

–1.0

–0.5

0.0

0.5

1.0

2000 2001 2002 2003 2004

US Tbill

Figure 11.5 Impact of 4 percent point drop in prime relative to baseline, 2000Q1–2004Q4


To avoid corruption and misuse of these tools, careful monitoring and sharp incentives need to be put in place.

11.6.2 Capital Management Techniques

Another vulnerability of these programs is that, in a fi nancially integrated economy, fi nancial arbitrage and capital fl ight might undermine them. As just discussed, an expansionary monetary policy might induce unstable exchange rates and overshooting as well. Capital controls – or capital management techniques – have been widely used to manage exchange rates and create running room for monetary policy (see, for example, Epstein et al., 2005).

South Africa has a long history of capital management techniques, including exchange and capital controls, dating back to at least 1939 (see Bruce-Brand (2002) for a brief history and overview). The current govern-ing legislation was set out in 1961. Their application has ebbed and fl owed over the years. With the reintegration of South Africa into the world fol-lowing the abolishment of the apartheid government in 1994, after consid-erable discussion, the government decided to progressively liberalize the capital and exchange controls.

Capital management techniques can help countries achieve more autonomy in the making of monetary policy, including reducing the variability and misalignment of the exchange rates, though there is still a great deal of debate on this point. A large literature has attempted to assess the ‘eff ectiveness’ of capital controls (see, for example, Lee and Jayadev (2005), Henry (2006) and Epstein et al. (2005), among many others, for surveys). Studies which have analysed the ‘eff ectiveness’ of capital controls have looked primarily at two things: fi rst, at their ability to maintain a ‘wedge’ between domestic interest rates and foreign rates; or, second, at their ability to avoid ‘currency crises’, that is, either large-scale changes in exchange rates or the forcing of countries off of a peg through reserve loss; more recent studies have looked at the ability of controls (especially on infl ows) to prevent overvalued currencies and currency instability and the ability of controls to provide autonomy for central bank expansion.

Many studies fi nd that capital controls do seem to be eff ective in insulat-ing domestic interest rates and exchange rates from international factors, but usually only to a moderate degree, especially if the time period is a long one; and only if the overall set of macroeconomic policies are intern-ally consistent. More recent evidence suggests, however, that capital con-trols can help to avoid fi nancial instability is stronger (Epstein et al., 2005). Countries that had controls on outfl ows and/or infl ows were more likely


to be able to avoid the contagion from the ‘Asian fi nancial crisis’. In the long literature on controls on infl ows for Chile, there is evidence that Chile was able to avoid an overvalued exchange rate and tilt toward a longer maturity structure of borrowing. In the case of Malaysia (Kaplan and Rodrik, 2002) controls on outfl ows temporarily allowed more expansion-ary policy. Hence, the evidence can be read that, properly implemented, controls can reduce instability and somewhat enhance macroeconomic autonomy, at least to a limited extent. South Africa has a lot of experience in implementing capital controls and should build on that experience to tailor these capital management techniques to modern circumstances.

11.7 CONCLUSION

This chapter has tried to develop an ET monetary policy framework for South Africa. I have emphasized that while monetary policy by itself cannot solve the unemployment problem in South Africa, it must make a contribution to doing so, and that simply stabilizing infl ation will not do the trick. An ET framework will work best if it is part of an overall ET macroeconomic strategy and if the central bank, as an institution, is com-mitted to cooperating with the government to implement such a policy. I have also argued that the central bank itself must bring to bear a coordi-nated set of tools, including interest rate policy, credit allocation policy, capital management techniques, and institutional development, to make an ET approach eff ective.

NOTES

1. Thanks to K.S. Jomo Terry McKinley, Butch Montes and Jose Antonio Ocampo for their encouragement, to my colleagues at PERI, Bob Pollin, James Heintz and Leonce Ndikumana for their indispensable contributions on our larger project on employment targeting in South Africa and to members of the ‘Alternatives to Infl ation Targeting’ group for their suggestions. A fi nal thanks to my project co-organizer, Erinc Yeldan for his contributions. All errors, of course, are mine.

2. See Robert Pollin et al. (2006) for more details on a general macroeconomic ET frame-work for South Africa of which this central bank policy can be seen as a part.

3. Ibid. 4. International Financial Statistics (IFS), lending rate. 5. IFS. 6. http://www.csae.ox.ac.uk/resprogs/smmsae/datasets.html. 7. http://www.reservebank.co.za/. 8. IFS. 9. http://www.reservebank.co.za/.10. Ibid.


REFERENCES

Amsden, A. (2001), The Rise of the Rest: Challenges to the West from Late-Industrializing Economies, Oxford: Oxford University Press.

Aron, J. and J. Muellbauer (2002), ‘Interest rate eff ects on output: evidence from a GDP forecasting model for South Africa’, IMF Staff Papers, 49 (November), 185–213, accessed at (www.csae.ox.ac.uk/conferences/2002-UPaGiSSA/papers/Aron-csae2002.pdf); non-technical summary at http://www.csae.ox.ac.uk/res-progs/smmsae/nontechs/nontech07.html.

Ball, L. and N. Sheridan (2003), ‘Does infl ation targeting matter?’, International Monetary Fund working paper no. 03/129.

Bernanke, B. et al. (1997), ‘Systematic monetary policy and the eff ects of oil price shocks’, Brookings Papers on Economic Activity, 1, 91–157.

Bernanke, B.S., T. Laubach, A.S. Posen and F.S. Mishkin (1999), Infl ation Targeting: Lessons from the International Experience, Princeton, NJ: Princeton University Press.

Bruce-Brand, A.M. (2002), ‘Overview of exchange controls in South Africa’, state-ment to the Commision of Inquiry into the Rapid Depreciation of the Exchange Rate of the Rand, Reserve Bank of South Africa.

Dooley, M.P. (1995), ‘A survey of academic literature on controls over internat-ional capital transactions’, National Bureau for Economic Research working paper no. 5352.

Du Plessis, S. and B. Smit (2005), ‘Economic growth in South Africa since 1994’, conference paper for the Economic Policy under Democracy: A 10 year Review Conference, Stellonbosch University, 28-29 October.

Epstein, G. (2007), ‘Central banks as agents of development’, in Ha-Joon Chang (ed.), Institutional Change and Economic Development, Helsinki: United Nations University Press, accessed at www.wider.unu.edu/publications/working-papers/research-papers/2006/on-GB/rp2006-54/_fi les/78091779622373109/default/rp2006-54.pdf.

Epstein, G., I. Grabel, K.S. Jomo (2005), ‘Capital management techniques in developing countries: an assessment of experiences from the 1990s and lessons for the future’, in Gerald Epstein (ed.), Capital Flight and Capital Controls in Developing Countries, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 301–33.

Farrell, G.N. (2001), ‘Capital controls and the volatility of South African exchange rates’, South African Reserve Bank occasional paper no. 15, July.

Galindo, L.M. and J. Ros (2008), ‘Alternatives to infl ation targeting in Mexico’, International Review of Applied Economics, forthcoming.

IMF (2003), South Africa: Selected Issues, International Monetary Fund country report no. 03/18, January.

International Monetary Fund (IMF) (2004a), International Financial Statistics, CD ROM.

IMF (2004b), South Africa: Selected Issues, International Monetary Fund country report no. 04/379, December.

Jayadev, A. and K.K. Lee (2004), ‘The eff ects of capital account liberalization on growth and the labor share of income: reviewing and extending the cross-country evidence’, in Gerald Epstein (ed), Capital Flight and Capital Controls in Developing Countries, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 15–57.


Kaplan, E. and D. Rodrik (2002), ‘Did the Malaysian capital controls work?’, in Sebastian Edwards and Jeff rey A. Frankel (eds), Preventing Currency Crises in Emerging Markets, Chicago, IL: University of Chicago Press, pp. 393–441.

Lee, K. and A. Jayadev (2005), ‘Capital account liberalization, growth and the labour share of income: reviewing and extending the cross-country evidence’, in G. Epstein (ed.), Capital Flight and Capital Controls in Developing Countries, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 15–57.

Pollin, R.N., G. Epstein, J. Heintz and L. Ndikumana (2006), ‘An employment-targeted economic program for South Africa’, United Nations Development Program Poverty Center, Brazil.


APPENDIX 11.A1 VARIABLES, DEFINITIONS AND SOURCES FOR VAR MODEL

Variable Defi nition Units Frequency Source Properties

RGDP Growth

Growth rate of real GDP, annualized

Percent Quarterly South African Reserve Bank

Stationary

Prime Prime lending rate


Stationary

Infl ation The infl ation rate based on CPI metropolitan, total


First diff erence is stationary

Exchange Rate

Nominal rand/US dollar exchange rate (an increase is depreciation)

Rand per US dollar

Quarterly IMF, Inter-national Financial Statistics (IFS)

Log fi rst diff erence is stationary

248

12. Infl ation targeting and the design of monetary policy in IndiaRaghbendra Jha1

12.1 INTRODUCTION

With around 20 central banks adopting it as their basic monetary policy framework, infl ation targeting (IT) has emerged as an important mon-etary policy framework. Over time IT has become more fl exible in its inter-pretation of target and permitted other goals to be included in the basic framework. Central banks have enhanced their communication with their respective publics about their targets and modus operandi.

Some authors have argued that for transition economies under going sustained fi nancial liberalization and integration in world fi nancial markets, IT is an attractive monetary policy framework. Consequently there is pressure for such economies to adopt IT.

This chapter evaluates the case for IT in India. It begins in Section 12.2 by stating the objectives of monetary policy, especially that infl ation control cannot be an exclusive concern of monetary policy in a country with mass poverty. An evaluation of the rationale for IT and nuances of implementation are spelt out in Section 12.3. Section 12.4 provides some evidence on the eff ects of IT in developed and transition econo-mies. Section 12.5 discusses India’s experience with using nominal targets whereas Section 12.6 discusses some recent developments in Indian mon-etary policy. Section 12.7 reviews some reasons why India is not ready for IT. Section 12.8 shows that even if the Reserve Bank of India (RBI) – India’s central bank – wanted to, it could not pursue IT since the short-term interest rate (the principal policy tool used to aff ect infl ation in coun-tries using IT) does not have signifi cant eff ects on infl ation. Section 12.9 concludes and sketches the contours of an alternative to an IT policy. A central feature of this policy is the use of capital controls. Appendix 12.A1 revisits the role that such capital controls played in ensuring that India did not get severely aff ected by the East Asian fi nancial crisis of the late 1990s, even though India’s macroeconomic fundamentals at that time were com-parable to those of countries that did experience such a crisis.

Infl ation targeting and the design of monetary policy in India 249

12.2 THE OBJECTIVES OF MONETARY POLICY IN INDIA

An overriding short-term concern of monetary policy is stabilization of the price level. However, India has had a long-standing problem of poverty and its alleviation has to be the cornerstone of the success of any policy, including monetary policy. Higher economic growth, along with some supporting redistributive measures has a crucial role in reducing poverty, even more than governance. Dollar and Kraay (2001) show for a broad cross-section of countries including India, that the incomes of the poorest 20 percent of the population rise in proportion to average income.2

China is an important example of the poverty reducing eff ects of eco-nomic growth (Table 12.1).

For almost three decades Chinese per capita GDP has grown at more than 8 percent per annum. Poverty has declined at an average of 9 240 000 persons per year.

In contrast India’s growth and poverty reduction record has been less spectacular (Table 12.2).

Because of lower growth the reduction in poverty in India has been far lower than in China notwithstanding the fact that inequality in China has grown more sharply than in India (Jha, 2004). India’s national poverty headcount ratio fell only by about 12 percentage points between 1951–52 to 19973 and the rate of poverty reduction was higher in the 1980s than in the reform period, post-1991, indicating that the quality of growth in the

Table 12.1 Growth and poverty alleviation in China

Year Annual poverty

reduction announced

by the government

(10 thousand)

Average annual growth

rate of GDP per capita (%)

Average annual

growth rate of farmers’

consumption level (%)

Average annual

growth rate of farmers’ net income

per capita (%)

1978–85 1786 8.3 10.0 15.11985–90 800 6.2 2.5 3.01990–97 500 9.9 8.0 5.01997–02 436 7.7 3.4 3.81978–02 924 8.1 5.6 7.2

Source: Chinese Statistical Abstract, various issues.


1980s was diff erent from that in the 1990s, so that even with lower growth greater reduction in poverty could take place in the 1980s. This trend has recently been reversed. Jha et al. (2006) report that the rural headcount ratio, which is usually higher than the urban headcount ratio, was 20.6 percent in 2004 compared to higher than 26 percent in 1999. Hence there exists scope for poverty reduction through higher growth in India.

Furthermore, unemployment has been rising in the post-reform period. In the organized sector employment barely changed between 1991 and 2001; from 1997 it actually fell. Total employment (organized and unor-ganized) is growing at about 1 percent per annum – half the projected growth rate of the labor force. Consequently, the unemployment rate for males in the rural sector rose from 5.6 percent in 1993–94 to 9 percent in 2004. In addition India also has substantial underemployment.

If India is to reduce poverty rapidly, its GDP must grow at 8 percent or more on a sustained basis. Kelkar (2004) argues that growth could accelerate essentially because of a broad series of fi nancial sector reforms, increased globalization and widening and deepening of product and fi nan-cial markets, and benefi cial structural changes – particularly on the supply side. In addition to its ‘surplus labor’ India is set to reap an important demographic dividend as the proportion of population in the working age group (15–64 age bracket) is expected to climb from 60.9 percent in 2000 to over 66 percent by 2030. The labor force is less nutritionally deprived and increasingly literate. Economic theory and international experience indi-cate that this could lead to sharp rises in labor productivity and an upward shift in the trend rate of growth. However, this labor force must be produc-tively employed for these productivity gains to be realized (Jha, 2005).

Low interest rates (to enhance investment) and a slightly undervalued exchange rate with low volatility (to boost exports) are critical to sustain-ing high growth rates. An appropriate monetary policy must address these requirements.

Table 12.2 GDP and per capita GDP growth in India

Period GDP growth (%)

Aggregate (average annual)

Per capita (average annual)

1972–82 3.5 1.21982–92 5.2 3.01992–2002 6.0 3.9

Source: Kelkar (2004).


12.3 EVALUATING THE RATIONALE FOR INFLATION TARGETING

The time inconsistency literature argues that a discretionary policy setting leads to higher infl ation without any gains in output (Barro and Gordon, 1983; Kydland and Prescott, 1977). The high welfare cost of infl ation raises the appeal of rules-based policy. Under a rules-based regime central banks set explicit values for intermediate targets, which they can control, and which are strongly related to the ultimate goals of monetary policy (viz. stabilization of output and infl ation), which cannot be directly controlled.

In recent times emerging market economies, including India, have experimented with three nominal targets: exchange rate, money supply growth and infl ation.4 The relative advantages/disadvantages of exchange rate and money growth targeting are portrayed in Table 12.3.

Bernanke and Mishkin (1997) argue that IT, in contrast to exchange rate targeting but like monetary targeting, enables monetary policy to focus on domestic considerations and to respond to shocks to the domestic economy.5 IT, like exchange targeting and unlike monetary targeting, has the advantage that people easily understand it. Since the central bank has a numerical infl ation target, the chances of slipping into a time inconsist-ency trap are reduced. Svensson (1999), Bernanke and Mishkin (1997) and White (2004) argue that IT is ‘decision making under discretion’ with central banks following a targeting rule which sets interest rates to reduce the deviation between conditional infl ation forecast (the interme-diate target of monetary policy) and the infl ation target to zero over the horizon.

In an emerging market economy such as India the problems of mon-etary management, in general and infl ation control, in particular, get compounded by low policy credibility (Calvo and Mishkin, 2003). This can lead to sudden stops of capital infl ows, making such countries prone to fi nancial crises. This increases the appeal of rules-based monetary regimes (like IT). Taylor (2002) argues that rules-based policies enhance the anticipation eff ects of monetary policy. Given less developed fi nancial markets such anticipatory eff ects are likely to be lower. Yet monetary policy could still have considerable eff ects through movements of wages and property prices. With an IT regime shocks from the monetary regime may be lower.

However, this rationale for IT is incomplete. For instance, Kirsanova et al. (2006) show, in the context of the UK, that an IT regime in itself cannot be considered to be an optimal monetary policy framework and optimal policy response must include terms of trade or exchange rate terms – an


Table 12.3 Advantages and disadvantages of the nominal anchors of exchange rate targeting and monetary targeting

Anchor: Exchange rate targeting

Advantages 1. This fi xes the infl ation rate for internationally traded goods and thus directly contributes to keeping infl ation under control. It is especially useful for sharply reducing infl ation in emerging market economies

2. If the exchange rate peg is credible, it anchors infl ation expectations to the infl ation rate in the anchor country to whose currency it is pegged

3. An exchange rate provides an automatic rule for the conduct of monetary policy that avoids the time-inconsistency problem

4. An exchange rate is simple and direct and, therefore, is well understood by the public

Disadvantages 1. An exchange rate target leads to loss of independent monetary policy (Obstfeld and Rogoff , 1996). Hence the ability of the monetary authorities to respond to shocks is compromised

2. The exchange rate peg may persuade large-scale foreign borrowing. In the case of emerging market economies such loans are invariably denominated in foreign currency. Large accumulation of such loans may lead to a crisis. In most developed countries a devaluation may have little direct eff ect on the balance sheets (since debts are denominated in home currency) but not so in emerging market economies since debts are denominated in foreign currency

3. Bernanke and Mishkin (1997) argue that exchange rate pegs can lead to fi nancial fragility

4. Although exchange rate targeting may be initially successful in bringing infl ation down a successful speculative attack can lead to a resurgence of infl ation

Anchor: Monetary targeting

Advantages 1. An advantage over exchange rate targeting is that monetary targeting enables a central bank to adjust its monetary policy to cope with domestic considerations

2. A monetary target is easily understood by the public – but not as well as an exchange rate target

3. Monetary targets have the advantage of being able to promote almost immediate accountability for monetary policy

Disadvantages 1. Typically the link between money growth and infl ation is subject to long and uncertain lags

2. The demand for money may not be stable, there may be instability of velocity and the money supply may not be controllable (Jha and Rath, 2003). This is especially true of broad monetary targets such as M2 or M3 and less so of narrow money


argument likely to hold with greater certainty for developing and tran-sition countries such as India.

Even if IT leads to price stability it may not guarantee fi nancial sta-bility. The RBI (2004) notes that the 1990s – a decade of relative price stability – witnessed a number of episodes of fi nancial instability indi-cating that price stability is not suffi cient for fi nancial stability. Large movements in capital fl ows and exchange rates aff ect the conduct of monetary policy continually. Bernanke and Gertler (2001), Bernanke (2003), Bean (2003) and Filrado (2004) argue that even if price stability does not imply fi nancial stability in the short run, price stability does not endanger fi nancial stability and price stability, and fi nancial stability would reinforce each other in the long run. In an economy with relative price stability the interest rate should not remain passive (as it would in an IT regime) in the presence of a sudden capital outfl ow. Thus the RBI (2004) notes ‘[while] there is very little disagreement over the fact that price stability should remain a key objective of monetary policy, reserva-tions persist about adopting it as the sole objective of monetary policy’ (p. 56).6

Empirical evidence suggests that in emerging market economies central bank interest rates react more strongly to changes in the exchange rate rather than changes in the infl ation rate or output gap (Mohanty and Klau, 2004). Hence, at this time, the standard tool to target infl ation – short-term interest rate – is unlikely to be particularly useful.7

12.3.1 The Mechanics of Infl ation Targeting

The modus operandi of a typical IT regime is as follows. The central bank revises its infl ation and output forecast at a frequency determined by monetary policy committee meetings using updated information. If the conditional infl ation forecast is higher than the target, the central bank raises the interest rate to minimize such deviation by the end of the targeting horizon, and vice versa. Households and fi rms then decide upon their consumption and investment plans. Blinder (1998) and Taylor (1993, 2002) argue that this is close to what many policy makers do in practice.

It has become common to compare ex post the actual setting of policy rates by central banks with what would have been predicted by the Taylor rule which suggests that (short-term) interest rates (the federal funds rate in the USA or the bank rate in India) should be changed in response to deviation of infl ation from a target and an output gap – the so-called reac-tion function of central banks (Clarida et al., 1998; Mohanty and Klau, 2004; Svensson, 1999). In applying his rule to the USA for the 1987–92


period Taylor shows that it described the actual performance of policy well. IT is not applied mechanically and focuses on containing infl ation as a medium-term goal (Bernanke and Mishkin, 1997).8

The choice of the price index to be used by an IT regime is critical. Typically the consumer price index (CPI) or, preferably, a measure of core infl ation that ignores excessively volatile prices, for example, food and energy is used.9 Whether IT should respond to asset prices is debatable. Bean (2003) analyses an objective function minimizing output gaps and deviation from infl ation targets and argues that a middle solution between completely ignoring asset prices and including asset prices in the price index number, if these aff ect infl ationary expectations, be used for IT.

Central banks operate in an environment of considerable uncertainty about the functioning of the economy and global capital fl ows. Hence the conduct of monetary policy must be informed by examining a number of indicators and cannot rely on just one intermediate target. Most central banks, including those of developed countries, practice liquidity management involving estimating market liquidity, auton-omous of policy action, and initiate liquidity operations to steer mon-etary conditions and are thus able to switch between quantitative targets and interest rate targets in response to macroeconomic circumstances. Most central banks try to build in automatic stabilizers in the liquidity management framework, for example, by setting reserve requirements on an average basis to allow the fi nancial system the leverage to adjust to temporary/seasonal liquidity shocks on its own account without central bank action and exercising an explicit preference for encasing short-term interest rates in a corridor around some optimal rate rather than at a point target.

12.4 THE PERFORMANCE OF INFLATION TARGETING

There is considerable debate about whether IT improves performance in regard to infl ation and output. Ball and Sheridan (2003) argue that the adoption of IT does not lead to a systematic improvement in the growth-infl ation trade-off ; Hu (2004) argues otherwise. Fraga et al. (2003) focus on emerging market economies, including India, and show that economies working with IT have higher volatilities of output, infl a-tion, interest rates and exchange rates than developed countries using IT (Table 12.4).

Preparing for a switch to an IT regime requires considerable back-ground work. Financial markets should be suffi ciently developed


and global capital markets should have adequate confi dence in these markets, thus enabling the adoption of a suffi ciently fl exible exchange rate regime and the central bank should have a high degree of indepen-dence and be able to use short-term interest rates as the main operating instruments. Such conditions may not be satisfi ed in many transition countries.

Table 12.4 Volatility and average of selected variables for 1997:1–2002:2 (quarterly data)

Countries Volatility of basic variables Average

Infl ation Exchange rate*

GDP growth**

Interest rate

GDP growth

Infl ation

Developed Economies Australia 2.05 0.13 1.96 0.58 4.78 5.89Canada 0.83 0.04 1.30 1.14 3.57 1.96Iceland 2.45 0.15 3.13 3.02 4.17 4.05New Zealand 1.21 0.16 3.61 1.47 3.09 1.65Norway 0.77 0.10 2.25 1.46 2.66 2.44Sweden 1.11 0.12 2.41 0.44 2.58 1.24Switzerland 0.54 0.08 1.14 0.92 1.79 0.85United Kingdom 0.92 0.06 0.79 1.13 2.61 2.46Average 1.24 0.11 2.07 1.27 3.16 2.57Median 1.02 0.11 2.11 1.13 2.88 2.20Emerging Market Economies Brazil 2.09 0.31 2.06 7.06 1.81 5.89Chile 1.30 0.17 3.25 – 3.11 3.88Colombia 5.43 0.25 3.38 10.02 0.81 12.51Czech Republic 3.46 0.09 2.73 5.81 1.18 5.31Hungary 4.09 0.16 – 1.13 – 11.21Israel 3.18 0.10 3.36 3.34 2.98 4.35Mexico 5.98 0.07 3.17 7.26 4.05 11.72Peru 3.04 0.11 3.45 5.50 2.11 3.89Poland 4.13 0.11 2.40 4.14 3.85 8.40South Africa 2.13 0.26 1.11 3.65 2.26 6.51South Korea 2.36 0.14 6.38 5.52 4.31 3.73Thailand 3.25 0.14 6.13 6.72 0.08 2.88Average 3.37 0.15 3.40 5.47 2.41 6.69Median 3.22 0.14 3.25 5.52 2.26 5.60

Note: * refers to the coeffi cient of variation (standard deviation/mean); ** growth rate measured comparing the current quarter to the same quarter of the previous year.

Source: International Financial Statistics, IMF (quarterly data) (2003).


12.5 RECENT INDIAN EXPERIENCE WITH NOMINAL TARGETING

The RBI has never pursued a pure nominal targeting regime, opting for a combination of rules-based and discretionary measures with the former changing over time. In the 1980s and early 1990s the RBI opted for a nominal exchange rate peg externally, and monetary control internally. However, both policy mechanisms have faltered. An infl exibly pegged exchange rate has proved to be unsustainable in the presence of strong capital fl ows10 whereas the instability of the money demand function as well as its supply (Jha and Rath, 2003) indicates that monetary targeting, by itself, is no longer a feasible option.

Further, the RBI faces a persistent fi scal overhang and has to support high fi scal defi cits. If fi scal policy is imprudent and the central bank does not help fi nance the defi cit, the end result would still be infl ationary as the public debt/GDP ratio would turn unsustainable in the medium term and the price level could at least partially be determined by the fi scal theory of the price level.11

Fiscal defi cits are infl ationary and put pressure on real interest rates and crowd out private investment (Engen and Hubbard, 2004). There is a vicious cycle between infl ation and budget defi cits – high defi cits cause higher infl ation, which raise interest rates, thus increasing the defi cit by raising debt service payments. Further, higher infl ation reduces the real value of tax collections.

The literature has emphasized frameworks based on the clear mandates of central bank independence and fi scal responsibility legislation. Fiscal rules restrict government spending which checks excessive build-up of defi cits and public debt, imparting stability to the economy. Concurrently fi scal rules may restrict the government’s ability to take countercyclical policy measures and hence contribute to increased business cycle volatil-ity. Fiscal policy rules are likely to be eff ective if accompanied by strong commitments and increased transparency (Bayoumi and Eichengreen, 1995) – hence the widespread consensus for central bank independence backed by fi scal discipline to contain infl ation and stabilize infl ationary expectations.

Although price stability, output growth, reduction of exchange rate vola tility and fi nancial stability are monetary policy goals for the RBI, none of these are under its direct control. The RBI sets intermediate targets which it can control and which have a stable relationship with the ultimate goals of monetary policy. A narrow target such as base money may be fully within RBI control but may have only weak links with the ultimate objectives of monetary policy. A broad target such as nominal income


may be closely related to the ultimate objectives of monetary policy but not be amenable to RBI control. Both money supply and money demand have become unstable since the initiation of fi nancial sector reforms. Other nominal targets have similar problems. Hence a purely rules-based monetary policy regime seems unhelpful.

12.6 RECENT DEVELOPMENTS IN MONETARY POLICY DESIGN IN INDIA

With the progressive widening of fi scal defi cits from the 1960s, the burden of fi nancing was borne by the RBI and the banking system. The support of the banking system to the government’s borrowing program involved progressive increases in the statutory liquidity ratio to 38.5 percent by the early 1990s. Although interest rates on government securities were steadily raised to enhance their attractiveness, it became increasingly diffi cult to get voluntary subscriptions even at high interest rates. The cash reserve ratio was increased from 3 percent in the early 1970s to almost 25 percent (if incremental reserve requirements are taken account of) by the early 1990s. Nevertheless, liquidity growth remained excessively high during the 1970s and 1980s and spilled over onto infl ation. With expansionary fi scal policy there are limits to the eff ectiveness of monetary policy in containing infl a-tion. The combined defi cit of central and state governments has been close to 10 percent of GDP for more than 15 years and the share of net bank credit to the government in fi nancing the fi scal defi cit has hovered around 10 percent of GDP for much of the past decade.

The 1985 Chakravarty Committee on Monetary Policy recommended that price stability emerge as the ‘dominant’ objective of monetary policy along with commitment to fi scal discipline (RBI, 2002, p. 67). Price stabil-ity was seen to be critical to sustain the process of reforms begun in 1991 (RBI, 1993). In the latter half of the 1990s, as the economy slowed down, monetary policy pursued an accommodative stance with an explicit prefer-ence for a softer interest rate regime with a constant vigil on infl ation.

The RBI formally adopted a multiple indicator approach in April 1998. These are to (1) maintain a stable infl ation environment; (2) maintain appropriate liquidity conditions to support higher economic growth; (3) ensure orderly conditions in the exchange market to avoid excessive volatility in the exchange rate; and (4) maintain stable interest rates (RBI, 2002).

Besides broad money, which remains an information variable, other macroeconomic indicators including interest rates, rates of return on money, capital and government securities markets along with data on


currency, credit extended by banks and fi nancial institutions, fi scal posi-tion, trade, capital fl ows, infl ation rate, exchange rate, refi nancing and transactions in foreign exchange available on a high frequency basis are juxtaposed with output data for drawing policy perspectives when formu-lating monetary policy.

This ‘check list’ approach has been criticized as watering down the concept of a nominal anchor. However, it is diffi cult to fi nd a variable that would encapsulate the large number of factors that need to go into monetary policy making at this stage of transition from a relatively autarkic administered economy to a relatively open market-oriented one. The RBI uses a mix of policy instruments including changes in reserve requirements, and standing facilities and open market operations which aff ect the quantum of marginal liquidity and changes in policy rates, for example, the bank rate and repo/reverse repo rates, which impact the price of liquidity with short-term interest rates signaling the stance of monetary policy.

Shifts in monetary policy transmission channel necessitated policy impulses which would travel through both quantity and rate channels and episodes of volatility in foreign exchange markets emphasized the need for swift policy reactions balancing the domestic and external sources of monetization in order to maintain orderly conditions in fi nancial markets. Even within the set of indirect instruments the preference is for market-based instruments.

Another serious challenge comes from the capital account, especially the high volatility of capital fl ows vis-à-vis trade fl ows.12 Since external borrowings are denominated in foreign currency, large devaluations are infl ationary and cause serious currency mismatches with adverse eff ects on the balance sheets of borrowers. The need for reserves as self-insurance emanates from the volatile nature of capital fl ows (including sharp rever-sals) and refl ects weakness in the existing international fi nancial architec-ture. India’s ratio of net foreign assets to reserve money grew from 11.9 percent in 1990 to 44.5 percent in 1996, 65.8 percent in 2000 and 117.3 percent in 2003.

12.7 INDIA’S UNPREPAREDNESS FOR INFLATION TARGETING

That transition economies such as India may not be ready for IT is the considered view not just of the RBI but also International Monetary Fund (IMF) economists. Masson et al. (1997) argue that economic structures in developing countries (including India) are incapable of supporting an IT


regime in the short to medium runs, because such countries do not satisfy a number of prerequisites for the successful implementation of IT. The authors consider these to be:

1. Independence of the central bank: this refers both to operational conditions and the policy space within which the RBI can operate. There are limits to the eff ectiveness of monetary policy in containing infl ation when faced with expansionary fi scal policy. Domestic and fi nancial markets should have enough depth to absorb the place-ment of public and private debt instruments; and the accumulation of public debt should be sustainable. In the Indian case, while there is some evidence to suggest that the latter condition is satisfi ed (Jha and Sharma, 2004), the fi rst is defi nitely not (Sharma, 2004). If both condit ions are not satisfi ed the independence of monetary from fi scal policy is compromised and the interest rate transmission channel of policy is weak and incompletely evolved.

In addition, the central government can, even with fi nancial liber-alization, apply subtle pressure on the RBI to alter monetary policy. I give two instances of these. In the latter half of 2004, when infl ation topped 8 percent and real interest rates had become negative, the RBI wanted to raise the bank rate to lower infl ation but could not, under government pressure. Similarly in early 2005 the Governor of the RBI publicly voiced concern over volatile FII infl ows and suggested a fi scal approach to capping them. However, the Finance Minister almost immediately rebuff ed him.

2. Refraining from using other nominal anchors: successful adoption of IT requires that other nominal variables such as the exchange rate should not be targeted. However, India needs to maintain a stable and competitive exchange rate to encourage exports. Even in devel-oped economies, which have explicitly opted for it, IT is associated with high exchange rate volatility. In view of their vulnerability to exchange rate crises, developing countries such as India are wary of excessive volatility.

3. Predominance of demand as opposed to supply shocks: IT implicitly assumes that monetary policy has to respond primarily to demand shocks. Balakrishna (1991) has underscored the role of supply shocks in determining infl ation in India.13 These make infl ation dependent on monetary as well as non-monetary factors. If infl ation rises because of a demand shock, the pursuit of IT will stabilize both infl ation and output. However, if infl ation rises because of an adverse supply shock, the pursuit of IT will exacerbate the recessionary eff ect on output by reducing demand (McKibbin and Singh, 2003).


4. Practical diffi culties in the implementation of IT: the high frequency data requirements including those of a fully dependable infl ation rate for targeting purposes are not yet available (RBI, 2004).

12.8 CHECKING FOR VIABILITY OF INFLATION TARGETING IN INDIA

A prerequisite for the RBI to pursue IT is that there should be a stable and signifi cant relationship between the measure of infl ation to be controlled and short-term interest rates. I test for this using monthly data over the period April 1992 to March 1998 from the RBI’s Handbook of Statistics on the Indian Economy (2000). The variables used are as follows:

IIP: Index of industrial production (1980–815100); REER: Index of real eff ective exchange rate (36-country), 19855100; Narmon: Narrow money; Cmrate: Call money rate; Xrate: Exchange rate of Indian rupee vis-a-vis US dollar (monthly averages); CPI: Consumer price index for industrial workers (1982 5100); WPITR20: Trimmed WPI (Mohanty et al., 2000); WPI: Wholesale price index (1993–94 5100); WPIADM: Wholesale administered price index (ibid.). Monthly dummies were added and logs were taken of all variables except the Cmrate. Augmented Dickey Fuller tests (not reported here to conserve space) indicated that all series are I(1).14

The bivariate relationships between the three candidate infl ation meas-ures and the monthly economic indicators, the P values from bivariate Granger causality tests are reported in Table 12.5. Each entry gives the P values for the null hypothesis that the indicator does not cause the infl a-tion measure, that is, the probability of obtaining a sample, which is even less likely to conform to the null hypothesis of no Granger-causality than the sample at hand. These Granger causality results are reported up to eight lags.

The WPITR20 measure of infl ation assumes that the WPI is the head-line measure of infl ation and defi nes the trimmed mean infl ation index as:

WPITRa 51a1 2 2a a

100bb an21

i5k1 lwipi (12.1)

where WPITRa is the trimmed WPI computed by ordering the component price change data pi and associated weights wi and removing the compo-nents on each tail of the distribution by a percent. The numbers of com-ponents trimmed from the left and right tails of the distribution are k and l respectively. When a 5 0 the trimmed mean would equal the weighted


mean whereas in the case of a 5 50 it would equal the weighted median. The root mean square error (RMSE) for any level of trimming is defi ned by:

RMSEa 5 Åan

i51(pt

a 2 pt) 2/n (12.2)

Table 12.5 P Values from Bivariate Granger Causality Tests

CPI IIP Exrate Narmon REER Cmrate

Lags1 0.22 0.67 0.00* 0.64 0.352 0.4 0.72 0.00* 0.99 0.433 0.69 0.87 0.00* 0.61 0.934 0.1 0.5 0.00* 0.46 0.85 0.01* 0.25 0.00* 0.36 0.556 0.00* 0.13 0.00* 0.26 0.587 0.00* 0.12 0.00* 0.12 0.698 0.00* 0.16 0.00* 0.03* 0.82WPITR20Lags1 0.06 0.00* 0.07 0.01* 0.142 0.01* 0.00* 0.01* 0.00* 0.093 0.00* 0.00* 0.00* 0.00* 0.00*4 0.00* 0.00* 0.00* 0.00* 0.04*5 0.00* 0.00* 0.00* 0.00* 0.196 0.00* 0.00* 0.00* 0.00* 0.27 0.00* 0.00* 0.00* 0.00* 0.148 0.00* 0.00* 0.00* 0.00* 0.26WPIADMLags1 0.00* 0.33 0.08 0.75 0.452 0.00* 0.09 0.01* 0.56 0.233 0.00* 0.03* 0.00* 0.44 0.184 0.00* 0.00* 0.00* 0.2 0.975 0.00* 0.00* 0.00* 0.01* 0.446 0.00* 0.00* 0.00* 0.00* 0.417 0.00* 0.00* 0.00* 0.00* 0.48 0.00* 0.00* 0.00* 0.00* 0.12

Note: * 5 Signifi cant at 5 percent level; CPI: Consumer price index; WPITR: Trimmed whoesale price index; WPIADM: Price index for the administered goods; IIP: Index of indus-trial production; Narmon: Narrow money; Exrate: Exchange rate Rs/$; REER: Real eff ective exchange rate; Cmrate: Call money rate.


where pat is the trimmed WPI with a trimming ratio of a percent from

each of the tails of the price distribution at time t, pt is the 36-month centered moving average change in WPI at time t, and n is the number of samples. Mohanty et al. (2000) conclude that this RMSE is mini-mized for a 5 20. Data on this variable are available in Mohanty et al. (2000).

The results of the Granger causality test indicate a weak relation between Cmrate and the measures of infl ation. In fact only WPITR20 seems to have a causal relation with Cmrate. On the other hand, the links between the measures of infl ation and IIP, Narmon, Xrate and REER are much stronger. Hence the causality tests do not provide support for using interest rates as instruments in a policy of IT.

A drawback of the crude Granger causality testing is that it provides no information about whether the sign of the (dynamic) bivariate relationship is theoretically correct. Further, there may be omitted variables. A VAR on the variables lCPI, lIIP, lNarmon, lREER and Cmrate indicates that the conclusions of Table 12.5 are broadly correct.15

Figure 12.1 shows that the 95 percent confi dence band for the

–0.02

0 10step

model 1: Cmrate –> LCPI

20 30

–0.01

0

0.01

95% CI for OIRFOIRF

Note: The band indicates the 95 percent confi dence interval for the orthogonalized impulse response function.

Figure 12.1 Orthogonalized impulse response function (OIRF) of call money rate on log CPI


orthogonalized impulse response function of Cmrate on lcpi is very wide, hence adding to our agnosticism about the effi cacy of IT in India.

The error correction model for lCPI is not signifi cantly responsive to any of the error correction terms. Hence it appears that IT may be diffi cult to pursue in the Indian context.

12.9 CONCLUSIONS: OPTIONS FOR INFLATION TARGETING IN INDIA

This chapter has argued that the primary objective of Indian monetary policy, at least in the medium term, has to be the attainment of higher economic growth. Moreover, since India has high infl ation aversion, this objective does not confl ict with short-term stabilization.

The design of monetary policy in India is circumscribed by the fact that the liberalization of fi nancial markets is far from complete so that the interest rate transmission channel is incomplete. Further, the banking system has strong monopoly elements. Moreover, as the fi nancial sector liberalizes some major government-owned mutual fund operations have had to be bailed out. The exacerbation of such contingent liabilities along with already high fi scal defi cit aggravates monetary policy diffi culties in the Indian context.

Against this background this chapter has argued that the multi- objective formulation pursued by the RBI has merit and that such monetary policy should be pursued to maintain stable interest and infl ation rates and a slightly undervalued currency in order to engineer higher export-led growth.

This policy has, however, led to substantial capital infl ows with attend-ant build-up of reserves and necessitated considerable sterilization opera-tions. This has now emerged as a signifi cant problem with its continuance at the current pace seemingly unsustainable if, for no other reason, than the fact that such reserves attract low yields. Currently two policy pack-ages to address this issue have been discussed. The fi rst is geared towards fi scal correction and monetary expansion. A second policy measure is weighted towards real exchange rate appreciation (more in line with IT) and would involve relatively larger current account defi cits. Real appreci-ation could be secured by nominal appreciation or higher infl ation. Both policies would lead to low infl ation rates and reduced infl ows of foreign capital and, therefore, lower accumulation of reserves at given rates of sterilization. Policy packages that use import liberalization would, like real appreciation, permit higher absorption via higher current account defi cits but without penalizing exports. The optimal package for India is


a judicious combination of these two broad sets of policies with greater emphasis on fi scal consolidation and import liberalization, rather than real exchange rate appreciation. These are essential elements of an appro-priate monetary policy regime for India.

Since rapid export growth is important, it makes sense to err on the side of undervaluation of the exchange rate, enabling India to capture a larger share of world markets. Growing exports, in turn, raise the incen-tive to invest. The propensity to save also rises in response to the increased profi tability of export-oriented investment. Moreover, an undervalued exchange rate is likely to boost saving by raising the share of profi ts in national income. This argument does not imply that unlimited real depre-ciation is feasible or desirable, just that there should be a bias towards mild undervaluation because it can play a supportive role to complemen-tary outward-oriented trade policies in generating a virtuous circle of higher saving, investment and growth. Along this path import demand would grow concomitantly and getting a current account surplus is not inevitable.

Clearly India has been conducting some form of real exchange rate tar-geting leading to a sharp rise in foreign exchange reserves. Lal et al. (2003) indicate that this has come at high economic costs – a claim that has been successfully disputed.16 IT would require India to pursue a clean fl oat reducing the need for large reserves. But the price to be paid is the pos-sibility of a highly unstable or inappropriate exchange rate. India’s policy makers were wise to reject this regime and opt for managed fl oating plus selective controls on capital fl ows. Reserves are now at a very comfortable level and continue to rise rapidly. The question of whether and how to absorb foreign infl ows is far more pertinent now than it was in the 1990s.

Clearly sterilization has outlived its usefulness. Some sterilized reserve accumulation can continue to maintain the present ratio of reserves to GDP. Further increases in the ratio should be avoided except as a purely short-term response to manifestly short-term infl ows. The policies espoused here have the advantage that, in addition to promoting balance of payments adjustment, they are desirable independently of the balance of payments and of the ‘temporary’ or ‘permanent’ character of the infl ows. Naturally, due to political constraints, these policies can only be pursued at a moderate pace. If there is continued acceleration of infl ows, despite the adoption of the suggested strategy, the government should consider tightening capital infl ow controls17 so that the strategy is not derailed. A corollary would be that capital account convertibility is eschewed.

It is not being suggested that India should resist an exchange rate appre-ciation indefi nitely. Once Indian GDP has grown in excess of 8 percent for more than two decades, so that real incomes have gone up substantially


and unemployment and poverty have dropped sharply, India could con-template changes in the monetary policy regime, essentially in the direc-tion of relaxing the undervaluation of the exchange rate. However, even at that point, given the importance of variables other than infl ation in an optimal monetary policy framework, there would be no argument for moving to IT.

NOTES

1. I am grateful to Gerald Epstein and three anonymous referees for helpful comments and Anurag Sharma for research assistance. The usual disclaimer applies.

2. At the very least there is no evidence that economic growth hurts poverty alleviation (Winters et al., 2002).

3. Results from the 1999–2000 National Sample Survey (Central Statistical Organisation, 2000) indicate a larger drop in poverty; however, this survey’s methodology is not com-parable with those of the earlier surveys, correcting for which reveals a modest drop in poverty.

4. Nominal income, another intermediate target, is both hard to target and poorly related to the ultimate objectives of monetary policy.

5. Another alleged advantage of an IT regime is that deviations from infl ation targets are routinely allowed in response to supply shocks by excluding some combination of food and energy prices, indirect tax changes, terms of trade shocks and the direct eff ects of interest rate changes on the index.

6. See also Epstein (2000). 7. For an application to India, see Section 12.8. 8. Seyfried and Bremmer (2003) discover that the Reserve Bank of Australia pays par-

ticular attention to infl ationary pressures, as measured by the GDP gap. They fi nd a relatively high degree of persistence and low speed of adjustment in the interest rate indicating that the central bank is interested in interest smoothing in addition to IT. See also Lomax (2005).

9. India with one wholesale price index and four CPIs still does not have a single price index with widespread acceptability for IT. Measures of core infl ation are not com-puted offi cially (Mohanty et al., 2000).

10. Joshi and Sanyal (2004) indicate the RBI has been targeting the index of real eff ective exchange rate (REER) of the Indian rupee with regard to the currencies of fi ve coun-tries, USA, Japan, UK, Germany and France, at the 1993–94 level. Patel and Srivastava (1997) note that such targeting has had more than a transitory eff ect. However this benign relationship may break as reforms lead to greater capital mobility.

11. In Latin America Jacome and Vazquez (2005) fi nd no causal relationship between central bank independence and infl ation, although the association between the two is strong.

12. The Indian rupee is convertible on the current account but capital account converti-bility is not permitted.

13. Callen and Chang (1999) review the literature on infl ation in India. The threshold of infl ation over which price stability should take precedence over the growth objective has been estimated to be between 4 and 6.5 percent (Samantaraya and Prasad, 2001).

14. Since the Mohanty et al. (2000) data set stops at 1998 we do not extend our analysis beyond that date.

15. Detailed results are reported in the full version of the paper available on the PERI website, accessed 6 May 2007 at http://www.peri.umass.edu.

16. Lal et al. (2003) argue that India’s growth rate in the 1990s could have been up to


2.7 percent per annum higher if foreign exchange infl ows during the decade had been fully absorbed. However, Joshi and Sanyal (2004) argue that if net foreign infl ows had been absorbed domestic spending (and not foreign exchange reserves) would have risen. Reserves as a proportion of GDP rose over the 1990s by an average of about 1.2 percent per annum. If the entire increase in reserves had been absorbed into investment each year, the ratio of investment to GDP averaged over the decade would have been 1.2 percent higher than it actually was. Given fi xed incremental capital output ratio (ICOR) (of 2.8 in the 1990s) the increase in India’s growth rate of GDP would have been only 1.2/2.8 5 0.4 percent per annum (approx.) higher over the decade. This sacri-fi ce would be even lower if (1) the ICOR would have risen (in line with the assumption of diminishing returns to capital); (2) some of the reserve accumulation spilled over onto higher consumption, thus reducing the growth rate; and (3) furthermore, the level of foreign exchange reserves in India was inadequate in 1991 and accumulating foreign reserves was necessary.

17. Through a Chilean-type tax, for example.

REFERENCES

Balakrishna, P. (1991), Pricing and Infl ation in India, Delhi: Oxford University Press.

Ball, L. and N. Sheridan (2003), ‘Does infl ation targeting matter?’, National Bureau for Economic Research, working paper no. 9577.

Barro, R. and D. Gordon (1983), ‘A positive theory of monetary policy in a natural rate model’, Journal of Political Economy, 91 (4), 589–610.

Bayoumi, T. and B. Eichengreen (1995), ‘Restraining yourself: the implications of fi scal rules for economic stabilization’, IMF Staff Papers, 42, 32–48.

Bean, C. (2003), ‘Asset prices, fi nancial imbalances and monetary policy: are infl a-tion targets enough?’, Bank for International Settlement working paper no. 140.

Bernanke, B. (2003), ‘Constrained discretion and monetary policy’, BIS Review, 5, 1–8.

Bernanke, B. and M. Gertler (2001), ‘Should central banks respond to movements in asset prices?’, American Economic Review, 91 (2), 253–7.

Bernanke, B. and F. Mishkin (1997), ‘Infl ation targeting: a new framework for monetary policy?’, Journal of Economic Perspectives, 11 (2), 97–116.

Blinder, A. (1998), Central Bank in Theory and Practice, Cambridge, MA: MIT Press.

Callen, T. and D. Chang (1999), ‘Modeling and forecasting infl ation in India’, International Monetary Fund working paper WP/99/119, Washington, DC.

Central Statistical Organisation (2000), National Sample Survey, New Delhi: Government of India.

Clarida, R., J. Gali and M. Gertler (1998), ‘Monetary policy rules in practice: some international evidence’, European Economic Review, 42 (4), 1033–67.

Dollar, D. and A. Kraay (2001), ‘Growth is good for the poor’, World Bank policy research paper no. 2587, pp. 1–50.

Engen, E. and R. Hubbard (2004), ‘Federal governments and interest rates’, National Bureau for Economic Research working paper no. 1068.

Epstein, G. (2000), ‘Myth, mendacity and mischief in the theory and practice of central banking’, accessed 6 May 2007 at www.umass.edu/peri.


Filrado, A. (2004), ‘Monetary price and asset price bubbles: calibrating the mon-etary policy tradeoff s’, Bank for International Settlement working paper no. 155.

Fraga, A., I. Goldfajin and A. Minella (2003), ‘Infl ation targeting in emerging market economies’, National Bureau for Economic Research, working paper no. 10019.

Hu, Y. (2004), ‘Empirical investigations of infl ation targeting’, Mimeo, Institute of International Economics, Washington, DC, February.

Jacome, L. and F. Vazquez (2005), ‘Any link between central bank independence and infl ation? Evidence from Latin America and the Caribbean’, International Monetary Fund working paper no. WP/05/75.

Jha, R. (2004), ‘Reducing poverty and inequality in India: Has liberaliz-ation helped?’, in A. Cornia (ed.), Inequality, Growth and Poverty in an Era of Liberalization and Globalization, Oxford: Oxford University Press, pp. 287–326.

Jha, R. (2005), ‘The political economy of recent economic growth in India’, in R. Jha (ed.), Economic Growth, Economic Performance and Welfare in South Asia, Houndmills: Palgrave Macmillan, pp. 28–51.

Jha, R. and D. Rath (2003), ‘On the endogeneity of the money multiplier in India’, in R. Jha (ed.), Indian Economic Reforms, Basingstoke: Palgrave Macmillan, pp. 51–72.

Jha, R. and A. Sharma (2004), ‘Structural breaks, unit roots, and cointegration: a further test of the sustainability of the Indian fi scal defi cit’, Public Finance Review, 32 (2), 196–219.

Jha, R., R. Gaiha and A. Sharma (2006), ‘Mean consumption, poverty and inequality in India in the sixtieth round of the national sample survey’, ASARC working paper no. 2006/11, Australian National University.

Joshi, V. and S. Sanyal (2004), ‘Foreign infl ows and macroeconomic policy in India’, Mimeo, Oxford University, pp. 1–54.

Kelkar, V. (2004), ‘India on the growth turnpike’, Narayanan Oration, presented at Australia South Asia Research Centre, Australian National University.

Kydland, F. and E. Prescott (1977), ‘Rules rather than discretion: the inconsist-ency of optimal plans’, Journal of Political Economy, 87 (2), 473–92.

Lal, D., S. Bery and D. Pant (2003), ‘The real exchange rate, fi scal defi cits and capital fl ows – India: 1981–2000’, Economic and Political Weekly, 38 (47), 4965–76.

Lomax, R. (2005), ‘Infl ation targeting in practice – models, forecasts and hunches’, speech by the Deputy Governor of the Bank of England to the 59th Atlantic Economic Conference, 12 March.

Masson, P., M. Savastano and S. Sharma (1997), ‘The scope for infl ation target-ing in developing countries’, International Monetary Fund working paper no. 97/130.

McKibbin, W. and K. Singh (2003), ‘Issues in the choice of a monetary regime for India’, in R. Jha (ed.), Indian Economic Reforms, Basingstoke: Palgrave Macmillan, pp. 11–50.

Mohanty, M. and M. Klau (2004), ‘Monetary policy rules in emerging economies: issues and evidence’, Bank for International Settlement working paper no. 149.

Mohanty, D., D. Rath and M. Ramaiah (2000), ‘Measures of core infl ation for India’, Economic and Political Weekly¸ 35 (5), 273–82.


Obstfeld, M. and K. Rogoff (1996), Foundations of International Macroeconomics, Cambridge, MA: MIT Press.

Patel, U. and P. Srivastava (1997), ‘Some implication of real exchange rate target-ing in India’, Mimeo, Indian Council for Research on International Economic Relations, New Delhi, March.

RBI (Reserve Bank of India) (1993), Annual Report 1991–92, Mumbai: RBI.RBI (2000) Handbook of Statistics on the Indian Economy, Mumbai: Reserve Bank

of India.RBI (2002), Reserve Bank of India Bulletin, Mumbai: RBI.RBI (2004), Report on Currency and Finance, Mumbai: RBI.Samantaraya, A. and A. Prasad (2001), ‘Growth and infl ation in India: detecting

the threshold level’, Asian Economic Review, 43 (3), 414–20.Seyfried, W. and D. Bremmer (2003), ‘Infl ation targeting as a framework for

monetary policy: a cross-country analysis’, Australian Economic Review, 36(3), 291–9.

Sharma, A. (2004), ‘Fiscal defi cits, fi nancial crises and adjustment in a semi-open economy’, unpublished PhD dissertation, Australian National University.

Svensson, L. (1999), ‘Infl ation targeting as a monetary policy rule’, Journal of Monetary Economics, 43 (3), 607–54.

Taylor, J. (1993), ‘Discretion versus policy rules in practice’, Carnegie–Rochester Conference Series on Public Policy, 39, 195–214.

Taylor, J. (2002), ‘The monetary transmission mechanism and the evaluation of monetary policy rules’, in N. Loayza and K. Schmidt-Hebbel (eds), Monetary Policy: Rules and Transmission Mechanisms, Santiago: Central Bank of Chile.

White, W. (2004), ‘Making macro prudential concerns operational’, speech delivered at Symposium on Financial Stability at De Nederlandsche Bank, Amsterdam, March.

Winters, A., N. McCulloch and A. McKay (2002), ‘Trade liberalisation and poverty: the empirical evidence’, Department of Economics, University of Sussex discussion paper no. 88, Brighton.


APPENDIX 12.A1 CAPITAL CONTROLS: OR WHY DID INDIA ESCAPE THE EAST ASIAN CRISIS?

It is instructive to compare India and the East Asian countries in 1996 (that is, just before the East Asian crisis of 1997). The fi rst six columns of Table 12.A1 indicate that, in most respects, India’s ‘fundamentals’ (fi scal balance, infl ation, current account balance, non-performing assets, debt-exports ratio and debt-service ratio) were worse or no better than the crisis-countries. All these countries were on loose dollar peg and India was only marginally diff erent from the rest in this regard except for the fact that India did not allow its real exchange rate to appreciate and was able to maintain its real exchange targeting posture.

The critical diff erence between India and the crisis-countries can be seen in the last two columns of Table 12.A1. India managed to keep short-term debt under control, both in relation to total debt and in relation to foreign exchange reserves and thus avoided an unstable debt structure, an outcome that was the direct result of controls on debt-creating short-term infl ows.

Table 12.A1 Various countries: indicators of crisis-vulnerability, 1996

FB/GDP (%)

ΔP/P (% p.a.)

CAB/XGS (%)

NPA (%)

NCEDT/XGS (%)

TDS/XGS (%)

SDT/EDT (%)

SDT/RES (%)

India –9.0 9.0 –11.7 17.3 103.6 21.2 5.3 27.1Indonesia –1.0 8.0 –13.0 8.8 180.5 36.6 25.0 166.7Korea 0.0 4.9 –14.6 4.1 82.0 9.4 49.4 192.7Malaysia 0.7 3.5 –6.4 3.9 40.4 9.0 27.9 39.7Philippines 0.3 8.4 –9.9 n.a. 80.1 13.4 19.9 67.9Thailand 0.7 5.8 –19.5 7.7 110.9 12.6 41.5 97.4

Note: FB/GDP: Fiscal balance as a proportion of GDP; ΔP/P: Rate of consumer price infl ation; CAB/XGS: Current account balance as a proportion of exports of goods and serv-ices; NPA: Non-performing assets of commercial banks as a proportion of total advances; NCEDT/XGS: Non-concessional external debt as a proportion of exports of goods and services; TDS/XGS: Debt service as a proportion of exports of goods and services; SDT/EDT: Short-term external debt as a proportion of total external debt; SDT/RES: Short-term external debt as a proportion of foreign exchange reserves.

Sources: FB/GDP, NPA: Bank of International Settlements Annual Reports (1997–98 and 1999–2000) and Government of India, Economic Survey (1999–2000).CAB/XGS, NCEDT/XGS, TDS/XGS, SDT/EDT, SDT/RES: World Bank, Global Development Finance (1999).ΔP/P: IMF International Financial Statistics (2000).


India was able to resist the pressure to adopt capital account convert-ibility essentially because of three reasons: fi rst, the ideology of laissez-faire is still not dominant in India, and second, foreign banks, which are normally a strong pressure group in favor of capital account convertibil-ity, had a very small presence in the country. Finally, India was ‘too big to be bullied’ into adopting capital account convertibility by Wall Street, the IMF and the US Treasury (Joshi and Sanyal, 2004).

271

13. Towards an alternative monetary policy in the PhilippinesJoseph Anthony Lim1

13.1 INTRODUCTION

The modern macroeconomic history of the Philippines had been marked by periodic balance of payment crises with massive devaluations that resulted in high infl ation. These crises were followed by drastic monetary and fi scal recessionary programs implemented by the International Monetary Fund (IMF), as a precondition for the release of emergency funds.

Monetary targeting was implemented in the Philippines from the 1980s to the 1990s under IMF sponsorship. It is generally accepted as contribut-ing to the depth of recessions in the Philippines in the last 25 years. Since it was replaced by infl ation targeting only in 2002, it is diffi cult not to include in this chapter a discussion of the experience with monetary targeting.

The next section describes the monetary targeting experience of the Philippines. Section 13.3 discusses the shift from the monetary targeting regime to the infl ation targeting regime starting 2002. Section 13.4 places monetary policy in the context of the complex macro situation and devel-opment needs of the Philippines. The last section gives detailed recommen-dations for an alternative monetary policy.

13.2 MONETARIST TARGETING AND POLICY IN THE PHILIPPINES: CRITICIZING THE DEMAND-SIDE CURE FOR INFLATION AND CURRENT ACCOUNT DEFICITS

The economic crises which occurred from the late 1940s to the present had been connected with balance of payment and foreign exchange crises. These have led to some very sharp recessions – especially the economic col-lapse in 1984–85 – that destroyed any chance of the Philippines becoming an East Asian success story.

Carrying a strong belief that current account defi cits and infl ation


are due to excessive aggregate demand caused by monetary expansion, monetarism’s response to balance of payment crises involved damaging pro-cyclical monetary and fi scal austerity that deepened the crisis and recession. Starting in the early 1960s the IMF had become the standard funder of last resort during crises. In the early 1980s quarterly monetary targeting became the norm. These monetary targets became very tight every time balance of payments deteriorated and infl ation increased. Monetary targets were based on targets on the monetary base and achieved through: (1) high required reserve ratio; (2) high policy rates of the central bank; and (3) open market sale of central bank bills and government securities in order to reduce the monetary base. It was the use of the third instrument that was most damaging as it directly reduced liquidity and credit in the fi nancial sector.

13.2.1 Supply-Side Causes of Infl ation

Figure 13.1 gives a picture of infl ation rate based on the consumer price index (CPI) and GDP growth rate. The graph shows that high infl ation usually happens simultaneously with lower growth or recessions, rather than during periods of high growth and high aggregate demand.

This is particularly true for the periods 1984–85 (economic collapse), 1990–91 (another recession) and 1998 (Asian crisis period). This is because these periods were periods of signifi cant currency devaluations resulting from balance of payment crises. This brings about stagfl ation that explains the high infl ation and recession.

In fact, the role of currency devaluations and oil price shocks in explain-ing periods of high infl ation in the Philippines is very clearly illustrated in Figure 13.2.

It can be seen that the devaluation in 1970 brought high infl ation in 1970–71. The fi rst oil price shock in 1973–74 brought about high infl a-tion in 1973 and, especially in 1974. Again the second oil price shock and worldwide infl ation in 1979–81 brought about the high infl ation during the same years. The massive devaluations in 1983 and 1984 led to high infl a-tion in the economic collapse period of 1984–85. The moderate devalua-tion in 1990–91, plus the oil price shock due to the fi rst Gulf War crisis brought about the infl ation in 1990 and 1991. The signifi cant devaluation during the Asian crisis in 1998 brought a slight uptick in infl ation, but nowhere near the high infl ation that occurred in previous devaluations.

Thus, by just using two types of supply-side shocks – currency depre-ciation and oil price shock – one can explain practically all the above-10 percent infl ation in modern Philippine history. High infl ation periods are not triggered by high domestic demand but by supply-side shocks.

Towards an alternative monetary policy in the Philippines 273

If one uses monthly or quarterly data on infl ation, one will also see that agricultural price shocks make their impact on price infl ation due to weather disturbances leading to food shortages.

13.2.2 Monetary Contraction and Impact on GDP and Unemployment

Using monetary contraction to fi ght infl ation aggravates recessions and unemployment, since the original supply shocks (devaluation, oil price shocks or food shortages) already have recessionary impact. It is import-ant to point out that the serious stagfl ationary periods 1984–85, 1991 and 1998 were all periods when the Philippines undertook severe monetary and credit contraction.

Figure 13.3 shows the relationship of lending rate and GDP growth rate. Lending rates rise during crises because of the higher infl ation, but credit and monetary contraction goes beyond this since the authorities purposely contract the monetary base and increase required reserves and the central bank policy rate. As expected, one sees a strong inverse rela-tionship between the lending rate and GDP growth rate.

–10

0

10

20

30

40

50

1950

1953

1956

1959

1962

1965

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

2001

2004

Per

cent

(%

)

GDP per capita growthCPI inflation rate

Source: National Statistics Coordination Board; International Financial Statistics (IMF).

Figure 13.1 GDP per capita growth versus CPI infl ation rate


The lending rates in Figure 13.3 actually underestimate the costs of borrowing since during periods of monetary contraction, many fi rms are credit-rationed. Figure 13.4 shows the relationship between growth of real money and GDP growth rate. Declines in real money are associated with declines in GDP growth in 1960, 1964, 1974, 1984–85, 1991 and 1998.

Figure 13.5 shows that the recessions of 1984–85, 1991 and 1998 created signifi cant upticks in the unemployment rate.

13.2.3 Monetary Contraction and External Defi cits

Much of the demand-suppressing monetary austerity programs were implemented not only to reduce infl ation but to reduce aggregate demand to ensure that current account defi cits were also reduced during balance of payment crises. And it is often overdone to produce current account surpluses.

Figure 13.6 shows a graph of the current account defi cit (as percentage of GDP) and GDP per capita growth rate over the years. Concentrating

–10

0

10

20

30

40

50

60

70

80

90

1949

1952

1955

1958

1961

1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

Per

cent

(%

)

CPI inflation rateGrowth rate of exchange rate


Figure 13.2 CPI infl ation versus growth of exchange rate


on three recession periods – 1984–85, 1991 and 1998 – it is clear that every crisis was preceded by a year or two of very signifi cant current account defi cits. This is because import demand grows during boom years as developing countries are dependent on imported raw materials and capital goods. It is also clear that after every recession, the current account defi cit improved, and in the case of the crisis in the mid 1980s and the Asian crisis the current account turned positive after the recession. It should be emphasized that the improvement in the current account balance was caused partly by the massive devaluation and partly by the recession which had been aggravated by the monetary contraction dictated by monetarist policy. Monetarist theory and the IMF interpret large trade defi cits as ‘overspending’ and require monetary and fi scal tightness. Thus, although monetary targeting is supposed to target only one variable – infl ation – it is also aff ected by foreign exchange depletion and capital outfl ows during crises.

Finally the last possible reason for monetary tightening during balance of payment crises is to stem the fl ight of capital and attract them back into the country, and at the same time stave off the damaging exchange rate

–10

–5

0

5

10

15

20

25

30

35

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

Per

cent

(%

)

Lending rateGDP per capita growth


Figure 13.3 GDP per capita growth versus lending rate


collapse. This was mainly the excuse in the 1984–85 debt crisis and the 1998 Asian crisis. However, in both cases it was not the high interest rates that stopped the currency depreciation and balance of payments outfl ow but mainly the devaluation-cum-recession which improved the current account by improving the trade accounts. Figure 13.7 tracks the current account and balance of payment balances. It can be seen in the graph that the balance of payments improvement from 1984 to 1987 following the 1984–85 collapse was mainly due to very signifi cant improvements in the current account. In the 1998 Asian crisis even the improvement of the balance of payments in 1998 was less than the current account improve-ments, indicating net capital outfl ows in a period of high interest rates. But in subsequent years the positive balances in the current account exceeded that of the balance of payments, which meant that there was continuing net capital outfl ow in the balance of payment account from 1999 to 2002 (most likely due to the lackluster performance of the Philippine economy

–30

–20

–10

0

10

20

30

40

50

1952

1955

1958

1961

1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

Per

cent

(%

)

Real money supplygrowth rate

GDP per capita growth


Figure 13.4 GDP growth rate versus real money supply growth


and the political troubles of the Estrada and Arroyo governments). In the crises of 1984–85 and 1998 the current account went into signifi cant positive territory because of the joint eff ects of devaluation and output and demand contraction. This proves that monetarist aggregate demand suppression is successful in stopping foreign exchange outfl ows and current account defi cits – not by stemming capital outfl ows and encourag-ing return of foreign exchange due to high interest rates but, as previous graphs prove, by creating sharp recessions and severe unemployment, which reduce current account defi cits. Infl ation rates are also lowered because recessions kill consumer and investment demand.

An important lesson from the discussion above is that the roles of the current account, the exchange rate regime and issues concerning how to stem capital fl ight should be incorporated in any alternative to the current monetary policies.

From the 1980s to 2000 the sole purpose of monetary policy was to fi ght ‘overspending’ that had ‘caused’ current account defi cits and high infl ation. It ensured that monetary policy could not be used for

–8

–6

–4

–2

0

2

4

6

8

10

12

14

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Per

cent

(%

)

GDP Growth rateUnemployment rate

Source: National Statistics Coordination Board; Labor Force Survey, National Statistics Offi ce; International Financial Statistics (IMF).

Figure 13.5 Unemployment rate versus GDP growth rate


countercyclical and growth purposes. The adoption of fi nancial lib-eralization policies starting in the 1980s also ensured that monetary policy could not be used to fi nance targeted or prioritized sectors of the economy. This complemented the prescription of the multilateral agen-cies for the country to abandon industrial policy and undertake trade liberalization.

At the same time the policy is not matched by a confi rmation of the theory behind the demand theory of infl ation and ‘overspending’. Figure 13.8 graphs the infl ation rate on the y-axis and unemployment rate on the x-axis for the years from 1959 to 2004 to fi nd out any trace of a Phillips curve. One can see there is nothing that looks like it in the graph in any sub-period. This is because infl ation is caused by supply shocks and not demand-led. (The graph looks like it is producing ‘natural rates of infl a-tion’ rather than ‘natural rates of unemployment’.)

–10

–8

–6

–4

–2

0

2

4

6

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

Per

cent

(%

)

GDP per capita growthCurrent account balance, % of GDP


Figure 13.6 Current account balance as percentage of GDP versus GDP per capita growth


13.3 FROM MONETARY TARGETING TO INFLATION TARGETING

13.3.1 A Switch in Regime: From Monetary Targeting to Infl ation Targeting

In the Philippines research work in the central bank itself pointed to the ‘overkill’ created by monetarist policy in trying to tame infl ation. This is mainly because the quantity theory of money equation assumes a unitary coeffi cient of money to prices with output holding steady, despite massive monetary contraction. The use of a single estimated infl ation equation assumes all explanatory variables (exchange rate movements, output growth, and so on) to be exogenous and that only the monetary variable will decrease infl ation. A study by a staff member of the central bank

–12

–10

–8

–6

–4

–2

0

2

4

6

8

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

Per

cent

(%

)

Current account balance, % of GDPBOP, % of GDP


Figure 13.7 Current account and balance of payments, percentage of GDP


(Dakila, 2001) shows that with a simultaneous equation system where exchange rate movements and output movements are endogenized, the degree of monetary tightening required to achieve a stipulated reduction of infl ation is lessened. This is because monetary tightening cum recession-ary policies tends to eventually cause some appreciation of the currency and a fall in output growth (both of which tend to be defl ationary).

When single-digit infl ation rates were achieved in the mid 1990s under the Ramos administration, the central bank relaxed monetary targeting to what was called the ‘modifi ed monetary targeting’ policy. This allowed the monetary targets to be exceeded as long as infl ation targets were being met. This was the start of a hybrid system of infl ation and monetary targeting. The offi cial reason given was that, because of fi nancial liberalization, the link between quantitative monetary targets and infl ation had weakened due to ‘structural breaks’ in the income velocity of money and volatilities and instabilities in the money multiplier. Thus the high liquidity and large monetary expansion in the mid 1990s failed to have any impact on infl a-tion as it remained below double-digit rates and continued to decline until the Asian crisis (see Guinigundo, 2005).

The laxer monetary policy in the mid 1990s of course was shattered by the ‘contagion’ eff ects of massive depreciation pressures during the Asian crisis. The major ways used to stave off the depreciation (and indirectly

–1

9

19

29

39

49

59

8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00

Unemployment rate

Infla

tion

rate

Source: Labor Force Survey, National Statistics Offi ce; International Financial Statistics (IMF).

Figure 13.8 Infl ation versus unemployment rate


avoid infl ationary pressures) were constant and periodic raising of the reverse repurchase rate and the raising of liquidity reserve ratios.2

However, the massive currency depreciation (around 40 percent) during the Asian crisis had a very low ‘pass-through’ to infl ation (see Figure 13.2). It did not even lead to (and only approached) double-digit infl ation. This easily allowed infl ation rates to fall after the crisis. Infl ation rates reached a very low 3 percent in 2002 and 2003, replicating the infl ation rates in the developed world.

By January 2000, monetary authorities were already studying the alleged successes of developed and developing countries that had adopted infl ation targeting (from a prior regime of exchange rate pegging or monetary targeting). The decision was to make the big switch to infl a-tion targeting starting January 2002. The switch entailed the following el ements: (1) continuation of the announcement of infl ation targets based on a band over a two-year time period; (2) adoption of a more passive monetary quantitative policy, and employing more the repurchase and reverse repurchase rates (policy rates) as the monetary instrument; (3) increased sophistication in infl ation rate estimation and in using single and multi-equation models to forecast infl ation and setting of infl ation targets; (4) the use of forward-looking models with monetary instruments reacting to and aiming to infl uence infl ationary expectations rather than actual infl ation; (5) creation of an advisory committee to recommend monetary policies based on the new infl ation targeting regime; and (6) the issuance of a quarterly infl ation report explaining the central bank’s policies and achievement or non-achievement of the infl ation targets.

But the most important elements are: (7) the stipulation of escape clauses that exempts the central bank from achieving the infl ation targets. This means that if the infl ation target is not achieved, the central bank can opt to not do anything if the reasons are: (a) sudden changes in prices of agricultural commodities and products; (b) natural calamities or cata-strophic events; (c) volatility in oil prices; and (d) sudden changes in gov-ernment policies, such as tax structures.

Another important element is: (8) the setting up of a ‘core’ infl ation rate, as opposed to the overall CPI or ‘headline’ infl ation rate. The core takes out oil and agricultural products whose prices are easily aff ected by external shocks and weather disturbances. This is an important element since even if overall ‘headline’ infl ation rate is not achieved, as long as the ‘core’ infl ation rate is within the infl ation target, the central bank can also opt not to do anything.

A central bank report (Guingundo, 2005) mentions two important things: (1) the current infl ation targeting method allows for ‘ample room for judgment and discretion of policy makers’; and (2) in explaining why


overshooting of the infl ation target in the fourth quarter of 2004 did not lead to increases in policy rates by the central bank, the publication gives the following explanation: (a) the infl ationary pressures caused by supply-side shocks (oil price increase) were not susceptible to monetary action, since the latter would work on the demand side; (b) the central bank fore-casts indicated that the pressures would subside in 2006; and (c) there were downside risks to the overall strength of economic activity. The last point is extremely important as it proves that current central bank policy does put weight on the strength or weakness of ‘economic activity’.

13.3.2 A More ‘Benign’ Policy?

So far ‘infl ation targeting’ in the Philippines from 2002 to 2006 has not led to very drastic monetary tightening unlike ‘monetary targeting’.

The improvement in infl ation targeting over monetary targeting can be due to a number of factors. First, the nature of the policy tools makes infl a-tion targeting more benign than monetary targeting. To illustrate, main-taining a monetary base target will entail credit tightening since it does not allow money supply and domestic credit to grow with the economy and with infl ation. On the other hand, maintaining the policy rate of the central bank (that is, keeping the reverse repurchase rate constant within reasonable limits) does not entail credit tightening and allows money supply and domestic credit to grow moderately. Thus to make monetary targeting and infl ation targeting equally restrictive will entail increasing policy rates drastically, which is much more obvious and prone to public criticism compared to reducing the monetary base target (which had not been transparent and kept in IMF memorandums). The Philippines, in its infl ation targeting history (2002 to present), has not increased the policy rate drastically and so infl ation targeting policy in the Philippines has not mimicked monetary targeting by drastically cutting money supply and credit.

More importantly, the adoption of infl ation targeting was not done as a precondition to IMF conditionality but was a conscious switch decided by the central bank. Thus, using policy rates as the key instrument allows the central bank to implicitly inject growth objectives in deciding what rates to maintain, even if on paper the main goal of infl ation targeting is infl ation reduction. Second, the escape clauses and the use of ‘core infl ation’ allow the Philippines to use similar instruments as the US Fed and to decide on policy rate changes based on the strength of the economy and not only on infl ation targets. The policy has been quite lax since world and domestic infl ation has not been very high between 2002 and 2006.

Figure 13.9 shows that since 2003, the core infl ation has always been


below the headline infl ation. Infl ation targets of the central bank were exceeded in periods when oil prices were rising in the world market (2004–05) – when infl ation rates went up beyond 6 percent and 7 percent. But policy rates were raised only infrequently (the last one in October 2005) since infl ation targeting was implemented in 2002. The justifi cations were: (1) core infl ation was below the headline infl ation; and (2) overshooting of the infl ation target was due to the oil price shock – a supply rather than a demand factor. Between 2002 and 2006, the central bank raised policy rates by only 75 basis points (due to the higher infl ation led by rising oil prices in 2005 and 2006) compared to the more than 400 basis points done by the US Fed in the same period. Of course, the Philippine policy rate started at a higher level of 9 percent compared to the US Fed rate of 1 percent.

13.3.3 Alternate Tightening and Loosening in 2007

2007 saw an initial monetary tightening but ultimate loosening of mon-etary policy. In the fi rst half of the year the government mopped up liquid-ity by allowing banks and their customers to park their money in special deposit accounts with the central bank, receiving interest rates equivalent to the repurchase rate. The offi cial reason is too much liquidity due to massive infl ows of remittances of overseas workers. What is left unsaid, however, is that the real danger of the high liquidity is that, coupled with little credit going to private businesses, the high liquidity could possibly lead to the creation of bubbles in the fast rising equities and real property markets.

But when infl ation was seen to be hovering around 3 percent and below

0

2

4

6

8

10

Qua

rter

ly a

vera

ge in

per

cent

(200

0 =

100

)12

1995 20061998 1999 2000 2001 200219971996 200520042003

Core inflationHeadline inflation

Source: Bangko Sentral ng Pilipinas (BSP).

Figure 13.9 Headline and core infl ation rates: fi rst quarter 1995 to fi rst quarter 2006


(way below the infl ation target of 4 percent to 5 percent for 2007), the central bank reduced their policy rates by as much as 200 points in the fi rst 11 months of 2007. An additional strong reason to do this is the alarm-ing appreciation of the peso due to the massive infl ow of remittances and portfolio infl ows.

In 2008 and 2009 when the global fi nancial crisis erupted, the peso started to depreciate due to strong capital outfl ows. In 2008, when food and fuel infl ation brought CPI infl ation rate in the Philippines to the teens (see Figure 13.9), the Bangko Sentral ng Pilipinas (BSP) increased policy rates by 100 basis points in the period from June 2008 to August 2008. When food and fuel infl ation waned in the fourth quarter of 2008 to the present period (March 2009), the BSP reduced policy rates by 125 basis points between December 2008 to March 2009. The current overnight deposit rate and overnight lending rates of 4.75 percent and 6.75 percent respec-tively, are the lowest since infl ation targeting started in January 2002.

13.4 THE NEED FOR A HOLISTIC VIEW AND AN ALTERNATIVE MONETARY POLICY

Even with the laxer monetary policy, this has not led to adequate credit expansion for investment needs, better employment prospects and a more stable external fi nancial sector. It is clear that the fi nancial, monetary, fi scal, external and real sectors are integrally linked and aff ect one another critically, and one cannot expect monetary policy alone to solve the macro problems.

13.4.1 The Recent Fiscal Crisis

From the mainstream point of view, the biggest challenge to the current Philippine macroeconomy after the Asian crisis is the large fi scal defi cits and public debt burden in the period 2002 to 2006. This problem is resur-facing in the global recession years of 2009 and beyond as tax collection is expected to fall with the economic slowdown and government spending will have to increase to pump prime the economy.

Fiscal surplus had been achieved by the Ramos administration before the Asian crisis. The Asian crisis, however, brought back fi scal defi cits. Table 13.1 shows fi scal defi cits worsening after the crisis, with the national government defi cit reaching more than 5 percent of GDP in 2002. Public sector borrowing requirements reached more than 6 percent in the same year as the defi cit of government corporations, especially that of the National Power Corporation, became larger.

285

Tab

le 1

3.1

Fisc

al a

nd p

ublic

fi na

nce

fi gur

es fo

r the

Phi

lippi

nes

Item

1998

1999

2000

2001

2002

2003

2004

2005

2006

Nat

iona

l Gov

t Rev

enue

s (P

Bill

ion)

462.

547

8.5

514.

856

7.5

578.

463

9.7

706.

781

6.2

979.

6

As %

of G

DP:

Nat

iona

l Gov

ernm

ent (

NG

) Rev

enue

s 17

.416

.115

.315

.614

.614

.814

.515

.116

.3

of w

hich

: Tax

15

.614

.513

.713

.612

.812

.812

.513

.014

.3N

atio

nal G

over

nmen

t Exp

endi

ture

s 19

.319

.919

.319

.719

.919

.518

.417

.817

.4N

atio

nal G

over

nmen

t Sur

plus

/Defi

cit

(−)

−1.

9−

3.8

−4.

0−

4.0

−5.

3−

4.6

−3.

8−

2.7

−1.

1

Publ

ic S

ecto

r Bor

row

ing

Req

uire

men

ts−

4.2

−4.

6−

5.2

−5.

2−

6.8

−6.

4−

5.8

−3.

4n.

a.

As %

of N

G R

even

ues:

NG

Deb

t Ser

vice

Pay

men

ts35

.642

.944

.348

.461

.973

.585

.183

.287

.2In

tere

st21

.622

.227

.430

.832

.135

.436

.936

.731

.7Pr

inci

pal

14.0

20.7

16.9

17.6

29.8

38.1

48.2

46.5

55.6

As %

of G

DP:

As o

f Ju

ly 2

006

Tot

al N

atio

nal G

over

nmen

t Deb

t56

.159

.664

.665

.771

.078

.279

.072

.369

.6D

omes

tic31

.932

.931

.834

.437

.139

.741

.540

.238

.2F

orei

gn24

.226

.832

.731

.333

.938

.537

.532

.031

.4

As %

of G

DP:

Tot

al P

ublic

Sec

tor D

ebt

94.6

101.

510

8.0

106.

011

0.2

118.

210

9.8

93.4

n.a.

Dom

estic

35.2

32.8

32.1

32.7

34.4

35.7

35.4

33.0

n.a.

For

eign

59.5

68.7

75.9

73.3

75.9

82.5

74.4

60.4

n.a.

Sour

ce:

Ban

gko

Sent

ral n

g Pi

lipin

as (B

SP);

Bur

eau

of T

reas

ury,

Nat

iona

l Eco

nom

ic a

nd D

evel

opm

ent A

utho

rity.


The gravity of the fi scal problem is due to the fact that the post-Asian crisis economic recovery from 1999 to 2005 had failed to improve tax eff ort. The tax eff ort peaked at more than 17 percent of GDP in 1996 and 1997 and consistently fell after that and bottomed at 12.5 percent in 2004 (Table 13.1). The falling tax eff ort despite signifi cant GDP growth forced the government to undertake substantial tax reforms to respond to down-grades by the rating agencies of the sovereign debt. The tax eff ort improved to around 14.3 percent in 2006 mainly due to the expanded coverage and higher value-added taxation (raised from 10 percent to 12 percent).

The falling tax eff ort and rising public debt burden after the Asian crisis brought about a serious situation wherein public debt service – principal and interest debt payments – made up 85 percent of government revenues in 2005 and 87 percent in 2006 (Table 13.1).3 The improving fi scal defi cits from 2003 to 2006 were brought about only because total non-debt expen-ditures of the government (as percentages of GDP) – especially social and economic services – were cut drastically as interest payments increased. Table 13.1 shows total national government expenditures falling to 17.4 percent of GDP in 2006 from more than 19 percent in 1998. Figure 13.10 shows the gravity of the situation as economic and social services as percentages of GDP continuously fell in recent years while the share of interest payments went up.

0

1

2

3

4

5

6

7

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Per

cent

Social services

Defense

Net lending

Interest payments

Economic services

General public services

Source: Department of Budget and Management (DBM).

Figure 13.10 National government expenditures, percentage of GDP


Despite improvements in the tax eff ort in 2006, the high public debt burden is expected to continue for some time as Table 13.1 indicates. Public investment (and other economic and social spending) has yet to recover from cutbacks, while private investments remain low despite high economic growth. In the fi rst quarter of 2007, despite a whopping 6.9 percent GDP growth, the tax eff ort fell to 12.2 percent (from 13 percent in the same period of 2006). The inability of revenue generation to keep pace with economic growth does not bode well for public investment and the infrastructure needs of the country, which have been neglected in recent years due to fi scal tightness. This is most relevant in 2009 when pump-priming is vital to prevent the economy from being adversely aff ected by the global economic recession, which already has reduced exports by 40 percent to 50 percent in the fi rst two months of 2009.

During the period of high issuance of government securities to fi nance the defi cits in 2002 to 2006, there was no signifi cant rise in the treasury bill rates since the fi nancial institutions – teaming with liquidity – preferred government securities to private loan assets (after having been burnt in the Asian crisis). This enabled the Treasury to issue treasury bonds at even lower rates.

13.4.2 More Open Capital Accounts and More Volatile Exchange Rate Movements

Figure 13.2 shows that since the 1980s the exchange rate has been very volatile. Obviously the exposure to fi nancial and capital account liberal-ization has created a Pandora’s box of dangerous short-term capital fl ows and volatile movements.

Even the years following the Asian crisis – from 1999 to 2005 – have been periods of volatile short-term portfolio fl ows with international capital shunning the country due to political instability and fi scal prob-lems. Starting 2005, short-term capital came into the country due to the increases in VAT and expectations of the narrowing of the fi scal defi cits. But they intermittently fl owed out also due to political instability, increases in the US interest rates in 2005 and early 2006, and ultimately the sub-prime fi asco in the US that brought stock markets tumbling down for high-yielding emerging markets like the Philippines.

In late 2006 and throughout 2007 the strong appreciation of the peso due to remittances of overseas workers, short-term capital infl ows and a generally weak dollar internationally is causing major concerns among exporters, overseas workers and domestic manufacturing competing with imports. On the other hand, other domestically oriented sectors are happy as infl ationary pressures are stemmed and economic growth is stimulated.


Trade defi cits may be expected to deteriorate (although this has not hap-pened very signifi cantly yet). But the latest picture of the Philippines has changed from one of high current account defi cits during periods of high growth (see Figure 13.6) to consistently positive current account balances (despite trade defi cits) because of the massive infl ux of foreign exchange earnings of the overseas workers.

13.4.3 Low Financial Confi dence and Credit to the Private Sector

The post-Asian crisis period has been marked by low fi nancial confi dence due to the trauma of non-performing assets suddenly increasing and due to stringent fi nancial supervision to achieve higher capital adequacy ratios and loan loss provisions. Figure 13.11 shows M2 (money plus quasi-money) and domestic credit as percentages of GDP.

It is clear that the Philippines has not achieved substantial fi nancial deep-ening as M2 and domestic credit fell drastically from its 1997 peak. The lack of fi nancial deepening is partly caused by the various recession and monetary

0.000

10.000

20.000

30.000

40.000

50.000

60.000

70.000

80.000

90.000

1948

1951

1954

1957

1960

1963

1966

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

2005

Per

cent

Domestic credit, % of GDPM2, % of GDP

Source: International Financial Statistics (IMF).

Figure 13.11 M2 and domestic credit, percentage of GDP


tightening periods. Another big reason is the decline in fi nancial confi dence due to the fi nancial crises. This is very clear in the decline of domestic credit (as percent of GDP) from the mid 1980s until 1992 due to the fi nancial and economic collapse of 1984–85. It is happening again in the post-Asian crisis period as both domestic credit and M2 as percentages of GDP declined from 1998 until 2004. But the picture of the Philippines in 2000 to 2008 changed from one of high current account defi cits during periods of high growth (see Figure 13.13) to positive current account balances (despite trade defi cits) because of the massive infl ux of foreign exchange earnings of the overseas workers. The positive current account balance is being threat-ened in 2009 and beyond as the global economic recession has signifi cantly reduced exports and is threatening the remittances of the overseas Filipinos workers who are at risk of losing their jobs.

The Asian crisis increased fi nancial regulations on banks. The Basel international standard for minimum capital adequacy (net worth to risk asset) ratios of banks is currently at 8 percent. The Philippines has been more stringent and imposed a minimum capital adequacy ratio of 10 percent starting in 2001. The actual current average capital adequacy ratio for Philippine banks ran to a high 15 percent to 16 percent in 2005. This situation, together with the perennial political crisis, has so far discour-aged banks from aggressively lending to the private sector. The banking system is awash with liquidity with a strong appetite for government securities – whether peso or dollar denominated – rather than private lending. This will constrain investments and employment generation in an economy with still underdeveloped long-term capital markets.

13.4.4 Persistently High Unemployment and Low Investment Rates

Figure 13.4 shows persistently high unemployment in the latest economic recovery period of 2000–04 despite positive economic growth. Figure 13.12 gives more detail on the employment picture. It shows that a major trend in the employment picture is the downward employment absorption capacity of agriculture and the low and stagnant employment absorption capacity of industry. The only sector adequately absorbing the growing labor force is the service sector.

It must be pointed out that the industrial and agricultural sectors are the main tradable sectors. With increased trade liberalization, globaliz-ation and competition among countries, these sectors are now exhibiting increasing output-employment ratios as output increases are not matched by equivalent employment increases. This means labor shedding and labor-cost cutting in areas whose products are facing stiff competition from imports. Thus, services, which is largely a non-tradable sector, becomes


the biggest absorber of employment. But this is not enough to absorb the expelled labor from the tradable sector, and the new labor force entrants (see Lim and Bautista, 2006). Of course the result is that unemployment remains persistently high, hovering at 10 percent and 11 percent.

This situation leads us to explore whether a more employment-sensitive monetary policy can help alleviate unemployment and underemployment through an integrated scheme of credit allocation to labor-intensive and employment-generating activities.

Related to this is the fact that investment rates have fallen since the Asian crisis (as is true for most of the East Asian economies). Investments as a share of GDP fell from above 24 percent before the Asian crisis in 1997 to 21 percent in 2000 and to 14.8 percent in 2006. The steady and continuous fall of the investment rate, partly due to the collapse in public and infrastructure investments due to the fi scal bind, does not bode well for future supply capacity and productivity improvements in the economy, even as the GDP growth rate has improved considerably in 2004 to 2007. This is another area where monetary policy may play a more active role.

A lax and accommodating monetary policy is even more urgent in the global recession period of 2009 and beyond when exports are plunging and

0

5

10

15

20

25

30

35

40

45

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Per

cent

ServicesIndustry

Unemployed

Agriculture

Source: Labor Force Survey, National Statistics Offi ce.

Figure 13.12 Employment by sector and unemployed, as percentage of labor force


consumption and investments are being threatened by possible declines in remittances from overseas Filipinos and losses in domestic business con-fi dence because of losses in the stock market, depreciating currency and adverse global conditions.

13.4.5 External Shocks and High Oil Prices

The volatile world oil price movements starting in the second half of 2004 has increased the average infl ation rate in the Philippines from a low of 3 percent in 2002 and 2003 to more than double in 2005 and 2006. The con-tinuing threat of higher oil prices may lead to tighter monetary policies at home and abroad.

Thus, a policy to respond to increasing world oil prices and return to higher world infl ation and interest rate regimes is now an urgent task. If this supply-led infl ation leads to a policy of fi ghting infl ation through demand-suppressing methods of increasing interest rates, recessionary pressures may again become a major monetary policy.

13.5 COMPONENTS OF AN ALTERNATIVE MONETARY AND FINANCIAL POLICIES

13.5.1 An Undervalued Currency and Tax-Based Capital Controls on Infl ows

The fi rst component of the recommended alternative scheme is a reason-ably pegged exchange rate regime, targeted at an undervalued level during good and healthy periods of global trade, and at a slightly overvalued level during bad times of global trade contraction. (This recommendation is consistent with the policy prescriptions of Frenkel and Taylor, Chapter 2, this volume.)

The peso has depreciated signifi cantly in 2008 and 2009 due to the massive capital outfl ows from the stock market and the signifi cant fall in exports during the current global economic recession. A weak currency is not healthy during collapsing global trade and shrinking export markets for this brings about stagfl ationary tendencies on the domestic economy without any positive eff ects from the export market. This puts the central bank in a dilemma as a lax monetary policy may aggravate the peso depre-ciation. This is true because of a fl exible exchange rate regime coupled with complete capital account liberalization.

A healthy economy requires an exchange rate that does not appreciate too much during good times when the economy and world economy is


strong. It also requires an exchange rate that does not go on a freefall col-lapse during the bad times. This requires capital controls and a reversal on the capital account liberalization that most emerging markets had under-taken in the 1990s. It would therefore be practical to propose a tax-based (or market-based) capital controls on infl ows like Chile and Malaysia, or quantitative restrictions on entry and exit of ‘hot money’ to and from the short-term equity and bond markets of the country, like China and India. Because of the diffi culty in undertaking this (note the recent unsuc-cessful Thai attempt to tax ‘hot money’), it is suggested that a regional ASEAN13 eff ort to undertake a common policy on capital controls be explored.

13.5.2 Incorporate Output and Employment Targets to the Current Infl ation Targeting Regime

If one envisions a transition away from the infl ation targeting regime, it would be wise to recommend that output and employment goals be explicitly stated as part of the objectives of monetary policy. After all, not adversely aff ecting economic activity has explicitly been cited by the central bank as one of the reasons why policy rates were not increased drastically despite infl ationary pressures from world oil prices in 2005 and 2006.

13.5.3 A More Active Role in Stimulating the Economy

The above monetary policy can have some benefi cial impact, especially on the current problems of high fi scal defi cits, lack of fi nancial confi -dence and unemployment. The more accommodating monetary policy may be complementary to the moves to improve and increase fi nancial loans in the post-Asian crisis period, and to off set the natural conserva-tive tendencies in credit expansion due to higher capital adequacy ratios and loan-loss provisions. It is crucial in the current economic slowdown brought about by the global economic recession. Furthermore, this will create a better atmosphere for involving credit allocation in employment generating activities to be discussed in the next section. Finally, since oil prices have returned to normal levels and world interest rates are low in the current period of global economic recession, the more accommodating policy may allow some room for monetizing the fi scal defi cits inasmuch as the pass-through of monetary increases to infl ation has been accepted as weak and unstable. Fiscal expansionary policies – especially in social, economic and infrastructure spending – are vital in returning the system to quality and employment-generating growth. And due to the high debt-


to-GDP and debt service ratios of the Philippines, it is diffi cult to support fi scal expansion with higher public debts. Monetizing the fi scal defi cit will be diffi cult to implement practically since the current authorities and the IMF have succeeded in institutionalizing the non-monetization of fi scal defi cits. Strong political will from the national government and monetary authorities will again be required.

13.5.4 Credit Allocation to Stimulate Investments in High Value-Added and Employment Generating Sectors

Inasmuch as unemployment is a major problem in the post-Asian crisis period and investment rates are low particularly in sectors that may have positive externalities for the economy and inasmuch as the fi nan-cial sector is reluctant to use market mechanisms to lend to the private sector, it is worth considering whether targeted credit programs can help in alleviating the unemployment and low investment problem. However, the Philippines since the 1980s has moved away from subsidized and targeted credit schemes as the fi nancial liberalization school has pre-dicted distortionary and adverse eff ects from such policies, and because the bad experiences with the Marcos administration has convinced many economists and technocrats that subsidized and targeted credit to potential cronies is detrimental. Relying more on the fi nancial markets and the private sector, they fi gure, is more benefi cial than government interventions. Thus, by September 2002, the central bank rediscount window has been liberalized to allow a generalized and uniform access to the facility by all sectors of the economy at market rates. The use of the facility has been reoriented for money supply management (comple-menting open market operations) instead of selective credit allocation (such as to exports and small-scale industries) and development fi nanc-ing (Guinigundo, 2005).

There are some targeted credit schemes outside the scope of the central bank administered by the Land Bank of the Philippines (LBP) and Development Bank of the Philippines (DBP) targeted at agricultural cooperatives, farmers’ groups and small-scale industries with funds from the Department of Agriculture (DA), Department of Agrarian Reform (DAR), other agencies and multilateral organizations. The Department of Trade and Industry (DTI) also has some credit lines targeted to small- and medium sized enterprises (SMEs). The current system has become rather schizophrenic as the formal system and the big government agencies such as the National Economic Development Authority (NEDA) and the central bank promote fi nancial liberalization and reduction of targeted credit in the formal sector. But the Arroyo government currently promotes


microfi nance lending by government-supported agencies as one of its key anti-poverty strategies.

Although there are some striking successes by some rural banks, coopera-tive banks and other microfi nance units in providing credit to develop local economies and some economic sectors, this is not being mainstreamed. The targeting of SMEs and micro-enterprises, largely in the informal sector, for microfi nance has positive aspects inasmuch as these entities have high employment generating potentials. However, the provision of microfi nance has oftentimes been politicized and lacks a holistic approach of providing the SMEs and micro-enterprises access to markets and link-ages in the formal sector, access to technology and good management practices, skills development and technical assistance for product develop-ment. This, unfortunately, is consistent with the overall trade and industrial policy that the government has been following for several decades, which prohibits ‘picking winners’ and frowns at promoting priority economic sectors (closely subscribing to the policies of the ‘Washington Consensus’ view and World Trade Organization (WTO) rules). Thus credit allocation in this setting becomes directionless and limited to providing small ‘liveli-hood’ programs for some targeted poor areas rather than to permanent, productive and growing industries and employment for a vibrant economy to benefi t a wider pool of poor and low-income families.

It is therefore recommended that a targeted credit allocation program be set up by the central bank and related institutions (such as key state banks) to give prioritized credit to key sectors that satisfy the following:

1. Exhibit strong potential for successful take-off , viability and high repayment but may require lumpy investments and/or suff er the ‘fi rst mover’4 problem.

2. Have strong employment generating eff ects, and/or strong interlink-ages with the other sectors of the economy, which will lead to multi-plier eff ects in the economy.

3. Have technology or knowledge spillover eff ects, as well as economies of scale.

4. Are part of an integrated set of industries that suff er coordination failure problems.

Market failures, endogenous growth and strategic trade theories are now mainstream economic theories that justify the above interventions. It is only the strong resistance of institutions that support the Washington Consensus that prevent developing countries from applying the policy implications of the new economic theories.


13.5.5 Heterodox Policies of Income Policies, Price Controls and Price Stabilization

Since most of the infl ationary pressures are from supply-side shocks, other means may be more productive than an immediate monetary demand-reduction response.

In many of the weather and natural calamity shocks the government has, for practical reasons and with some success, gone into temporary price con-trols and constant monitoring of basic foodstuff s and imposing heavy pen-alties on hoarding. Importation of agricultural products is also undertaken during periods of agricultural shortages. These policies should be continued and enhanced. Improvements are especially needed in the area of equitable, transparent and effi cient distribution of imported foodstuff s during periods of agricultural shortages. The ineffi cient distribution system and lack of transparency in many of the sales of the National Food Authority (NFA) during periods of food and agricultural shortages call for an overhaul in the system and more participation of the private sector in the distribution of temporarily imported foodstuff s during periods of shortages.

The oil price shocks had led many in the Philippines to question the deregulated structure of the oil industry. There is a perception that there is an asymmetry of quick price increases during times when world prices are rising, but slow and lagging price decreases when world prices are falling. The question of a domestic oil cartel has arisen (led by Shell, Caltex and a previously government-owned oil company, Petron), as many new players are fi nding it hard to compete in a setting of volatile world prices. Regulatory boards still regulate electricity charges and transportation fares, while the Department of Energy has clout to stop unreasonable price increases in oil and gasoline products. Thus, the importance of regulation in public utilities and oil/gasoline products become crucial in fostering competition and stop-ping cartel-like pricing. The Philippines also lacks anti-trust legislation that will remove monopoly and predatory pricing in key economic sectors.

If world oil prices increase signifi cantly again, it is recommended that an oil price stabilization fund (used moderately successfully in the 1970s by the Marcos government) be explored to tackle the situation. Equally important is the need to develop, in the medium and long run, alternative fuel sources and to re-enact laws giving special tax incentives for fi rms providing these alternative fuel sources.

The Philippines has an incomes policy that deliberately keep wages low and lagging behind price increases even as their monitoring and imple-mentation are grossly inadequate. With labor productivity increasing, this has led to lower infl ation but has also led to a long-run decline in real wages, which may have contributed to worsening income distribution and


poverty. A more balanced approach that this chapter recommends involves the following: (1) depending more on tripartite agreements derived by gov-ernment, labor and employers to impose agreed-upon ceilings on price and wage increases; (2) enact laws and rules that punish monopoly and cartel-like pricing (especially for the oil and telecommunications industries); and (3) do more monitoring and temporary regulation of key products, such as food, oil, telecommunications and rent-control housing, especially during periods of price instabilities.

The above set of alternative monetary policies points to a need for a change in mindset from a dichotomy between the fi nancial and real sectors to a viewpoint wherein the fi nancial sector is integrally linked and sup-portive of the real sector. The strategy is to coordinate and link the current targeted credit programs in the anti-poverty strategy campaign with the initiatives of the economic departments and agencies in developing and promoting key priority economic sectors that have high value-added, high technology spillover, multiplier eff ects and employment generating poten-tials. Successes in this arena will hopefully spill over to the formal and big business sector of the economy.

13.5.6 Testing the New Alternative

Inasmuch as the central bank already has a long-term and a short-term macroeconometric model, it is suggested that the new alternative be tested employing similar types of macro model. The new econometric model poses some challenges: (1) How to model a more discretionary monetary policy based on multiple objectives of both output/employment gener-ation and price stability (various simulation scenarios and quadratic loss functions might be used). (2) How to defi ne and model ‘overheating’ and excessive aggregate demand in the model to identify when some mon-etary contraction or aggregate demand reduction may be benefi cial to the economy and to identify when monetary contraction will be unreasonably recessionary. (3) How to model the incomes policies and temporary price controls. (4) How to model increased credit allocation and incorporating its impact on the growth of key economic sectors with high value-added and employment generation. No doubt diff erent scenario simulations are needed (with low, medium and high scenario assumptions) to make the models more useful to policy makers.

13.5.7 A Diffi cult Endeavor

The above alternative set of policies runs counter to the current macro policies of the Philippine government and go beyond policies traditionally reserved


for the central bank. Thus the adoption of the alternative monetary policy is a formidable endeavor that requires a change, not only in the infl ation target-ing regime, but also in the current macro framework of the country.

To achieve this, one does not need to drastically change the mandate and autonomy of the central bank. What is needed is perhaps a gradual but sure and progressive change in its attitude. The fi rst hurdle of shifting away from overemphasizing infl ation over growth has been achieved. The next step is just to make this more explicit. The hardest part is convincing not only the central bank but the entire government, business and fi nancial sectors of the country that high and sustained growth requires an active industrial policy supported by monetary and credit policies. The last hurdle is the most diffi cult one as it entails a rejection of the Washington Consensus and the adoption of a new and fresh perspective. This is what heterodox economists, including those involved in this volume, are working and striving for.

The current global fi nancial crisis – triggered by a deregulated fi nancial system existing side by side infl ation targeting regimes that jacked up global interest rates, which ultimately led to the subprime fi asco – may give many economists and policy makers cause to question the old mon-etary paradigm and pave the way for a framework that provides stronger and healthier links between the fi nancial/monetary sector and the real and productive sector of the economy.

NOTES

1. The author is grateful to Francis Dakila, Digna Paraso and Zeno Abenoja of the Bangko Sentral ng Pilipinas for their support and help. He would also like to thank Gerald Epstein and Erinç Yeldan for their positive comments and suggestions in improving this chapter. All mistakes are the author’s.

2. These are cash and short-term government securities which a bank is allowed to keep as required reserves.

3. The increase in 2006 was, however, due more to conscious prepayment of foreign debt (as the peso became strong) and domestic debt (as interest rates started to fall).

4. Haussman and Rodrik (2002) refer to the market failure problem of ‘information spill-over’ or ‘fi rst mover’ problem as the problem wherein the fi rst mover bears all the risks. If they succeed, others will imitate them and reduce their market share. If they fail, they will bear all the losses.

REFERENCES

Dakila, F. (2001), ‘Evaluation of alternative monetary policy rules: a simulation-based study for the Philippines’, PhD thesis, University of the Philippines, Diliman.


Frenkel, R. and L. Taylor (2006), ‘Real exchange rate, monetary policy and employment’, DESA working paper no. 19.

Guinigundo, D. (2005), ‘Infl ation targeting: the Philippine experience’, Bangko Sentral ng Pilipinas, Metro Manila.

Haussman, R. and O. Rodrik (2002), ‘Economic development as self-discovery’, National Bureau for Economic Research working paper no. 8592.

Lim, J. and C. Bautista (2006), ‘External liberalization, growth and distribution in the Philippines’, in L. Taylor (ed.), External Liberalization in Asia, Post-Socialist Europe, and Brazil, New York: Oxford University Press, pp. 267–310.

Lim, J. and M. Montes (2002), ‘Structural adjustment after structural adjustment, but why still no development in the Philippines?’, The Asian Economic Papers, 1 (3) (Summer), 90–119.

299

14. Monetary policy in Vietnam: alternatives to infl ation targetingLe Anh Tu Packard1

14.1 INTRODUCTION

One of the most important challenges facing policy makers is to determine how monetary policy should be conducted in order to meet their country’s national development goals. In recent years a growing number of central banks have convinced each other that the siren song of infl ation targeting is worth pursuing,2 even though a strong theoretical case that this mon-etary rule possesses superior welfare properties has yet to be established. Infl ation targeting calls for the ‘explicit acknowledgment that low and stable infl ation is the overriding goal of monetary policy’, which implies that a low infl ation target should have supremacy over other development objectives.3

For Vietnam, the quest for a pro-development monetary policy has become more urgent because the country is entering a new developmental phase that will be shaped by the terms of its accession to the World Trade Organization (WTO) and commercial treaties with its trading partners. Mindful of both the opportunities and risks that come with this phase, the Vietnamese government has been looking into macroeconomic and monetary policy guidelines to manage this period of unprecedented expo-sure to the world economy. In keeping with recent fashion among central banks, the State Bank of Vietnam (SBV) expressed interest in exploring the feasibility of infl ation targeting. However, the general consensus is that at present Vietnam does not meet the necessary conditions to implement an infl ation targeting regime because the central bank lacks adequate tools to carry out an eff ective infl ation targeting monetary policy. Additionally, other offi ces of economic management in the Ministries of Finance, Planning and Investment, Industry and Trade do not view infl ation tar-geting as a matter of urgency, nor are they willing to cede to the SBV that degree of power. There is also the larger question of whether an infl ation targeting regime is compatible with Vietnam’s development priorities, and whether it would help improve economic performance over the long run.


The context for discussing monetary policy in Vietnam is as follows: upon launching sweeping reforms during the 1990s, the country has gen-erally followed the East Asian ‘developmental state’ model. In the view of its political leaders, monetary policy should serve as a tool to meet the country’s socioeconomic development goals, which are rapid and sustainable growth, modernization, industrialization and poverty reduc-tion. According to the 1998 Law on the State Bank of Vietnam, the task of the central bank is to stabilize the value of the currency, secure the safety of the banking system and facilitate socioeconomic development in keeping with the nation’s socialist orientation (Kovsted et al., 2002, Thao 2004). SBV senior offi cials consider this a mandate to control infl ation and promote economic growth (Nghia, 2005). The implied assumption shared by policy makers and the public is that the nominal exchange rate and the domestic price level are closely linked. This link became highly visible during the Doi Moi (Renovation) reforms when Vietnam was grap-pling with hyperinfl ation and large depreciation of the parallel market exchange rate. Unlike many other developing countries, Vietnam has a strong domestic constituency for low infl ation because of its earlier trau-matic experience with hyperinfl ation, which heightened public sensitivity to price movements. That said, a strong preference for low infl ation and willingness to carry out policies to support a low infl ation environment does not mean that the central bank has a clear mandate to adopt formal infl ation targeting.

This chapter aims to contribute to the search for the right mix of macro-economic and monetary policies that can best serve Vietnam in the coming period of greater openness and intensifi ed competition in both domestic and global markets. It examines the factors that should guide monetary policy, taking into account the current state of Vietnam’s transition to a more market-oriented economy and the challenges posed by dollarization, fi nancial repression, informal and underdeveloped fi nancial markets, and rapid international economic integration. Not surprisingly, it fi nds that critical gaps in knowledge, institutional arrangements, tools and rules are impeding the eff ectiveness of monetary policy.

Sharing the view that the main task of the central bank should be to maintain macroeconomic prices that are conducive to rapid and sustain-able economic growth, an alternative to infl ation targeting is proposed. To support Vietnam’s transition to a more market-oriented economy, the central bank should instead target a real exchange rate (RER) that is stable and competitive. It is argued that this key relative price has a more powerful infl uence on the allocation of labor and capital, and on the com-position of domestic output, than administrative levers typically employed by centrally planned economies. Maintained over an extended period, a

Monetary policy in Vietnam: alternatives to infl ation targeting 301

stable and competitive RER promotes an effi cient allocation of resources and employment-creating growth, reinforces macroeconomic and fi nan-cial stability, and encourages fi nancial market development.

The chapter argues that a stable and competitive RER is a superior intermediate target for several reasons. First, this target clearly is consist-ent with the Law on the State Bank of Vietnam,4 which states that the SBV’s task is to stabilize the value of the currency. Since Vietnam now has multiple important trading partners, the value of the domestic cur-rency should be stabilized against a trade-weighted basket of currencies that take into account diff erences in their respective rates of infl ation. Second, it improves the transparency of monetary policy and strength-ens confi dence in the central bank’s ability to conduct monetary policy eff ectively. In other words, the central bank is assigned a task that is realistic and therefore doable. Third, a stable and competitive RER can contribute substantially to economic growth and employment creation if it is supported by complementary fi scal, monetary and industrial poli-cies. Fourth, it can have positive medium- to long-term impacts on struc-tural change and development through a variety of channels: resource allocation, changes in production techniques and growth of capital stock including stock of human capital (Frenkel and Taylor, 2005).5 Fifth, compared to a strict focus on infl ation targeting which tends to slow economic growth and lower employment growth (Epstein, 2003), a RER target is more likely to be a more eff ective stabilizing force and can do a better job in dampening output volatility during periods of global turbulence.

A stable and competitive RER’s long-term positive impact on resource allocation and the composition of output takes place through its infl u-ence, both direct and indirect, on key macroeconomic prices such as the domestic interest rate, the relative price of traded to non-traded goods, the relative cost of capital and labor, and the import-export price ratio. The RER, used in conjunction with appropriate commercial and indus-trial policies, can serve as a development tool in coordination with other monetary policy instruments to strengthen the economy’s overall com-petitiveness, increase aggregate productivity, maintain external balance, contain infl ation and stabilize asset markets (Frenkel and Taylor, 2005). International evidence from cross-country empirical research provides support for this view: instability (variability) of the RER is found to be negatively related to growth (Corbo and Rojas, 1995; Montiel, 2003), and overvaluation of the RER (in other words, an uncompetitive RER) has been linked with slower growth (Montiel, 2003; Razin and Collins, 1997).

The chapter is organized as follows: Section 14.2 provides a brief history


of Vietnam’s banking system. Section 14.3 analyses the macro economy from the perspective of identifying transmission mechanisms. Section 14.4 examines the issues surrounding the framework for monetary policy. Finally, Section 14.5 describes the merits of a stable and competitive RER as a superior alternative to infl ation targeting, and off ers some concluding comments.

14.2 VIETNAM’S BANKING SYSTEM: BRIEF HISTORY

From 1976 to 1989, like other centrally planned economies, Vietnam’s single-tier banking system was owned and controlled by the state. The SBV provided nearly all domestic banking services through a vast branch network. Bank lending was state directed, and credit ration-ing was imposed because fi nancial resources were scarce. During this period, SBV offi ces served as the interface between state planning, the national budget and state entities including some 12 000 state-owned enterprises (SOEs).6 Under central planning, the SBV was not required to carry out many traditional functions of commercial banking, such as credit analysis or risk management, and its main task was to ensure that fi nancial resources were allocated to economic units in accordance with the plan.

The quantity of currency outside banks was very high (the ratio of cur-rency outside banks to nominal GDP reached 9.2 percent in 1986) as the government attempted to monetize sharply rising fi scal defi cits as revenue growth failed to keep pace with rising expenditures (World Bank, 1991).7 SOEs in Vietnam lacked fi scal discipline as they operated under the soft budget constraint that was common among socialist countries. To cir-cumvent credit rationing, they engaged in unauthorized credit creation through various means such as abuse of the check payment system and use of supplier credits8 as a substitute for borrowing in credit markets. These practices had infl ationary consequences, created fi nancial problems for the SBV and contributed to deterioration in the consolidated balance sheets of SOEs.

Before money and capital markets were established during the 1990s, household liquid and semi-liquid assets mainly consisted of the domestic currency, gold, hard currency notes, and easily tradable commodities, such as rice. Remittances from overseas Vietnamese9 contributed to the dollarization of the economy. Continued eff orts by households and other economic agents to protect themselves from infl ation by reducing their domestic currency holdings (causing the ratio of currency outside the


banking system to GDP to decline from 9.2 percent in 1986 to 6.6 percent in 1988) only worsened the infl ationary spiral.

The 1987–89 macroeconomic crisis and hyperinfl ation provided the impetus for the comprehensive and coordinated Doi Moi reforms. In 1988 the Prime Minister signed Decree No. 53/ND which ended the monobank system and created a two-tier system consisting of the SBV as the central bank and four state-owned commercial banks (SOCBs). In addition, the government ended the state monopolies on fi nancing foreign trade and on providing long-term fi nance. The intent was to increase management autonomy and responsibility, and to introduce the pressure of competition in order to improve bank performance (ibid.).

In 1990 the government promulgated two banking ordinances for a two-tier banking system. These ordinances transformed the SBV into a central bank with oversight over the domestic banking system and provided the legal framework for commercial banks and other fi nancial institutions. The government liberalized entry into the banking system and lifted rules on sectoral specialization of the SOCBs. Commercial banks were given responsibility for the operation and control of their fi nances and implementation of universal banking activities. The decision to raise the interest rate for household deposits in the formal banking system increased confi dence in the domestic currency and encouraged households to deposit their dong assets in bank accounts (Figure 14.1). In 1989 RERs on households rose sharply, and encouraged a steady rise in the value of household deposits from VND 207 billion in March 1989 to VND 1348 billion by January 1990 (Figure 14.2).

As regards the current conduct of monetary policy, the broad division of labor currently is as follows: the government’s task is to prepare a plan

–4

–2

0

2

4

6

8

10

12

14A

pr-8

9

May

-89

Jun-

89

Jul-8

9

Aug

-89

Sep

-89

Oct

-89

Nov

-89

Dec

-89

Jan-

90

Feb

-90

Mar

-90

Apr

-90

May

-90

Nominal household demand deposit rate: monthlyNominal household 3-month deposit rate: monthlyInflation rate: monthly

Real household demand deposit rate: monthly

Sources: State Bank of Vietnam, World Bank (1991, p. 86, 1992).

Figure 14.1 Infl ation and interest rates, Vietnam 1989–90


for monetary policy that includes an infl ation forecast, while the National Assembly’s tasks are to set an annual target for the infl ation rate that is consistent with the state budget and economic growth objectives, and to supervise the implementation of monetary policy. As it currently stands, Vietnam’s monetary policy strategy is but a component of the broader fi ve-year socioeconomic development strategy that is formulated by the government and the ruling Communist Party. Within this framework, the SBV’s role is to come up with a concrete action plan for the banking sector, which includes setting targets for the amount of liquidity needed by the economy. Specifi cally, the SBV announces annual targets for total liquidity (M2 growth) and credit growth. However, the goal of keeping within credit growth targets does not appear to receive high priority, as actual credit growth in both 2004 and 2005 were well above the targeted growth rate of 25 percent.

The SBV is also tasked with stabilizing the exchange rate,10 an appro-priately vague mandate that ought to direct the authorities to maintain a trade-weighted infl ation-adjusted exchange rate that promotes sustainable (non-infl ationary) growth over the medium to long term. Like China and Singapore, the Vietnamese currency is offi cially pegged to a basket of cur-rencies, and like China, the authorities do not disclose the currency compo-sition in the basket nor the basket weights. From January 2000 to December 2006, Vietnam’s real eff ective exchange rate has depreciated measurably, especially in comparison to China and its South East Asian neighbors (Figure 14.3). This depreciation corresponds to a moderate acceleration of Vietnam’s GDP growth rate from 6.8 percent in 2000 to 8.4 percent in 2005, and a somewhat more rapid acceleration of the infl ation rate.

0

200

400

600

800

1000

1200

1400

1600

0

2

4

6

8

10

12

14

Apr-89 Jun-89 Aug-89 Oct-89 Dec-89

Real household 3-monthdeposit rate: monthlyHousehold monthlybank deposits

Sources: SBV, World Bank (1992).

Figure 14.2 Household deposits, Vietnam


14.3 DESCRIPTION OF THE MACRO ECONOMY

14.3.1 GDP and Macro Aggregates: Mechanisms of Adjustment

To carry out monetary policy eff ectively, policy makers in Vietnam need to have a much better grasp of the actual mechanisms of transmission and adjustment than they do at present. For example, the transmission of monetary policy via the interest rate channel is unclear because credit market segmentation, fi nancial repression and credit rationing add addit-ional layers of murkiness to the process. Through its short-term policy rate and commercial bank reserve requirement, the SBV is able to infl uence the commercial bank lending rate and activity levels of enterprises that borrow from the formal fi nancial sector. However, its infl uence over credit growth in the informal fi nancial sector and informal lending rates is not at all clear.11

Furthermore, the picture is obscured by the country’s ongoing struc-tural transformation that has led to a gradual fl attening of the investment spending curve, as investment spending becomes more sensitive to interest rate changes. This is illustrated in Figure 14.4, where the expected link between gross capital formation by enterprises and real lending rates has not emerged until after 1994. Even so, investment spending in Vietnam continues to be less sensitive to interest rate movements compared to other countries with more developed fi nancial sectors. This is because retained earnings continue to be the main source of fi nancing for business capital spending.

To elaborate further on the problem of incomplete information, both

70.00

80.00

90.00

100.00

110.00

120.00

130.00

Jan-

2000

May

-200

0

Sep

-200

0

Jan-

2001

May

-200

1

Sep

-200

1

Jan-

2002

May

-200

2

Sep

-200

2

Jan-

2003

May

-200

3

Sep

-200

3

Jan-

2004

May

-200

4

Sep

-200

4

Jan-

2005

May

-200

5

Sep

-200

5

Jan-

2006

May

-200

6

Sep

-200

6

Jan

2000

= 1

00

Vietnam China Thailand

Singapore Indonesia Malaysia

Figure 14.3 Real eff ective exchange rate


aggregate money supply and important elements of the money demand function are unknown (Hauskrecht and Nguyen, 2004) because the economy is partially dollarized and there is no reliable data about the quantity of US dollars and stock of gold outside the banking system that are used as a medium of exchange and a store of value.12 In particular, in Vietnam the aggregate ‘true’ stock of money is hard to estimate because it includes M2 (recorded by the SBV), foreign deposits held in banks, as well as two signifi cant unobserved variables, private sector foreign currency holdings and gold in circulation. It is also likely that the domestic and foreign currency will have diff erent velocities (ibid.) with diff erent trajec-tories, thus complicating the formulation based on the simple formula-tions that rest on the ‘quantity theory of money’.

Figures 14.5A and 14.5B present the velocity time path for currency outside banks (V1) and for total liquidity M2 (V2), which includes cur-rency outside banks, domestic currency deposits and foreign currency deposits. Two mutually off setting infl uences on velocity deserve mention. Ongoing structural reform of the fi nancial sector and improvements in the payments system increases velocity. At the same time, in a multi-currency economy a large-scale portfolio switch to the domestic currency can lower velocity. Figures 14.5A and 14.5B illustrates this: from 1991 to 1994 households and fi rms switched to the domestic currency and reduced their non-bank foreign currency and gold holdings as they trusted more the government’s ability to control infl ation. This brought about a decline in velocity and also led to greater monetary deepening.

–5.0

0.0

5.0

10.0

15.0

Per

cent

(%

)

20.0

25.0

30.0

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Lending rate, workingcapital (short-term), real

Inflation rate (CPI) Gross capital formation byenterprises, % of GDP

Figure 14.4 Infl ation rate, lending rate and gross capital formation by enterprises


Accurate tracking of domestic credit growth is also critical to the eff ec-tive conduct of monetary policy. This is yet another problem for the central bank, because key variables that aff ect fi nancial sector development and domestic credit growth in Vietnam are diffi cult to estimate. These include the magnitude of inter-fi rm credit as percent of aggregate credit creation and the quality of their accounts receivable (which may pose a signifi cant risk to the banking system). Finally complicating the task is the murky link between bank credit growth, the infl ation rate and actual borrowing by business enterprises (see Figure 14.4) due to the coexistence of formal and informal fi nancial markets, and the role of inter-fi rm credit.

As regards exchange rates, the impact of the central bank’s exchange

Figure 14.5A Velocity time paths, 1986–2002

Velocity 2: nominal GDP/M2Velocity 1: nominal GDP/currency outside banks

Inflation rate

–10123456789

0

2

4

6

8

10

12

14

16

1986 1988 1990 1992 1994 1996 1998 2000 2002

Per

cent

per

ann

um (

%)

Vel

ocity

rat

e

Velocity 2: nominal GDP/M2Velocity 1: nominal GDP/currency outside banks

Inflation rate

–100.0

0.0

100.0

200.0

300.0

400.0

500.0

600.0

700.0

800.0

900.0

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Per

cent

per

ann

um (

%)

Vel

ocity

rat

e

Figure 14.5B Velocity time paths, 1986–2005


rate policy on the real economy is diffi cult to determine because it is not possible to predict how Vietnam’s informally pegged exchange rate regime aff ects the growth of monetary aggregates. The SBV does not provide information on its interventions in the foreign exchange market, and there is no explicit sterilization policy. Yet, thus far, SBV actions to manage the informal peg do not appear to have negative consequences. The M2 growth rate has not been overly volatile and the infl ation rate has been low.

14.4 MACROECONOMICS AND INSTITUTIONAL FRAMEWORKS OF CENTRAL BANKING

14.4.1 Issues Surrounding Scope for Infl ation Targeting

Although there is interest in infl ation targeting on the part of the SBV, and a steady stream of international expert advice is provided on infl a-tion targeting for the SBV senior management, the consensus view is that at present the conditions to support a formal infl ation targeting monetary framework are not met. The reasons are evident (see Sections 14.1 and 14.2) when we consider the four main conditions outlined by the International Monetary Fund (IMF) that are deemed necessary to support such a framework (Carare et al., 2002):

The central bank has a clear mandate to make infl ation targeting the ●

primary objective of monetary policy and is publicly accountable for meeting this objective.The infl ation target will not be subordinated to other objectives and ●

monetary policy will not be dominated by fi scal priorities.The fi nancial system is developed and stable enough to implement ●

the infl ation targeting framework.The central bank has adequate policy instruments to be able to infl u- ●

ence infl ation.

At present, the SBV has limited scope to implement monetary policy using market-based indirect instruments to infl uence infl ation,13 although this has long been its declared objective, because fi nancial markets are thin and not well developed (the government securities market is segmented and illiquid). In addition, as explained above, the government is only at the early stage of building the necessary foundations (including timely access to a high frequency databank of key economic and fi nancial vari-ables needed for policy analysis) for developing a ‘reasonable understand-ing of the links between the stance of policy and infl ation’.


Nevertheless, there are many additional reasons why a rigid infl ation targeting framework is not appropriate for Vietnam, even if the conditions for infl ation targeting are met. First, it gives primacy to the wrong target (infl ation), forcing policy makers to operate in a framework that implicitly accords higher priority to infl ation than to other more pressing develop-ment objectives. For example, it obliges the central bank to automatically adopt a tightening stance whenever the infl ation indicator rises above its target range, or risk being branded as incompetent for failing to stick to the infl ation target. A rigid infl ation targeting framework also sets false standards for judging the quality of monetary policy, distracting policy makers from more serious and arduous eff orts to understand the actual workings of their economy.

Second, it is not so easy to determine what should be the right rate of infl ation to target, and the SBV may fi nd it much too tempting to simply follow the lead of other central banks even when that may not suit Vietnam’s particular circumstances.

Third, it sends the wrong message about what is needed to ensure good policy making. The implicit underlying justifi cation for infl ation targeting is that policy makers cannot be trusted to make sound policy decisions. The assumption is that they tend to give in to short-sighted political demands that can harm national social welfare over the long run. Therefore, to protect against this, policy makers must be bound to tight rules and explicit objectives, and they must be held publicly accountable to meeting these objectives. Although there are valid points in these argu-ments, tying their hands with rigid rules and wrong targets could very well push the economy onto a diff erent path that departs sharply from the country’s development goals.

14.4.2 Quality of Monetary policy

Surprisingly, the government managed to achieve favorable macro-economic results in spite of having to operate somewhat in the dark (given the critical gaps in information described in some detail above) with the crudest of monetary tools to infl uence aggregate demand. One explanation for this success is that these constraints did not prevent the government from pursuing the ‘right’ fi scal and monetary policies (see Section 14.2). The economic administration seems to have maintained macroeconomic stability and succeeded in keeping infl ation under control over a prolonged period, from 1990 to 2005. The gains from achieving this credibility can be seen in the progress made in monetary deepening, as the ratio of M2 to nominal GDP more than doubled from its nadir in 1993. The re-intermediation of foreign currency previously


held outside the banking system also indicates greater confi dence in the banking system.

Thus far, policy making credibility has remained strong – despite poor information and weak monetary instruments – because key fi scal and monetary policies have been well managed. Both the stock and fl ow of government debt (including debt denominated in foreign currency) have stayed at prudent levels, the outlook for fi scal balance remains healthy and the expected trajectory for the current account defi cit does not give cause for concern. However, the monetary authorities do not have adequate tools at present to protect the economy from exogenous shocks, which means that Vietnam’s vulnerability to external shocks will increase as its economy becomes more integrated with the global economy.

Moreover, the risk of policy error is likely to increase if the govern-ment fails to address the issue of critical gaps in information. As noted in Section 14.2, the coexistence of formal and informal fi nancial markets and the role of inter-fi rm credit in liquidity creation have made the relationship between bank credit growth and actual borrowing by business enterprises blurry and hard to predict. Without better information, the central bank runs the risk of misinterpreting data and may respond inappropriately, with dire consequences. For example, the SBV may attribute a ‘too high’ rate of credit growth to excessive monetary or fi scal easing, when in fact these high numbers may actually be the result of credit reallocation due to a secular rise in the formalization of credit and decline in informal sector lending.

14.5 INVESTIGATING ALTERNATIVES TO INFLATION TARGETING

14.5.1 The Real Exchange Rate is a Better Target

An alternative policy target should be consistent with and support wide-ranging development priorities. In other words, an important criteria for the monetary target is that it should play a positive development role and actively support the economy’s structural transformation. To this end, we need to specify the critical components of this economic transformation in order to sharpen monetary policy’s role, and to ensure that it becomes a coherent part of the nation’s development strategy.

Taking into account these considerations, a stable and competitive real exchange rate (SCRER) is regarded to serve as a better intermediate target because it helps to advance Vietnam’s national priorities, which


are rapid and sustainable growth, modernization, industrialization and poverty reduction. In contrast to infl ation targeting regimes that have been found to push many economies into a lower employment growth trajectory, SCRER targeting would contribute to growth and employment creation through its impact on resource allocation, rewarding fi rms that adopt forward-looking production technologies and encouraging them to develop promising new businesses.

To lay the foundations for sustainable growth, strategies that give priority to developing the medium to large enterprise (MLE) sector may be more eff ective than strategies that advocate concentrating resources on developing the SME sector. This is because mid-size enterprises have better capacity for learning and for technological innovation, and can create more jobs faster. Development of this sector will accelerate for-malization of the economy (enabling policy makers to better monitor economic activity), promote human capital development, technological development and development of management skills, and strengthen the competitiveness of domestic fi rms. Given that the long-term survival and growth of enterprises depends on their being able to maintain a healthy profi t rate, the stability of employment growth in the formal sector is closely linked to an environment that is conducive to MLE growth. An important aspect of this environment is that monetary policies send con-sistent signals to affi rm the basic stability of key macroeconomic relative prices including the RER. This is needed so that enterprises will be confi -dent enough to proceed with their investment plans in order to develop in areas that are most likely to be profi table.

A SCRER targeting framework for monetary policy is key to promot-ing rapid expansion of the MLE sector. This is necessary to ensure that employment in the formal sector (which is dependent on MLE growth) will increase at a rate that can absorb Vietnam’s rapidly growing labor force (about 1.2 million new entrants to the workforce every year). The economic well-being of Vietnamese workers depends on this, because wage rates in the formal sector are signifi cantly higher than informal sector wage rates. A SCRER target also helps the government to reduce its reliance on administrative levers to bring about desired changes in the economy. Offi cials have less justifi cation to yield to pressure from fi rms in import-substitution sectors for special protections. If fi rms in sectors such as paper, steel and cement are unable to survive and prosper in a favorable price environment created by a SCRER, the government should conclude that they are unlikely to achieve long-term commercial viability. Consequently, the economy would achieve better resource allocation if these fi rms were to close down their operations.


NOTES

1. Earlier versions of this chapter were presented to the May 2005 CEDES/Amherst Research Conference in Buenos Aires and the July 2005 Da Nang Symposium on Continuing Renovation of the Economy and Society. Financial support for this project has been provided by the Ford Foundation, UNDESA and the Rockefeller Brothers Foundation. My gratitude and thanks go to two anonymous referees, Gerald Epstein, Erinç Yeldan, Jaime Ros, Lance Taylor, Per Berglund, Dang Nhu Van and Phillipe Scholtes for their insightful comments and valuable ideas. I am responsible for all remaining errors and omissions.

2. Bernanke and Mishkin (1997). 3. However, Mishkin (2000), an advocate of infl ation targeting, acknowledges that price

stability is ‘a means to an end, a healthy economy, and should not be treated as an end in itself’ and that ‘central bankers should not be obsessed with infl ation control’.

4. Law on the State Bank of Vietnam is dated 12 December 1997. 5. Frenkel and Taylor (2005) emphasize that the RER must be kept at a stable and com-

petitive level for a relatively long period if the positive eff ects are to take place. The reason is that responses to the new (competitive) set of relative prices take time because they involve restructuring fi rms and sectoral labor market behavior. This takes place over time via changes in the pattern of output among fi rms and sectors, and adjust-ments in technology and organization of production.

6. In 1989 the SOE sector was made up of about 12 000 enterprises, of which 3100 were in industry; while the remaining were in trade, construction, agriculture and services. Most SOEs were provincial or district enterprises that were managed by the Industrial Bureaus of the provincial or district People’s Committees (World Bank, 1991). The reform of state enterprises, a key component of the Doi Moi reforms, subjected the SOEs to a hard budget constraint. By 1992 the number of SOEs fell by nearly half to 6545 enterprises, and their labor force was cut from 2.7 million to 1.7 million (IMF, 1998).

7. The expenditures included the costs of maintaining a large military force, direct subsi-dies to SOEs and indirect subsidies associated with price controls.

8. This is done by delaying or failing to repay credit extended by their suppliers which gen-erally were other SOEs. The Vietnamese term employed by SOE managers to describe this practice is chiem von nhau (conquering each other’s working capital).

9. This usually took place through informal channels due to unfavorable regulations gov-erning formal money transfers. Recipients were forced to take the money in Vietnamese currency at an exchange rate which eff ectively gave them half or sometimes only a third of the amount they could get in the open market (Beresford and Dang Phong, 2000).

10. SBV intervention in the foreign exchange market to support the targeted rate of exchange takes the form of buying and selling foreign currency or engaging in foreign exchange swaps.

11. As in other countries where credit market segmentation play an important role, fi rms in Vietnam that have access to the formal banking system become key actors in the process of credit creation. They act as fi nancial intermediaries to credit-constrained fi rms by providing the latter with trade credit. In other words, institutional factors help to turn inter-fi rm credit into an imperfect substitute for bank credit.

12. Both function as ‘a quasi second legal tender’ or ‘parallel currency’ in the economy (Hauskrecht and Nguyen, 2004). The government can track the quantity of currency outside banks and the quantity of dong and dollar deposits. However, the quantity of gold and hard currency held by households and other economic agents that are used as a medium of exchange and store of wealth is not known.

13. As regards indirect monetary policy instruments, during the mid 1990s the SBV intro-duced and has been using required reserves, refi nancing and discount lending facilities, open market operations and foreign exchange interventions. The refi nancing rate and the discount rate together defi ne the upper and lower band for the open market opera-tions rate although on occasion this rate does move outside the band (Camen, 2006).


REFERENCES

Beresford, Melanie and Dang Phong (2000), Economic Transition in Vietnam: Trade and Aid in the Demise of a Centrally Planned Economy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

Bernanke, B. and F. Mishkin (1997), ‘Infl ation targeting: a new framework for monetary policy?’, Journal of Economic Perspectives, 11 (2).

Carare, A., A. Schaechter, M. Stone and M. Zelmer (2002), ‘Establishing initial conditions in support of infl ation targeting’, International Monetary Fund working paper WP/02/102.

Camen, U. (2006), ‘Monetary policy in Vietnam: the case of a transition country’, Bank for International Settlement papers no. 31, accessed at www.bis.org/publ/bppf/bispap31t.pdf.

Epstein, Gerald (2003), ‘Alternatives to infl ation targeting monetary policy for stable and egalitarian growth: a brief research summary’, working paper no. 62.

Frenkel, R. and L. Taylor (2006), ‘Real exchange rate, monetary policy, and employment’, DESA working paper no. 19, United Nations, February; extended version accessed in Alternatives to Infl ation Targeting, 2, from the Political Economy Research Institute, accessed at www.un.org/esa/desa/papers/2006/wp/9_2006.pdf.

Hauskrecht, A. and T.H. Nguyen (2004), ‘Dollarization in Viet Nam’, Indiana University, Kelley School of Business, Department of Business Economics and Public Policy working paper, accessed at http://ideas.repec.org/p/iuk/wpaper/2004-25.html.

IMF (International Monetary Fund) (1998), ‘Vietnam: selected issues and statisti-cal annex’, IMF staff country report no. 98/30.

IMF (1999), ‘Vietnam: selected issues’, IMF staff country report no. 99/55.IMF (2003), statement by Sean Nolan, IMF Division Chief, Asia and Pacifi c

Department, at the Consultative Group Meeting for Vietnam, accessed at www.imf.org/external/np/dm/2003/120203.htm.

IMF (2004), staff report for the 2004 Article IV consultation.Kovsted, J., J. Rand, F. Tarp, N.D. Tai, N.V. Huong and T.M. Thao (2002),

‘Financial sector reforms in Vietnam: selected issues and problems’, CIEM/NIAS discussion paper no. 0301, Hanoi.

Mishkin, Frederic S. (2000), ‘Infl ation targeting in emerging market countries’, American Economic Review, 90 (2) (May), 105–9.

Montiel, Peter (2003), Macroeconomics in Emerging Markets, Cambridge: Cambridge University Press.

Monetary Authority of Singapore (2001), ‘Singapore’s exchange rate policy’, accessed 8 August 2006 at www.mas.gov.sg.

Nghia, L.X. (2005), ‘Orientations of the banking sector development strategy towards international integration’, paper presented at the 18 January seminar on Preparation for the 2006–2010 Five Year Social Economic Development Plan in the Banking Sector, Hanoi.

Razin, Ofair and Susan M. Collins (1997), ‘Real exchange rate misalignments and growth’, National Bureau for Economic Research working paper no. w6174.

Thao, N.D. (2004), ‘Strategy for the development of banking services in Vietnam in the context of international economic integration’, Project VIE/02/009 Banking Service Sector, Hanoi, August.


Thanh, V.T., D.H. Minh, D.X. Trong, H.V. Thanh and P.C. Quang (2001), Exchange Rate in Vietnam: Arrangement, Information Content and Policy Options, Hanoi: Central Institute for Economic Management, February.

Tho, T.V., N.N. Duc, N.V. Chinh and N. Quan (2000), Kinh Te Vietnam. 1955–2000. Tinh Toan Moi, Phan Tich Moi [Vietnam Economy. 1955–2000. New Estimates, New Analysis], Hanoi: Statistical Publishing House, December.

UNIDO (2004), ‘Vietnam Agenda 21. National Conference and Forum on Sustainable Development in Vietnam’, Document prepared by UNIDO Country Offi ce in Vietnam for the Forum on Sustainable Industry and Business, Hanoi, December.

World Bank (1991), ‘Transforming a state owned fi nancial system: a fi nancial sector study of Vietnam’, report no. 9223-VN, Washington, DC, April.

World Bank (1992), ‘Vietnam: restructuring public fi nance and public enterprises’, report no. 10134-VN, Washington, DC, April.

315

accountability 3, 4aggregate demand 16, 28, 42, 118, 124,

187, 190–91, 199 see also eff ective demand

anti-infl ationary policy 6, 75, 87anti-unemployment policy 73–5Argentina 23, 131, 179–202Augmented Dickey–Fuller Test 260

balance of payments 58, 272, 276, 279‘black’ market see parallel marketsBrazil 22, 131, 139–55, 158Brazilian Central Bank 139–43, 145,

147–8, 151–2

call money rate 260–61capital adequacy ratio 289capital controls 22–4, 41, 44, 55, 190,

193, 197–8, 233, 243–4, 248, 264, 269–70, 291–2

capital fl ight 139, 142, 243capital fl ows, international 40, 204,

215–18, 220, 223capital management techniques 18–24,

225, 233–5, 243–4central banks 3–24

credibility of 159–60, 166, 172, 173and development 5 net worth of 194–5

central bank independence 4, 8, 87, 187Chakravarty Committee 257Chile 24, 158, 190, 244China 24, 112, 249class 71–87class confl ict 71, 75, 93cointegration 163, 166Colombia 158Comitê de Política Monetária 141computable general equilibrium

model see general equilibrium model

conditional least squares 119Conselho Monetário Nacional 140contagion eff ects 139, 151 convertibility 181–3, 185, 187, 190corto 160, 167–9, 172crawling band 159crawling peg 45credit allocation techniques 20–21credit targeting 24currency crisis 139, 142, 147–8current account 139, 145, 152, 180,

271, 274–9, 288–9Czech Republic 77, 78, 80

defl ation 93, 97–101demand management 72, 86, 129–30devaluation 179–82, 187, 272–3, 275–7

see also exchange rate, real exchange rate

development channel 16 dirty fl oat 140, 153, 155disinfl ation 142, 145, 154, 171Doi Moi reforms 300, 303

East Asia 248, 269–70economic growth 132, 179–81, 191,

249, 257, 263see also infl ation and economic

growth, SCREReff ective demand 29, 32–3, 36

see also aggregate demand 28emerging market economies 158employment 24–5

targeting 18, 24, 227–47equilibrium, macroeconomic 34–7, 47expectations 181–2, 184–5, 192, 196–7exchange rate 28

appreciation 139, 142, 145, 148, 152, 153, 155

depreciation 139, 142, 144, 145, 148, 151, 153, 155

Index


devaluation 179–82, 187and economic development 28, 30–31and economic policy 28, 37and external balance 28and fi nance 29and infl ation 29and monetary policy 159, 167–71over-valuation 301pass-through 8, 139, 171–4pegged 291and resource allocation 28, 29targeting 139–42, 146–7, 150, 155,

179–99, 252, 310–11volatility 180, 187, 191–2, 197see also real exchange rate, nominal

exchange rateexpansionary policy 72–3, 75expansionary fi scal contraction 203export sector 108

see also tradable goodsexternal balance 29external fragility 220, 223

Federal Reserve 41fi nancial crisis 146fi nancial instability 5fi scal balance 23, 180–81, 187–9, 190,

284–7, 292–3fi scal dominance 3, 205, 214fi scal overhang 256fi scal policy 160, 190–91, 196–9fl exible exchange rate 160–61

see also exchange rateFrance 77–8, 80

GEAR (Growth, Employment and Redistribution) 228–9

gender 93–112gender bias 108–9

and central bank policy 111–12see also infl ation targeting

gender discrimination 95gender equality 110gender equity 110–12general equilibrium model 24, 211–23Granger Causality 260–62Greenspan, Alan 3

Haiti 121–2heterodox economics 72

Hodrick-Prescott fi lter 97, 99, 102 hyperinfl ation 117

IMF (International Monetary Fund) 5, 8, 9, 119

and conditionality 5 see also infl ation targeting

impulse response function 164, 238, 262–3

income distribution 4, 8incomes policies 21, 131–2, 160India 248–70infl ation

aversion 78–87confl ict theories of 75core 254, 281–3and economic growth 9, 56,

116–36expected 48and inequality 132and poverty 73–4, 76and unemployment 93–112

infl ation preferences 71–87infl ation rate targets 9, 18infl ation targeting 4, 41–2, 45, 55

alternatives to 172–4, 291–7and asset price stability 9and budget defi cit 149–50and central bank accountability 8and economic growth 10–11, 56,

116–36and escape clauses 281–2and exchange rate stability 16and foreign exchange reserves 15and gender bias 108–9, 111and IMF 5, 44and industrial policy 30–31and infl ation rates 8in open economies 44and price stability 3and primary fi scal surplus 150and real exchange rates 44–67and sacrifi ce ratios 9in small economies 55and trade balance 11–17and unemployment 11–12, 15, 24,

56, 71–87, 93–112infl ation targeting countries, list 6infl ation targeting, socially responsible

alternatives 17–24

Index 317

infl ation variability 158ILO (International Labour

Organization) 4India 23, 95–107interest rate 38, 45, 57, 60–61, 116–30,

159–60, 167, 170, 172, 183–5, 187, 192–8

see also real interest rateinterest rate parity 192–3Ireland 95, 122, 135Israel 95, 131, 135Instituto Brasileiro de Geografi a e

Estatística 141International Social Survey Program

77–92

J-curve 34Jacobian matrix 54, 64jobless growth 210–11

Kaldor 35Keynesian models 49, 50, 60

labor intensity channel 16labor market 29, 32–3Lebac 182–90Lerner Symmetry Theorem 30liberalization 4logistic regression 81, 83–6Lula 144–5, 147

macroeconomic channel 16macroeconomic instability 5Malaysia 24, 244Mexico 18, 55, 95, 131, 135, 158–78Minsky, Hyman 39monetarism 117monetary base 179, 185–90, 192, 194,

195monetary policy 41–2, 45, 93–4, 97,

102, 105, 110–11, 132, 182, 187, 190, 192–3, 197–9

pro-development 299money supply growth 94, 97,

105–6 and women’s employment

108–10monetary targeting 159, 190, 251–2,

256, 271–2, 275, 279–82modifi ed 280

macroeconomicsnew classical 71, 73new consensus 49, 55New Keynesian 72see also post-Keynesian economics

and structuralist macroeconomics

NAIRU (non-accelerating infl ation rate of unemployment) 72,117

narrow money 252, 260–61Nicaragua 122nominal anchor 258nominal exchange rates 37–39, 41, 42,

45, 46, 182, 192, 193, 195–7and middle income countries 39and targeting 159see also exchange rate, real exchange

ratenonlinearity 119non-tradables see non-traded goodsnon-traded goods 28, 30–36North American Free Trade

Agreement (NAFTA) 165

open market operations 45, 57outlier 119, 120–21output gap 166–7output-infl ation trade-off 158output variability 158

parallel market 300pass-through 163, 165–6, 171–4

see also exchange ratePeru 131, 158Philippines 23–4, 271–98Phillips Curve 71, 117, 166, 278political economy 71–2, 75–6

and gender 93–112policy space 7post-Keynesian economics 49, 72potential output 152–5poverty 73–4

alleviation of 249price rigidities 58price stability 9, 11, 14primary surplus 141,150primary surplus target 203, 210, 219,

223–4


productivity 29, 31–3, 35–7prudential regulations 198–9

rational expectations 71–2, 86real exchange rate 28–42, 37–68,

94, 97, 105–9, 179–202, 263–4, 269–70, 300

appreciation of 32, 39, 55, 171and capital fl ows 40–41and central bank intervention 39,

41depreciation of 31, 166, 168, 171,

180, 189, 198and employment 36, 65fi xed 55and output 163–5overvaluation 301and targeting 44, 45, 173, 251–2, 264and wage share 53see also exchange rate, nominal

exchange ratereal interest rate 94, 97, 102–6, 109,

141–3, 145–8, 150, 152, 155, 206, 208, 212, 220, 232

see also interest raterelative infl ation aversion see infl ation

aversionremittances 283–4, 287, 289, 291, 302Reserve Bank of India 248reserve requirements 45, 52Ricardo-Viner Model 30root mean square error 261Rwanda 122

sacrifi ce ratio 8, 97SCRER (stable and competitive real

exchange rate) 16, 17, 180–81, 187, 189–92, 196–9, 310–11

seigniorage 195

SELIC (Sistema Especial de Liquidação e Custodia) interest rate 141, 143, 147

Singapore 95, 100–108stagfl ation 105state-owned enterprises 302supply shock 10, 146South Africa 24, 55South Korea 24, 122State Bank of Vietnam 299–301sterilization 188–90, 194, 196–8structuralist macroeconomics 16–17,

44–5surplus labor 250Sweden 131

Taylor Rule 49, 57Tinbergen Rule 194tradables see traded goodstraded goods 28, 30–36, 96, 108–9transparency 3, 8, 18trilemma 8, 17–18, 39, 41, 45, 47, 52,

140, 181, 192–4Turkey 22–3, 203–26twin targeting 203–26

uncovered interest parity condition 193

under-employment 232unemployment, male 96unemployment, female 96unemployment preferences 71–87

VAR (vector auto-regression) models 236–38, 262

Vietnam 23, 299–314volatility 8, 9, 15

Zimbabwe 122

Beyond Inflation Targeting

Documents

Transcript of Beyond Inflation Targeting