The European Central Bank: Credibility, Transparency, and Centralization

T E C B C           , T           ,    C             Jakob de Ha...

Author: Jakob de de Haan | Sylvester C. W. Eijffinger | Dr Sandra Hamella

17 downloads 829 Views 1019KB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

T E C B C           , T           ,    C            

Jakob de Haan, Sylvester C.W. Eijffinger, and Sandra Waller

Book Series

The European Central Bank

CESifo Book Series edited by Hans-Werner Sinn Marc Gradstein, Moshe Justman, and Volker Meier The Political Economy of Education: Implications for Growth and Inequality Assaf Razin and Efraim Sadka, in cooperation with Chang Woon Nam The Decline of the Welfare State: Demographics and Globalization Jakob de Haan, Sylvester C. W. Eijffinger, and Sandra Waller The European Central Bank: Credibility, Transparency, and Centralization

The European Central Bank Credibility, Transparency, and Centralization

Jakob de Haan, Sylvester C. W. Eijffinger, and Sandra Waller

The MIT Press Cambridge, Massachusetts London, England

© 2005 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. MIT Press books may be purchased at special quantity discounts for business or sales promotional use. For information, please email [email protected] or write to Special Sales Department, The MIT Press, 5 Cambridge Center, Cambridge, MA 02142. This book was set in Palatino by SNP Best-set Typesetter Ltd., Hong Kong and was printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Haan, Jakob de. The European Central Bank: credibility, transparency, and centralization / Jakob de Haan, Sylvester C. W. Eijffinger, and Sandra Waller. p. cm. – (CESifo book series) Includes bibliographical references and index. ISBN 0-262-04226-6 (alk. paper) 1. European Central Bank. 2. Banks and banking, Central–European Union countries. 3. Monetary policy–European Union countries. I. Eijffinger, Sylvester C. W. II. Waller, Sandra. III. Title. IV. Series. HG2976.H325 2005 332.1¢1¢094–dc22 2004059701 10 9 8 7 6 5 4 3 2 1

Contents

Series Foreword Preface ix

vii

1 Introduction 1 1.1 The Start of the Monetary Union 1.2 Main Issues 2

1

2 The ECB: Structure, Strategy, and Policy 9 2.1 Introduction 9 2.2 Structure of the E(S)CB 10 2.3 Monetary Policy Strategy of the ECB 11 2.4 Monetary Policy of the ECB 16 2.5 ECB’s Evaluation of Its Monetary Policy Strategy 2.6 Conclusions 25

21

3 The ECB’s Strategy: An Assessment 27 3.1 Introduction 27 3.2 Predictability of Policy Decisions 27 3.3 Inflation Objective 36 3.4 Role of the First Pillar 47 3.5 Inflation Forecasts 56 3.6 External Value of the Euro 57 3.7 Asset Price Inflation 74 3.8 Interest Stepping versus Smoothing 78 3.9 Conclusions 80 4 Transparency, Accountability, and Credibility of the ECB 83 4.1 Introduction 83 4.2 Transparency and Disclosure: Theory and Evidence 85

vi

Contents

4.3 4.4 4.5 4.6 4.7

Transparency of the ECB 94 Accountability of the ECB 108 Blinder’s Survey on Credibility 112 Credibility of the ECB 117 Conclusions 123

5 Centralization or Decentralization 125 5.1 Introduction 125 5.2 The Decentralized Eurosystem 125 5.3 Risks of Decentralization 130 5.4 Diverging Business Cycles in the Euro Area? 137 5.5 Different Inflation Rates in the Euro Area 148 5.6 Differences in Monetary Policy Transmission and Financial Structure 160 5.7 Conclusions 167 6 New Member Countries 169 6.1 Introduction 169 6.2 Central Bank Independence and Convergence 170 6.3 Implications of Enlargement of the Monetary Union 6.4 Conclusions 210 7 Options for Reform 213 7.1 Introduction 213 7.2 Options for Reform 215 7.3 The ECB Proposal 220 7.4 Conclusions 223 8 Conclusions

225

Notes 229 References 239 Index 261

192

CESifo Book Series Foreword

This volume is part of the CESifo Book Series. Each book in the series aims to cover a topical policy issue in economics. The monographs reflect the research agenda of the Ifo Institute for Economic Research and they are typically “tandem projects” where internationally renowned economists from the CESifo network cooperate with Ifo researchers. The monographs have been anonymously refereed and revised after being presented and discussed at several workshops hosted by the Ifo Institute. Hans-Werner Sinn

Preface

This book offers our syntheses of the debate on the European Central Bank (ECB), focusing on two sets of issues: the transparency and credibility of the ECB, and its decentralized setup. We argue that these issues are relevant for evaluating the performance of the ECB since 1999. These issues play an important role in the discussion of the ECB’s monetary policy strategy and debate on institutional reform of the ECB. Parts of this book were presented at seminars and workshops at various places, including the Bank of England (London), CESifo (Munich), the ECB (Frankfurt), the University of Antwerp, the Center for European Studies at Harvard University, and the Bundesbank (Frankfurt). We would like to thank the participants in these seminars and workshops for their stimulating comments. Thanks are also due to Lex Hoogduin, Julius Horvath, and Jan-Egbert Sturm for their comments. We also would like to thank three anonymous referees for their very helpful suggestions. We also would like to express our gratitude to a number of our coauthors of papers that have, in one way or another, found a place in this book: Ivo Arnold, Helge Berger, Adam Elbourne, Petra Geraats, Robert Inklaar, Bas Kiviet, Linda Toolsema, Jan-Egbert Sturm, and Peter van Els. Special thanks are further in order to Maaike Beugelsdijk, Carin van der Cruijsen, David-Jan Janssen, Olaf de Groot, Richard Jong-a-Pin, Robert Inklaar, and Jörg Reddig for their assistance in gathering data and other research assistance. Last we would like to thank the Nederlandsche Bank for providing valuable data and Dana Andrus for editorial support.

The European Central Bank

1

Introduction

1.1 The Start of the Monetary Union On January 1, 1999, Europe entered a new era with the adoption of a single currency—the euro—by eleven of the fifteen member states of the European Union (EU). Greece joined the euro area in 2001. For the first time since the Roman empire a large portion of Europe now shares a common currency.1 The launch of the euro has created the world’s second largest single currency area in terms of economic size after the Unites States. It is the first time in history that countries of anything like this number, size, or global economic weight have gathered together on a voluntary basis to share a currency and pool their monetary sovereignty. With the start of the final phase of the Economic and Monetary Union (EMU), participating countries no longer have their own monetary sovereignty. Monetary policy in the euro area has been delegated to the European Central Bank (ECB). The Governing Council of the ECB is responsible for taking monetary policy decisions. This Council consists of the Executive Board of the ECB—made up of the president, the vice president, and four other members—and the central bank governors from the twelve euro countries. Together with national central banks, the ECB is part of the European System of Central Banks (ESCB). While the ECB is responsible for policy decisions, national central banks play a role in implementing monetary policy.2 The ECB’s primary objective, as laid down in the Maastricht Treaty, is price stability. In 1998 the ECB announced its interpretation of price stability: maintaining inflation below 2 percent in the medium term. It also developed a so-called two-pillar strategy to accomplish this objective. The instruments and procedures through which the Eurosystem—namely the ECB and the national central banks in the

2

Introduction

euro area3—conducts its operations are working smoothly. Financial operators learned to cope with the new money market in just a few weeks. Not so long ago, the idea of merging the monetary systems of the EU countries seemed fantastic. Nevertheless, a plan for doing so was drawn up by the Delors Committee. This committee—named after its chairman, Jacques Delors, who was at the time the president of the European Commission—was authorized by the European Council in June 1988 with the task of studying and proposing stages leading to economic and monetary union. Its report proposed three stages in a gradual move to monetary union, but stressed that the timing of each stage required a political decision. Many ideas of the Delors Committee found their way in the Maastricht Treaty, named after the place in the Netherlands where the leaders of the EU countries met to take a decision on EMU. Then, with the ink of this blueprint barely dry, the forerunner of the single currency, the Exchange Rate Mechanism (ERM) of the European Monetary System (EMS), almost collapsed, casting doubt about the project. Many skeptics asked what hope there could be for a monetary union among countries unable to keep national currencies aligned.4 Indeed, sometimes EMU was perceived as an ambitious project that would never fly, just like the emu, the large Australian bird. Still the European governments carried on. With the start of the monetary union in January 1999 their vision became reality. 1.2 Main Issues With the 1999 start of the monetary union, the ECB became responsible for monetary policy making.5 Chapter 2 describes the ECB’s monetary policy strategy and policy decisions in some detail. The monetary policy strategy of the ECB, which was announced on October 13, 1998, consists of three main elements: a quantitative definition of price stability, a prominent role for money in the assessment of risks to price stability (at the time called the first pillar), and a broadly based assessment of the outlook for price developments (the second pillar). In accord with the first pillar, the ECB uses a quantitative reference value for the annual growth rate of a broad monetary aggregate (M3) by which it can assess whether monetary developments pose a risk to price stability. The definition of price stability adopted was a year-on-year increase of the Harmonized Index of Consumer Prices

Introduction

3

(HICP) for the euro area, which was not to exceed 2 percent in the medium term. In December 2002 the ECB Governing Council decided to evaluate the strategy, taking into account the public debate and the outcomes of research undertaken by staff of the Eurosystem. The results of this evaluation were published in May 2003. The Governing Council agreed that in the pursuit of price stability, it would aim to maintain inflation rates below but close to 2 percent over the medium term. According to the Council, “this clarification underlines the ECB’s commitment to provide a sufficient safety margin to guard against the risks of deflation. It also addresses the issue of the possible presence of a measurement bias in the HICP and the implications of inflation differentials within the euro area.” The Council also decided that in the presentation of the arguments for a certain policy decision, the president of the ECB would start with what is now called the Economic Analysis (i.e., the previous second pillar). Most academic observers considered these decisions to imply that the first pillar had become less important. Although they were generally argued to be in the right direction, many such observers claimed that the steps taken by the Governing Council did not go far enough. Svensson (2003) stated, for instance, that “It would have been better to throw the reference value on the garbage heap of history where it belongs.” Chapter 3 offers our assessment of the ECB’s monetary policy strategy. We focus on issues like the definition of price stability, the predictability of policy decisions, and the role of the first pillar. In this chapter we also discuss the role of the external value of the euro in the ECB’s strategy. The rest of the book deals with two sets of issues that European central bankers are faced with. A first important issue for the ECB—as it is for every central bank—is its transparency. According to the Code of Good Practices on Transparency in Monetary and Financial Policies of the International Monetary Fund (IMF 2000a)6: The case for transparency of monetary and financial policies is based on two main premises. First, the effectiveness of monetary and financial policies can be strengthened if the goals and instruments of policy are known to the public and if the authorities can make a credible commitment to meeting them. In making available more information about monetary and financial policies, good transparency practices promote the potential efficiency of markets. Second, good governance calls for central banks and financial agencies to be

4

Introduction

accountable; particularly where the monetary and financial authorities are granted a high degree of autonomy. . . . In making the objectives of monetary policy public, the central bank enhances the publics understanding of what it is seeking to achieve, and provides a context for articulating its own policy choices, thereby contributing to the effectiveness of monetary policy. Further, by providing the private sector with a clear description of the considerations guiding monetary policy decisions, transparency about the policy process makes the monetary policy transmission mechanism generally more effective, in part by ensuring that market expectations can be formed more efficiently. By providing the public with adequate information about its activities, the central bank can establish a mechanism for strengthening its credibility by matching its actions to its public statements.

The ECB tries to be open. There are, for instance, regular press conferences, a monthly bulletin is being issued and the president of the ECB appears before the European Parliament four times each year. Recently the ECB also started to publish staff forecasts (i.e., projections) of inflation. All this has led Otmar Issing—member of the Executive Board and chief economist of the ECB—to conclude that “in terms of transparency, the ECB can emerge proudly from the comparison to any other leading central bank” (Issing 2001a, p. 14). However, financial market reports often suggest that the ECB has not been very successful in explaining its policies to financial market participants. For instance, in the Goldman Sachs Weekly Analyst of June 25, 1999, it was stated that “the ECB has not yet succeeded in giving an entirely clear and convincing message on monetary policy issues to financial markets.” In chapter 4 we analyze the transparency of the ECB. As follows from the citation of the IMF Code of Good Practices on Transparency, it is often argued that transparency may enhance the credibility of the policy maker. Although economists often base their models on the notion of credibility, little is actually known about it. For instance, how can a central bank become credible? Should a central bank be independent to be credible? In a survey among central bankers, Blinder (2000) asked his respondents how important credibility is. It appeared that there is a broad consensus among central bankers that credibility is very important. Blinder also asked his respondents how a central bank can build credibility. On a scale from 1 (unimportant) to 5 (of the utmost importance) the answer “granting a central bank independence” received an average score of 4.51, becoming the second highest score after “a history of honesty” (4.58). Do financial market participants agree with central bankers that credibility is important? And if so, why, and how can it be built? And how is the ECB rated in

Introduction

5

terms of credibility in comparison to, for instance, the US Federal Reserve and the Bank of England? Chapter 4 reports the outcomes of a survey similar to Blinder’s among private sector economists on the credibility of the ECB. The ECB is—at least in legal terms—politically independent. Not only does the Maastricht Treaty prohibit the ECB from taking orders from politicians, the members of the Executive Board are appointed for eight-year, nonrenewable terms, insulating them from pressures to please politicians in order to get re-appointed. National central bank governors have, at minimum, five-year contracts. As follows from the citation of the IMF Code on Good Practice on Transparency, independent agencies should be accountable. The final issue in chapter 4 is therefore the accountability of the ECB. The second main issue on which the book focuses is the decentralized structure of the central bank of the euro area. As will be explained in more detail in chapter 2, in the initial setup of the ECB presidents of national central banks have a majority in the Governing Council. This is unlike the US central bank structure, with just 5 members from the regional Federal Reserve Banks within the decision-making Federal Open Market Committee of 12 members. Also in comparison with the German Bundesbank—with 9 votes from the Land central banks within the Bundesbank’s Central Bank Council of 17 members—the ECB is very decentralized. Critics of the ESCB sometimes argue that its decentralization is a design flaw. For instance, The Economist stated that: The Governing Council is supposed to set interest rates according to conditions in the euro area as a whole, but there is a risk that national governors will be unduly influenced by conditions in their home country. Small countries may also carry undue weight in the system. . . . A weak centre, combined with strong national interests, could create conflicts that undermine the whole system’s credibility.7

There are at least three conditions that must be met for a common monetary policy to succeed without causing frictions among the members of the monetary union (Guiso et al. 1999). First, members must agree on the ultimate goals of the common monetary policy. This issue was largely settled through the Maastricht Treaty and the ensuing ratification process, leading to the adoption of price stability as the primary objective for the ECB. Second, the common monetary policy will be easier to implement if the member countries’ business cycles are

6

Introduction

aligned and if inflation rates are very similar. If various countries (or sizable regions) in the monetary union are not at the same points in the business cycle or have different inflation rates, decision making on the appropriate monetary policy stance becomes a difficult task. It is clear that countries in the euro area sometimes have different inflation rates and business cycles, although some authors have argued that monetary and economic integration will lead to more business cycle synchronization (e.g., see Frankel and Rose 1998). Third, the monetary policy transmission mechanism should operate in a similar fashion across the member countries of the monetary union. Differences in the transmission mechanism could mean that the appropriate size and timing of monetary policy decisions are difficult to assess. Moreover, if the burden of adjustment is not equally shared across countries, sizable distribution differences may create political tensions. Chapter 5 deals with these issues. In particular, in this chapter we examine the degree of business cycle synchronization in the euro area, the inflation differentials among the euro countries, and (the persistence of) differences in monetary transmission across these countries. Although the ECB is a very young institution, there is already a debate going on about its reform in view of the recent enlargement of the European Union. The new EU members (eight eastern European countries and two southern European countries) are expected to apply soon for euro area membership. Negotiations about EU membership started with Poland, Hungary, the Czech Republic, Slovenia, Estonia, and Cyprus in March 1998; in February 2000 negotiations were opened with Slovakia, Malta, Bulgaria, Romania, Latvia, and Lithuania.8 The European Union has outlined a three-step approach to the monetary integration of the accession countries. The applicants must first join the European Union, next enter the Exchange Rate Mechanism II (ERM II), and, finally, after fulfillment of the Maastricht convergence criteria, accede to the euro area. To qualify for EU membership, an applicant must have a strong market economy and stable institutions that guarantee democracy, the rule of law, and human rights. Furthermore the country must have largely adapted its laws to comply with EU legislation, known as the acquis communautaire, and have the ability to implement that legislation. In December 2002 the European Union decided that all applicant countries, except for Bulgaria and Romania, could become EU members in 2004. Bulgaria and Romania will enter at a later stage.

Introduction

7

At some point in time, the new EU members will join the euro area (chapter 6). Furthermore in the near future three EU members that are currently not members of the euro area might adopt the euro (the United Kingdom, Sweden, and Denmark). So membership in the Eurosystem could increase from the current 12 to 27. In the absence of modification of the current ECB statute, this enlargement could have severe consequences for the efficiency of monetary policy making in the euro area. A larger ECB Council would experience great difficulties in decision making. Without reform of the current ECB Statute, the size of the ECB Governing Council could increase from 18 to 33, making it by far the largest monetary policy-making institution among OECD countries. Due to this increase in membership, discussion and voting procedures will likely become more time-consuming and complicated. A second reason why the issue of reform of the ECB is on the political agenda is related to the decentralized nature of the ECB.9 Since almost all accession countries are small in economic terms relative to current euro area members, enlargement within the given institutional setup will significantly increase the degree of overrepresentation of the area’s smaller member countries in the Council in terms of relative economic size. For instance, in a monetary union with 27 members the current ECB statute implies that the representatives of its smallest 17 member states, representing only about 10 percent of the area’s aggregated GDP, could determine monetary policy decisions in the euro area. Overrepresentation, while not necessarily a problem per se, has the potential to introduce an unwelcome bias into the ECB’s decision making, if country representatives put at least some weight on national economic developments and these developments deviated notably from the behavior of euro area aggregates. Since the bulk of the accession countries are transition economies, they may be subject to idiosyncratic shocks and somewhat higher structural inflation than the core of the euro area. Chapter 6 therefore discusses business cycle synchronization, inflation differentials, and monetary policy transmission in the accession countries. There are various ways how the ECB can be reformed, like centralization (the Executive Board will become responsible for policy decisions), vote weighting (the vote of a national central banker depends on the size of the economy), representation (one central banker represents various central banks), extending regional central banks across

8

Introduction

national borders, and rotation (the governors of national central banks have rotating voting rights). The ECB has chosen for a system of rotation, in which national central banks have rotating voting rights, depending on the size of their economies. Many academics have criticized the ECB proposal and favor more centralization. Chapter 7 therefore presents various options for reform of the ECB. The final chapter of the book offers our main conclusions.

2

The ECB: Structure, Strategy, and Policy

2.1 Introduction The Maastricht Treaty made the European Central Bank (ECB) politically independent. Nowadays it is widely believed that central bank independence and an explicit mandate for the bank to restrain inflation are important institutional devices to ensure price stability. It is thought that an independent central bank can give full priority to maintaining low levels of inflation. In countries with dependent central banks other considerations (notably, re-election prospects of politicians and low levels of unemployment) can interfere with the objective of price stability. The example often mentioned is the German central bank. Because the Deutsche Bundesbank was relatively autonomous, Germany had one of the best post–Second World War inflation records among the OECD countries. The statute of the ECB is largely modeled after the laws governing the Bundesbank. This chapter starts with a brief description of the structure of the European System of Central Banks (ESCB) (section 2.2), followed by a discussion of the monetary policy strategy of the ECB (section 2.3) and actual policies since 1999 (section 2.4).1 The monetary policy strategy of the ECB has three main elements: a quantitative definition of price stability, a prominent role for money in the assessment of risks to price stability, and a broadly based assessment of the outlook for price developments. In December 2002 the ECB Governing Council decided to evaluate this strategy in the light of the experience, taking into account the public debate and the outcomes of research undertaken by staff of the Eurosystem. The results of this evaluation were published in May 2003. Section 2.5 discusses the results, while the final section offers our conclusions.

10

Chapter 2

ESCB

15 National central banks: 3 outs and 12 ins:

Governing Council

ECB

Executive Board

Executive Board of the ECB President Governors of the central banks of the Vice president countries in the euro area Up to four other members

General Council President Vice president and Governors of national central banks

Eurosystem Figure 2.1 Structure of the European System of Central Banks. (Source: Eijffinger and De Haan 2000)

2.2 Structure of the E(S)CB The European System of Central Banks consists of the European Central Bank and the national central banks. The ECB and the national central banks of the countries that have adopted the euro are often referred to as the Eurosystem. If all member states of the EU adopt the euro, the Eurosystem and ESCB will be synonymous. The ESCB is governed by the decision-making bodies of the ECB: the Governing Council, the Executive Board, and the General Council (see figure 2.1). The Governing Council of the ECB is the most important decision-making body of the ECB. It consists of the Executive Board of the ECB and the governors of the national central banks of the countries in the euro area. The Governing Council is responsible for formulating monetary policy, including decisions about intermediate objectives and interest rates. The Council decides, in principle, by simple majority. At the time of writing, 12 out of the 18 members in the Governing Council are presidents of national central banks, who have been appointed by their respective governments. However, when taking monetary policy decisions, the members of the Governing Council of the ECB are not expected to act as national representatives but rather in a fully independent personal capacity. The setup of the ECB is therefore based on the principle of “one person, one vote.” This way national central bankers are involved in monetary policy deci-

The ECB: Structure, Strategy, and Policy

11

sions, and in this respect the ECB differs from other central banks in federal countries (see chapter 5 for a further discussion). In looking ahead to the enlargement of the monetary union (see chapter 6), various proposals were put forward to reform the ECB. These reform proposals are discussed in chapter 7. The Executive Board of the ECB consists of the president, the vice president, and up to four other members. The Executive Board implements monetary policy decisions taken by the Governing Council. In doing so, it may give instructions to national central banks. The European Council appoints members of the Executive Board. Their term in office is eight years and is not renewable. A system of staggered appointments has been applied for the appointment of its first members. As long as not all EU member states use the euro as their currency, the General Council will also play a role within the ESCB. It consists of the president and vice president of the ECB and the governors of the national central banks of all EU member countries. As the central banks of the EU countries that have not (yet) adopted the euro continue to pursue national monetary policies, they will not participate in decisions related to the single monetary policy for the euro area. In the General Council they will, however, have the opportunity to discuss monetary policy issues and their exchange rate relations with the euro. 2.3 Monetary Policy Strategy of the ECB The primary objective of the ESCB is price stability.2 However, the Maastricht Treaty does not provide a specific definition of this objective. In October 1998 the Governing Council of the ECB defined price stability as follows: a year-on-year increase of the Harmonized Index of Consumer Prices (HICP) for the euro area, which does not exceed 2 percent in the medium term. Issing et al. (2001) point out that this quantification is in agreement with the stated preferences of European governments as expressed several times in the European Council’s Broad Economic Guidelines. The HICP is a comprehensive measure for inflation, reflecting the focus of the general public on consumer goods.3 The aim of an inflation rate “below 2 percent” clearly delineates the maximum rate of inflation deemed to be consistent with price stability. The wording “year-on-year increases” implies that persistent price decreases— that is to say, deflation in the measured price index—would not be

12

Chapter 2

considered to be consistent with price stability either. The Governing Council explicitly announced that price stability is to be maintained over the medium term, thereby acknowledging that price levels may be temporarily distorted by short-term factors. As Issing (2001a, p. 10) states: [T]he track record of the ECB . . . cannot be assessed on the basis of temporary deviations from price stability caused by external and unavoidable shocks. In view of the lags with which monetary policy affects the economy, a central bank cannot ensure price stability at each and every moment in time in the face of exogenous shocks.

The wording “for the euro area” highlights that areawide developments, instead of specific national or regional factors, are the only determinants of decisions regarding the single monetary policy. A yearon-year increase of the HICP for the euro area as a whole represents price stability, even if increases in national price indexes are above 2 percent per year. As Issing (2001a, p. 8) puts it: [R]egional inflation differentials [should] not be considered as a necessary source of problems . . . the differentials may be reflecting the appropriate economic adjustment to the temporary insurgence of disparities in productivity growth or to the realization of other asymmetric shocks.

At the time the monetary policy of the ECB was introduced, it rested on the strategy of two pillars (ECB 2004). The first pillar is a prominent role for money. As inflation in the long run is considered to be a monetary phenomenon, the ECB Governing Council announced a quantitative reference value for the annual growth rate of a broad monetary aggregate (M3). The focus on M3 was supported, according to the ECB (2001), by its favorable empirical properties such as a stable money demand function. Furthermore M3 is known to be a good indicator of inflation (Issing et al. 2001). A reference value for M3 growth was set at 4.5 percent but was not considered intermediate monetary target, “in order to avoid an automatic monetary policy reaction to fluctuations in M3 growth that may not be associated with inflationary pressures, but that may result, for example, from . . . financial innovations” (Issing 1999a, p. 20). Although the first pillar is sometimes presented as only referring to M3 growth (e.g., see Begg et al. 2000), the ECB concerns itself not only with the extent to which M3 growth deviates from the reference value, it also analyzes the underlying causes. Quite a lot of attention is currently being paid, for instance, to the growth rates of the components

The ECB: Structure, Strategy, and Policy

13

of M3, notably the growth rate of credit supplied to the private sector (see section 3.4 for further details). The second pillar is a broadly based assessment of developments regarding the outlook on prices and the treats to price stability in the euro area as a whole. A wide range of economic and financial indicators is used for this purpose. These variables can be grouped as (1) gap measures (i.e., measures of the discrepancy between output, or its factors of production, and their equilibrium values), (2) measures of cost pressure, (3) international prices and exchange rates, and (4) other asset prices. The indicators used are as follows: HICP and other price indicators in the euro area. This includes producer prices, which are a leading indicator for HICP inflation. In addition prices in the world markets for raw materials are closely monitored, especially the price of crude oil.4 Included in this category are the deflator of the euro area’s gross domestic product (GDP) as well as the deflators of its components. Also real estate prices are watched, where possible. •

• Indicators for the real sector. This includes the real GDP and its components, and also the production of the manufacturing, and other sectors. The capacity utilization rate also belongs to this category of indicators. • Confidence indicators. Among these are the economic sentiment indicator and various consumers and business sectors confidence indicators.

Indicators for the labor market. Employment and unemployment in the whole economy and by sector, as well as wages and unit labor costs (total and its components) and productivity belong to this category.

•

Exchange rates. The euro–dollar exchange rate, and the (real and nominal) effective exchange rates, are closely monitored by the ECB. The initial decline of the euro vis-à-vis the dollar has had quite some impact on inflation in the euro area (see section 3.6.1 for a discussion).

•

Financial market indicators. To this group of indicators belong various interest rates, as well as the term structure and stock market indexes. The long-term yield and the term structure are often considered to contain information concerning (expectations of) future inflation and economic growth.

•

Projections for inflation and economic activity. The Governing Council takes forecasts by others—like the Consensus forecasts—as well as the •

14

Chapter 2

Types of transactions Monetary policy operations Provision of liquidity

Absorption of liquidity

Maturity

Frequency

Open market operations Main refinancing operations

Reverse transactions

One week

Weekly

Longer term refinancing operations

Reverse transactions

Three months

Monthly

Fine-tuning operations

Reverse transactions Foreign exchange swaps Nonstandardized Foreign exchange swaps

Nonregular

Collections of fixed-term deposits Reverse transactions

Outright purchases Structural operations

Outright sales

Reverse transactions Issuance of debt certificates Outright purchases

Nonregular Standardized/ nonstandardized

Outright sales

Regular and nonregular Nonregular

Standing facilities Marginal lending facility

Deposit facility

Reverse transactions

Deposits

Overnight

Access at discretion of counterparties

Overnight

Access at discretion of counterparties

Figure 2.2 Monetary policy operations of the Eurosystem. (Source: ECB 2004)

projections of the staff of the ECB into account in its decision-making process (see section 3.5 for a discussion). In the early years the Governing Council of the ECB discussed monetary policy at every meeting, which takes place every fortnight. However, in November 2001 it was decided that the Council should, as a rule, assess the monetary policy stance only at its first meeting of the month. At these first meetings interest rate decisions are normally taken as well. The ECB has a range of instruments at its disposal for implementing monetary policy. To manage liquidity in the money market and steer short-term interest rates, it uses open market operations. In addition two standing facilities allow eligible counterparties to invest their daily liquidity surpluses or to cover their overnight liquidity needs (see figure 2.2).

The ECB: Structure, Strategy, and Policy

15

The two standing facilities are the marginal lending facility and the deposit facility. Both facilities have an overnight maturity and are available to banks on their own initiative. The deposit facility is used for mopping up liquidity from the banks at rates that are below market rates. The rate set at this facility acts as a floor (lower limit) for shortterm money market rates. The marginal lending (or Lombard) facility provides liquidity to the banks at rates usually above market rates. The rate set at this facility thus acts as a ceiling (upper limit) for short-term money market rates. The standing facilities thus constitute a corridor for the (interbank) money market rate and signal the ECB’s views on interest rates in the medium run. Within that corridor the money market rate is fine-tuned by means of open market operations. In assessing the situation at the money market, EONIA plays an important role. EONIA is short for Euro Over-Night Index Average. It is a two decimal place rate that has been calculated and published by the ECB at the end of every TARGET business day since January 4, 1999. The EONIA interest rate is the weighted average rate on transactions reported to the ECB every business day by a representative panel of banks, known as the EONIA Panel. EONIA is therefore representative of the vast majority of euro-denominated overnight lending. The importance of EONIA to the ECB is that it contains information on the supply and demand equilibrium in the overnight market. The total cash market volume reported for EONIA purposes is published daily. Although it varies considerably, depending on market conditions, the cash market averages around €40 billion. There are four categories of open market operations. First, the main refinancing operations (MROs) are the most important open market operations and represent the key monetary policy instrument of the ECB (ECB 2004). They are executed as standard weekly tenders for liquidity-providing transactions initially with a two-week maturity.5 In January 2003 the Governing Council of the ECB decided to shorten the maturity of the MROs from two weeks to one week, effective as of March 2004. Normally these operations consist of reverse transactions, meaning the central bank buys assets under a repurchase agreement or grants a loan against assets given as collateral. So these operations provide funds for a limited and pre-specified period only. Second, the longer term refinancing operations, which are executed as standard monthly tenders for liquidity-providing reverse transactions with a three-month maturity, have the purpose of fulfilling the additional longer term refinancing needs of the financial system. Third,

16

Chapter 2

fine-tuning operations will be executed as needed to smooth the effect on interest rates of unexpected shocks in money market liquidity. Finally, the Eurosystem may issue debt certificates and carry out reverse or outright transactions whenever it wishes to adjust the structural position of the Eurosystem vis-à-vis the financial sector. The ECB also introduced a minimum reserve system, by which credit institutions are required to hold compulsory deposits with national central banks. The size of the required reserves is determined by the size and composition of the liabilities on the balance sheet of the institution concerned. For most liabilities included in the reserve base the reserve ratio is 2 percent. The minimum reserve system serves two main purposes: to contribute to the stabilization of money market interest rates and to create sufficient structural demand for central bank money. Interest rates are stabilized by allowing credit institutions to use averaging provisions: reserve requirements are determined on the basis of average daily reserve holdings over the maintenance period. Since March 2004 the timing of the reserve maintenance period is such that it will always start on the settlement day of the main refinancing operation following the Governing Council meeting at which the monthly assessment of the monetary policy stance is pre-scheduled. Furthermore, as a rule, the implementation of changes to the standing facility rates are to be aligned with the start of the new reserve maintenance period. 2.4 Monetary Policy of the ECB Table 2.1 shows the decisions of the ECB on interest rates. Just before the ECB became responsible for monetary policy in the euro area, interest rates were lowered in a coordinated decision by the various national central banks in December 1998. In the first months of 1999 interest rates remained stable. Many observers were worried at the time whether the new institution would be able to withstand political pressures. The start of the ECB had not been very promising after the political tinkering about the presidency of the ECB (see Eijffinger and De Haan 2000). Furthermore, right after the ECB took over the monetary policy responsibilities from the Bundesbank, there was strong pressure—notably from the German Finance Minister Oskar Lafontaine—to cut interest rates. In April 1999 the ECB reduced interest rates by 50 basis points. At the press conference after the decision was taken, ECB President Duisenberg reported

The ECB: Structure, Strategy, and Policy

17

Table 2.1 ECB interest rates, January 1999–December 2003 Main refinancing operations

With effect from:

Deposit facility

Fixed rate tenders, fixed rate

Variable rate tenders, minimum bid rateb

Marginal lending facility

January 1, 1999a April 9, 1999 November 5, 1999 February 4, 2000 March 17, 2000 April 28, 2000 June 9, 2000 June 28, 2000 September 1, 2000 October 6, 2000 May 11, 2001 August 31, 2001 September 17, 2001 November 9, 2001 December 5, 2002 March 7, 2003 June 6, 2003

2.00 1.50 2.00 2.25 2.50 2.75 3.25 3.25 3.50 3.75 3.50 3.25 2.75 2.25 1.75 1.50 1.00

3.00 2.50 3.00 3.25 3.50 3.75 4.25 — — — — — — — — — —

— — — — — — — 4.25 4.50 4.75 4.50 4.25 3.75 3.25 2.75 2.50 2.00

4.50 3.50 4.00 4.25 4.50 4.75 5.25 5.25 5.50 5.75 5.50 5.25 4.75 4.25 3.75 3.50 3.00

Source: www.ecb.int. a. On December 22, 1998, the ECB announced that, as an exceptional measure between January 4 and 21, 1999, a narrow corridor of 50 basis points would be applied between the interest rates for the marginal lending facility and the deposit facility in order to facilitate the transition to the new regime by market participants. b. On June 8, 2000, the ECB announced that, starting from the operation to be settled on June 28, 2000, the main refinancing operations would be conducted as variable rate tenders.

on the reasons for the interest rate cut. He explained that the 12-month growth rate of M3 growth stood at 5.2 percent in February—namely above the reference value but at clearly a declined rate—while the inflation rate was somewhat below 1 percent. According to the ECB’s assessment it appeared unlikely that HICP increases would be out of line with the Eurosystem’s definition of price stability, while at the same time economic growth prospects worsened toward the end of the year. In November 1999 interest rates were raised 50 basis points. According to Duisenberg, the main argument for reversing what was then

18

Chapter 2

called the precautionary interest rate reduction was the fact that around the beginning of the summer the balance of risks to future price stability gradually moved toward the upside. Inflation rates were expected to increase mainly as the result of the increase in energy prices earlier that year, which was working its way through to consumer prices. An increase in interest rates should help to counteract further liquidity growth over the medium term—in the period July to September M3 growth stood at 5.9 percent—and contribute to maintaining inflation expectations safely below 2 percent. In 2000 interest rates were further increased in six subsequent steps. Except for the increase of June 2000, interest rates rose by 25 basis points. During this period there was sometimes confusion as to the role of the euro–dollar exchange rate. Some observers thought that the ECB had changed its policies and was now trying to stabilize the exchange rate. However, the ECB does not have an exchange rate target. Still the exchange rate is one of the indicators used in the second pillar and is therefore closely observed by the ECB. There was also a feeling that the euro was undervalued. Occasionally the ECB intervened, sometimes jointly with the central banks of the United States and Japan. According to the ECB press release of September 22, 2000, this concerted intervention was motivated by the “shared concern about the potential implications of recent movements in the euro exchange rate for the world economy.” In November 2001 the ECB intervened again, but now the central banks of the United States and Japan did not join in. In section 3.6 we will deal in more detail with the exchange rate policy of the ECB. At its meeting on May 10, 2001, the Governing Council of the ECB decided to reduce interest rates. This decision was regarded as an adjustment to somewhat lower inflationary pressures over the medium term. Furthermore M3 growth had been on a downward trend, standing at 4.8 percent over January to March 2001. The ECB also stated that the monetary growth figures were distorted upward by non–euro area residents’ purchases of negotiable paper included in M3. This distortion was estimated to be around half a percentage point of the annual growth rate of M3. There was a further distortion due to non–euro area residents’ holdings of other marketable paper included in M3. The size of this distortion was uncertain when the Governing Council took its decision. At the end of August a further rate cut of 25 basis points took place. In September the Fed and the ECB reduced their rates by 50 basis

The ECB: Structure, Strategy, and Policy

19

points to counteract the feared negative impact of the terrorist attacks on the United States, followed by a further cut of 50 basis points in November. At the beginning of 2002 euro banknotes and coins were introduced. The euro conversion was successfully carried out (see Sidebar 2). After a long period of unchanged interest rates, the Governing Council decided at its meeting of December 5, 2002, to lower the key ECB interest rates by 50 basis points, because “the arguments in favour of a cut in the key ECB interest rates have strengthened. The evidence that inflationary pressures are easing has increased, owing in particular to the sluggish economic expansion. Furthermore, downside risks to economic growth have not vanished.”6 In 2003 rates were cut twice. In March and June the Governing Council decided to reduce the interest rates by 25 and 50 basis points, respectively. The Council considered that the outlook for price stability over the medium term had improved, owing in particular to the subdued pace of economic growth and the appreciation of the exchange rate of the euro. Although the growth rate of M3 was far above the reference value—in fact, it increased to 7.1 percent in the period from November 2002 to January 2003, compared with 6.9 percent in the period from October to December 2002—it was not expected to be inflationary. As Duisenberg put it in his press conference on March 7, 2003: The continued strong monetary growth reflects an ongoing pronounced preference for liquidity in an environment of high financial, economic and geopolitical uncertainty. Although liquidity remains ample, it is not expected at this stage to give rise to inflationary pressures, given the current economic context and the expectation that some of the portfolio shifts will be reversed once the financial market uncertainty diminishes. Recent data on loans to the private sector, notably the weak growth in loans to nonfinancial corporations in late 2002, confirm this assessment.

Similar arguments were used to explain the interest rate reduction in June 2003. At the time, inflation rates were expected to decline to below 2 percent over the medium term, following recent movements in the exchange rate of the euro and given the sluggish growth performance of the euro area. According to the ECB, the strong expansion of M3 should not for the time being adversely affect this outlook. No specific reasons were given why, in contrast to the previous step, rates were now cut by 50 basis points.

20

Chapter 2

1 Too Little, Too Late? According to some observers, the ECB reduced interest rates too little and too late in 2001. In November 2001, Sinn argued, for instance: “The European Central Bank’s failure to lower interest rates at its last meeting, was very disappointing. Is the boom in Ireland and Finland so important for Europe that it prevents battling recession in the heart of the Continent? (Sinn 2001, p. 2). Sinn suggests that the decentralized decision-making structure of the ECB has led to suboptimal decisions. The table below shows the capacity utilization rates in the euro area countries around the time of the ECB decisions to which Sinn refers, namely the last two quarters of 2000 and the first two quarters of 2001. Nevertheless, in terms of the capacity utilization rate of manufacturing, Germany was still doing better at the time than the euro area as a whole, even better than Ireland.

Table 2.2 Capacity utilization rates (%) in the euro area, 2000–2001 Capacity utilization Country

2000:III

2000:IV

2001:I

2001:II

Euro area Austria Belgium Finland France Germany Greece Ireland Italy Netherlands Portugal Spain

83.9 85 84.2 87 87.9 86.0 77.8 78.5 78.9 85 82 80.3

84.7 85 84.5 86.7 89.1 86.3 78.4 81.5 79.8 84.6 80.9 80.8

84.4 84.2 84.8 87.3 88.8 86.9 78.2 79.8 79.5 85.2 82.5 80.1

83.6 84 82.7 86 86.9 85.7 79.3 80.4 79.4 84.8 82 79.7

Source: Datastream.

The ECB: Structure, Strategy, and Policy

21

2 The Changeover to the Euro The changeover to euro coins and banknotes on January 1, 2002, marked a big leap in the process of monetary integration in Europe. A remarkable feature of the changeover to euro currency was the different periods of dual circulation of national and euro currency in many countries. While Germany had formally no dual circulation period, Austria, Belgium, Finland, Greece, Italy, Luxembourg, Portugal, and Spain opted for the maximum of two months as allowed by the Maastricht Treaty. Between the extremes were the Netherlands (4 weeks), Ireland (6 weeks), and France (7 weeks). Although the experiences with the changeover differed widely, the changeover was successful. The conversion of automatic teller machines (ATMs) had been completed on the first working day (January 2) in Austria, Germany, Luxembourg, and the Netherlands. By January 4 nearly all countries, except Italy and Spain, had effected the ATM conversion. Saturday, January 5, was an important test for the success of the changeover as the volume of business was higher than on all other days of the first week counted together. At the end of this day, the European Commission concluded that “ . . . it would appear that, overall, queues in shops, and in large stores too, were still no longer than usual” (European Commission, Press Release, January 7, 2002). According to the Commission, on average, just under two-thirds of cash payments were made in euros. On January 11 the Commission concluded that the euro changeover was practically complete. On average, almost 85 percent of the number of cash payments were then made in euros, and over two-thirds of the vending machines had been converted to the euro (European Commission, Press Release, January 11, 2002). Source: Eijffinger (2002)

2.5 ECB’s Evaluation of Its Monetary Policy Strategy On May 8, 2003, the Governing Council of the ECB made the results of the evaluation of its monetary policy strategy public. The opinion of the Council was that so far the strategy had worked satisfactorily. According to Issing (2003, pp. 4–5): Since 1999, medium and long-term inflation expectations—as measured by survey data or financial market indicators—have remained consistent with the ECB’s definition of price stability. This is all the more remarkable given that the ECB started without a track record of its own and that it has experienced a number of sizeable adverse price shocks. As a result of these shocks, HICP inflation was above (and sometimes significantly above) 2% for quite some time. But, as the shocks gradually unwound, so inflation has returned towards

22

Chapter 2

Table 2.3 Rate cuts and consensus inflation forecasts Date

Rate cut (%)

Year-ahead consensus forecast

April 1999 May 2001 August 2001 September 2001 November 2001 December 2002 March 2003

0.50 0.25 0.25 0.50 0.50 0.50 0.25

1.6 1.8 1.8 1.8 1.7 1.8 1.7

Source: Kieler (2003).

levels compatible with price stability. Medium- and long-term inflation expectations remained well anchored throughout this period.

As table 2.3 shows, the year-ahead consensus forecast for inflation on the dates when the ECB cut interest rates varied in a narrow range from 1.6 to 1.8 percent. Although actual inflation exceeded 2 percent almost continuously since mid-2000, this did not lead to an increase in expected inflation. The ECB’s Survey of Professional Forecasts had shown the expectation over the five-year horizon to be 1.8 to 1.9 percent since the inception of the monetary union (Kieler 2003). The Governing Council made two clarifications in view of the evaluation. First, the Council refined its definition of price stability. The Council agreed to maintain inflation rates below but close to 2 percent over the medium term. According to the ECB press release, “This clarification underlines the ECB’s commitment to provide a sufficient safety margin to guard against the risks of deflation. It also addresses the issue of the possible presence of a measurement bias in the HICP and the implications of inflation differentials within the euro area.” As was pointed out by Issing (2003), the main reason for keeping inflation below but close to 2 percent was the need for a safety margin against potential risks of deflation. In instances of strong deflationary pressures, monetary policy could become ineffective if the central bank interest rate management is constrained by a liquidity trap or a “zerobound” problem. The analyses of the ECB showed that such constraints would not pose a significant threat if inflation remains sufficiently above zero. The aim to maintain inflation rates close to 2 percent further took into account a potential measurement bias in the HICP and the

The ECB: Structure, Strategy, and Policy

23

inflation differentials of a structural nature within the euro area. In chapter 3 we discuss these issues in some detail. Second, the Council refined the structure of its policy making, though not departing from the objective to base policy decisions on a comprehensive analysis of the risks to price stability. The Council noted that their perception of both pillars of the monetary policy strategy over time had changed and been enlarged. At the core was the aim to provide clearer communication and means of verifying information in coming to a unified decision on the risks to price stability. A new structure was given to the introductory statement of the president. It will start with the economic analysis to identify short- to medium-term risks to price stability, followed by the monetary analysis “to assess medium- to long-term trends in inflation in view of the close relationship between money and prices over extended horizons.” As in the past, monetary analysis was to take into account developments in a wide range of monetary indicators, including M3, its components and counterparts, notably credit, and various measures of excess liquidity. According to the Governing Council, the monetary analysis was mainly seen as a means of cross-checking, from a medium- to longterm perspective, the short- to medium-term indicators of economic performance. To stay focused on the longer term nature of the reference value for M3 growth, the Governing Council decided to discontinue the practice of an annual review of this reference value. The ECB (2003a, p. 24) explains this decision as follows: . . . the practice followed thus far by the Governing Council of annually reviewing the reference value may have contributed to the occasional misperception that the reference value is announced “for a specific year,” similar to the past practice of some monetary targeting central banks. Since the reference value is based on medium-term assumptions for real GDP growth and the income velocity of money—and since experience has shown that these assumptions change only gradually—there may in fact be no need for an annual review of the technical assumptions underlying the computation of the reference value.

Although the statements of the ECB were intended merely to be a clarification of its concern with price stability, many observers interpreted them to be a change in the ECB’s inflation objective. For instance, Sibert (2003) called it a “distinctly dovish change in policy.” When it was originally announced that inflation was to be kept at lower than 2 percent, the natural interpretation was not that the ECB was aiming to keep inflation just below 2 percent. Since the floor was zero,

24

Chapter 2

the obvious conclusion was that the ECB was aiming for around 1 percent, with the costs of deviating from this rising sharply as inflation either rose to two percent or fell to zero percent. Likewise the discussion of the two-pillar strategy was widely interpreted as diminishing the importance of the first pillar in the monetary strategy despite ECB officials’ stress on its continuity. For instance, Svensson (2003) summarized the Governing Council decision as follows: “Keeping the two-pillar strategy but reducing the prominence of the first pillar by putting it second and discussing the monetary pillar (relabeled ‘monetary analysis’) after the ‘broadly-based assessment’ (relabeled ‘economic analysis’), seeing it mainly as a means of ‘cross-checking’ the ‘economic analysis.’ This is a change in the right direction, but it is not enough.”7 As is evident from the preceding quotes, the ECB is being closely watched by many academic and professional economists. Also the European Parliament reviews its monetary policy regularly (see sidebar 3 for further details). In chapter 3, we offer our synthesis of the debate on the monetary policy strategy of the ECB.

3 ECB Watchers: Bachelors’ Wives and Maidens’ Children Are Well Trained Like many other central banks, the ECB is closely monitored by academic and professional watchers, who regularly publish their assessment of the ECB and their comments on its monetary policy. A prominent group of academic ECB watchers are the members of the Panel of Economic and Monetary Experts who have been submiting their briefing papers quarterly since June 2000 to the Committee on Economic and Monetary Affairs of the European Parliament.8 In accordance with Article 113(3) of the Treaty on European Union, this committee is responsible for maintaining a monetary dialogue with the president of the ECB. These academics have influenced the ECB’s policies to some extent. For example, the publication of inflation projections by the ECB for the first time in December 2001 was in response to a call in the monetary dialogue with the ECB president during 2000. It is remarkable that professional ECB watchers—above all, financial market experts—change their points of view very quickly. Against the background of a gloomy economic outlook, a stronger euro, easing inflationary pressure and turbulence in financial markets during the summer of 2002 came, in the course of the year, increasingly louder calls for more

The ECB: Structure, Strategy, and Policy

25

interest rate reductions. The same analysts had argued for an interest rate hike only a few weeks before but were now demanding—with the same amount of pressure—an easing of monetary policy. The economists of JP Morgan Chase, for example, expected the ECB to lower interest rates to 2 percent in the aftermath of the terrorist attacks of September 11, 2001. By the end of January 2002 they claimed to see the turning point in the interest rate cycle at 2.75 percent. With the first signs of economic recovery in the first quarter, the experts changed their minds and predicted an interest rate increase. It is doubtless vital for the accountability, transparency, and communication strategy of the ECB to engage an active debate over monetary policy issues and to exchange views with independent academic and nonacademic ECB watchers. Sometimes the critique, both from academic and professional ECB watchers, is not particularly practicable, and sometimes the critics characterize the ECB’s polices inappropriately. For instance, Begg et al. (2000) refer to the first pillar as a “money growth reference target.” Nevertheless, the communication strategy of the ECB can be criticized (see chapter 4 for a discussion). The ECB communications on the evaluation of the two pillars in its monetary policy strategy were misleading. At the press conference when the results were made public, the ECB officials claimed that the evaluation was merely a clarification. However, the statements on the price stability objective and the presentation of the outcome of the policy discussion were such that they were widely interpreted as indicating that the ECB had changed its monetary policy strategy. So there is some question whether words and deeds were in agreement, especially since some central bankers privately communicated that the first pillar had been pushed aside.

2.6 Conclusions Before the ECB started operations, the main elements of its monetary policy were decided on by the Governing Council in October 1998. The original strategy consisted of three elements. First was the Council’s formal definition of price stability, namely to abide by a yearly inflation for the euro area of not more than 2 percent in the medium term. Second was the prominent role given to money growth in the assessment of the risks to price stability. This became known as the first pillar of the ECB’s monetary policy. Third was to use a wide variety of economic and financial variables in providing a broadly based evaluation of any threats to price stability. This was the second pillar.

26

Chapter 2

In May 2003 the ECB announced the outcomes of its internal evaluation of its strategy. Although the ECB stressed continuity and stated that the decisions taken were all about clarification, these decisions were perceived widely as a change in its monetary policy strategy. The ECB in fact did still seek to maintain the inflation rate below or close to 2 percent over the medium term. The use of the two-pillar framework of the strategy was confirmed to be a tool to organize the information relevant for assessing the different risks to price stability. It was also announced that the introductory statement of the president at the ECB Press Conference following a Governing Council rate-setting meeting was to start with an analysis identifying the short- to mediumterm risks to price stability. This part of the assessment was to be called the economic analysis. After the assessment of the medium- to longterm risks to price stability, the monetary indicators were to be examined. This second part of the assessment was to be called the monetary analysis. The Governing Council emphasised that the monetary analysis would serve mainly as a means of cross-checking, from a medium- to long-term perspective, the short- to medium-term indications of the economic analysis. This decision has been widely seen as downgrading the importance of the (previous) first pillar.

3

The ECB’s Strategy: An Assessment

3.1 Introduction In the academic literature the monetary policy strategy of the ECB has been severely criticized. For instance, Begg et al. (2000, p. 6) argue: At a time when more and more central banks replace money-growth targeting with (expected) inflation targeting, and make efforts at being transparent, the ECB’s strategy is unavoidably seen as obscure, often even archaic.

This chapter provides an assessment of the ECB’s monetary policy strategy to examine whether this critique is justified. We start with an analysis of the predictability of interest rate decisions by the ECB. After a review of the literature, we present some new results based on analyses of forward rates and newspaper reports (section 3.2). The next two sections deal with the elements of the ECB’s strategy that have been most severely criticized, namely the definition of price stability (section 3.3) and the role of the monetary analysis (section 3.4). Subsequent sections discuss inflation forecasts (section 3.5), the role of the external value of the euro in the monetary policy strategy of the ECB (section 3.6), the questions of whether asset prices should play a role in the ECB policies (section 3.7), and whether interest steps should be small or large (section 3.8). The final section offers our conclusions. 3.2 Predictability of Policy Decisions As follows from the citation of Begg et al. (2000), some observers argue that the ECB sometimes behaved in a rather erratic and nontransparent way.1 Issing (2001a, p. 15) argues that one possible way to approach the transparency of monetary policy decisions is to analyze the predictability of the decisions taken by the Governing Council. “If the

28

Chapter 3

stability-oriented strategy were indeed to complex and ultimately not transparent, as claimed by some of our critics, then we should expect monetary policy decisions to be very imperfectly forecasted.”2 In order to examine this issue more systematically, we analyze the development of forward market rates on a daily basis, as well as newspaper reports. 3.2.1 Analysis of Forward Rates Financial market prices may contain information about inflation and interest rate expectations. Therefore they are often used for measuring market expectations. Söderlind and Svensson (1997) provide a survey of methods to extract information about market expectations from asset prices for monetary policy purposes. Traditionally interest rates and forward exchange rates have been used to extract expected interest rates, exchange rates, and inflation. More advanced approaches rely on implied forward interest rates (market consensus of future interest rates based on the relationship between spot rates) to extract expected future time-paths. Very recently, methods have been refined to extract not only the means but also the whole (risk-neutral) probability distribution from a set of option prices. Svensson (1993), Söderlind (1995), and Söderlind and Svensson (1997) use forward interest rates (nominal interest rates agreed upon today for an investment period starting in the future) as an indicator of the monetary stance of a central bank. They suggest the use of forward rates, because these reliably indicate market expectations of future short-term interest rates, inflation, and risk premiums in general. Forward rates presumably consist of four unobservable components: expected real interest rates, expected future inflation, and the relevant risk premiums (i.e., the forward-term risk premium and the inflation risk premium). Isolating any of these components, however, is a difficult exercise as the four components are interrelated. In recent years many papers focusing on the predictability of monetary policy in the United States have been published. The question of whether market participants were able to anticipate the Fed’s interest rate decisions has, for instance, been analyzed using Fed funds futures prices (e.g., see Krueger and Kuttner 1996; Kuttner 2001; Poole and Rasche 2000; Owens and Webb 2001). Following the approach by Poole and Rasche (2000), Gaspar et al. (2001) have analyzed whether the announcements of ECB decisions have a significant impact on the behavior of overnight rates (EONIA) in Europe. The bottom line of this research is that if the market antici-

The ECB’s Strategy: An Assessment

29

pates the systematic behavior of the central bank, it should only adjust to news (i.e., information innovations) but not to the central bank’s actual announcements of monetary policy decisions. The authors conclude that the ECB monetary policy announcements do not significantly affect the mean interest rates. These findings are consistent with the assumption that markets do not make systematic mistakes in anticipating the ECB’s monetary stance. The results of Gaspar et al. suggest that markets were able to predict the ECB’s interest rate decisions quite accurately over the period of analysis (January 1, 1999, to March 23, 2001). Table 3.1 is reproduced from this study. It shows the success of the market in predicting the ECB’s interest rate movements on the day before the meeting of the Governing Council, distinguishing between the times when the ECB moved interest rates and when it left rates unchanged. Of the ECB interest rate changes taken into account, only one, namely the April 1999 decision, caught the markets by surprise. Likewise in only one case the markets expected a move, while the ECB rates remained stable. Somewhat less supportive evidence for the predictability of ECB moves is reported by Ross (2002), who finds that while ECB rate hikes were generally well predicted, the markets have had more difficulty in fully and correctly anticipating ECB rate cuts in terms of either timing or magnitude. In the remainder of this section, we examine whether in the forward rates financial markets anticipated the interest rate steps of ECB. We calculate the differences between the actual money market rate and the three-month EURIBOR future. This difference yields information about the market participants’ interest rate expectations. We compare these expectations with the ECB’s actual interest rate moves. Figure 3.1 shows that the markets anticipated most of the ECB’s interest rate moves. Sometimes the move was expected to take place earlier, and sometimes the size of the step was over- or underestimated. Table 3.1 Have money markets anticipated ECB’s decisions? Market expectations

ECB

Move No move

Source: Gaspar et al. (2001).

Move

No move

Number of observations

88% 8%

12% 92%

8 36

30

Chapter 3

0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 12-3-98 4-3-99 8-3-99 12-3-99 4-3-00 8-3-00 12-3-00 4-3-01 8-3-01 12-3-01 ECB interest rate move

Interest expectations extracted from 3M futures

Figure 3.1 Interest rate expectations extracted from three-month futures

Let us turn to examine some of the ECB decisions (see table 2.1 for an overview of the decisions). A particularly controversial decision was the ECB’s first interest rate decrease in April 1999. The crises in Russia and Brazil severely affected confidence in the world economy and low oil prices created fears of deflation. The ECB was already under much pressure, but it was only April 8, 1999, when the central bank lowered its key rate by 50 basis points. For the market participants the decision came too late; they had in fact expected the rates to remain and their fears of deflation had dissipated. The decision of November 1999 was expected earlier. About one month before the ECB’s decision to raise the rate by 50 basis points, the markets had priced-in a tightening of monetary policy, in anticipation of a big step of 75 basis points. For some critics the increase was simply a correction of the earlier “unnecessary” decrease of April 1999; for others, it was the ECB’s reaction to the rising inflation. In either case, however, the move was not unexpected since it was announced by the ECB well in advance. The ECB decision of February 2000 caught financial markets by surprise. Some financial market participants were about to believe that the

The ECB’s Strategy: An Assessment

31

0.1

0

-0.1

-0.2

-0.3

-0.4

-0.5 1-1-01 1-15-01 1-29-01 2-12-01 2-26-01 3-12-01 3-26-01 4-9-01 4-23-01 5-7-01 5-21-01

Actual ECB move

Interest rate expectations extracted from 3M futures

Figure 3.2 Interest rate expectations in May 2001

ECB was acting sluggishly and only able to make big moves, and they were puzzled when the ECB raised rates by only 25 basis points. This was the beginning of a series of “trip steps.” The ECB decisions of March, April, June, and August were generally expected. Likewise the step of October 2000 had been expected by the markets, but compared to the earlier moves it had been anticipated very late. During the period from the end of August 2000 and the last days of September, the markets did not expect an interest rate move. Only a few days prior to the ECB meeting, forward rates disclosed an increase in interest rate expectations. The May 2001 decision was also a surprise to many financial market participants (see figure 3.2). Whereas the majority of the market participants had bet on a small step downward by the beginning of January and even on a 50 basis points step by mid-April, the ECB actually did not lower its key rate until May 10. Meanwhile the markets had accommodated to the belief that an interest rate decrease would not fit into the ECB’s monetary strategy at that time, with M3 remaining persistently above its reference value and HICP inflation above 2 percent. As explained in chapter 2, the ECB justified the move by the latest monetary developments and the temporary upward distortion of

32

Chapter 3

the monetary aggregate due to non–euro area residents’ purchases of negotiable paper included in M3. The decision of August 2001 did not come as a surprise. The next decision to reduce rates after the terrorist attack of September 11, was decided upon by means of teleconferencing. This was the first decision taken between two regular Council meetings, and it was widely accepted, though it had not been expected. At the regular meeting on September 13, the Council decided to leave rates unchanged. This decision was justified by the imminent political and economic uncertainty after the terrorist attacks and the announcement that the ECB would coordinate its activities with the Federal Reserve and other major central banks. These arguments were in line with markets’ perceptions. But what was not easy to understand was that the ECB finally lowered rates by 50 basis points with some lag behind the Fed. This gave way to speculations that the ECB had taken an ad hoc decision. The speculation was nourished by the subsequent change in the minimum bid rate for the next refinancing operation, which was settled on September 19, 2001. The step of November 2001 did not come as a surprise, although the markets had only anticipated a small step. All in all, we can conclude that the evidence is mixed. By and large the markets predicted the ECB moves quite accurately. Only in a few cases did the ECB take the market by surprise, while in other cases the timing of the ECB steps was not entirely in line with market expectations. 3.2.2 Expectations according to Newspaper Reports In a further attempt to analyze to what extent financial markets were caught by surprise, we systematically gathered the newspaper reports about ECB policy decisions in the Financial Times (FT) and the Frankfurter Allgemeine Zeitung (FAZ) in 1999 and 2000. Generally, journalists are in close contact with financial market analysts and are therefore able to examine whether a certain policy decision was in line with financial market expectations. Table 3.2 reports the outcomes from this analysis. After each meeting of the ECB’s Governing Council either a press conference took place and/or a press release was issued. The main decision is shown in the second column of table 3.2, while the third and fourth columns show whether according to the FT and the FAZ this decision was in line with financial market expectations. We only report those decisions on which the journalists had something to say about financial markets’ expectations.

The ECB’s Strategy: An Assessment

33

Table 3.2 Monetary policy decisions and financial market expectations, 1999–2000 Council meeting

Decision

FT report

December 22, 1998

Rate cut

Many analysts have been expecting an ECB rate cut in the first half of next year (FT, 12–23–98)

January 7, 1999

No change

January 21, 1999

No change

February 18, 1999

No change

April 8, 1999

Decrease 50 basis points

May 6, 1999

No change

July 1, 1999

No change

July 15, 1999

No change

July 29, 1999

No change

August 26, 1999

No change

September 9, 1999

No change

September 23, 1999

No change

FAZ report

Majority of observers expect constant rates for next three to four months (FAZ, 1–8–99) ECB will lower rates by 0.5 percentage points according to Salomon Smith Barney (FAZ, 1–22–99) In line with expectations (FT, 2–19–99) Cut bigger than expected (FT, 4–9–99) Most economists expect euro-zone rates to remain unchanged until at least September (FT, 5–7–99) In line with expectations (FT, 7–2–99) Most economists expect the 2.5% rate to stay until next year (FT, 7–16–99) In line with expectations (FT, 7–20–99)

Markets expect tightening before end of year (FT, 9–10–99) Most economists expect no rate rise before first quarter of 2000 (FT, 9–24–99)

Traders say that size of the cut surprised them (FAZ, 4–10–1999) No one had taken into account that ECB would change interest rates again (FAZ, 5–7–99)

Markets had expected no change (FAZ, 7–16–99)

Decision was expected (FAZ, 8–27–99) Decision was expected (FAZ, 9–10–99)

34

Chapter 3

Table 3.2 (continued) Council meeting

Decision

FT report

FAZ report

November 4, 1999

Increase 50 basis points

Markets had anticipated the move, economists welcome the size of the increase, some financial market participants had expected rise of 0.25 (FT, 11–5-99)

December 15, 1999

No change

January 20, 2000

No change

ECB did not disappoint financial market participants, decision was for majority of analysts ausgemachte Sache, next increase not expected before mid 2nd quarter of 2000 (FAZ, 11–5–99) Bundesverband deutscher Banken praises ECB for information policy of monetary policy stance (FAZ, 12–16–99) Rate rise of 25 basis points expected in Spring (FAZ, 1–21–00)

February 3, 2000

Increase 25 basis points

February 17, 2000 March 2, 2000

No change

March 16, 2000

No change

Increase 25 basis points

Economists divided over size of next move, some predicting a 50 basis point rise in March (FT, 1–21–00) Economist are divided how sharply ECB may rise rates (FT, 1–21–00); many analysts had been expecting either no change or a 50 basis point move (FT, 2–4–00)

Decision won broad approval, rise would have created confusion (FT, 3–3–00); most market participants had expected rates to remain unchanged (FT, 3–2–00) Rise came as little surprise to markets which had expected an increase (FT, 3–17–00)

Many market observers had expected this decision after hints of Duisenberg; decision was widely expected (FAZ, 3–17–00)

The ECB’s Strategy: An Assessment

35

Table 3.2 (continued) Council meeting

Decision

FT report

April 13, 2000

No change

April 27, 2000

Increase 25 basis points

May 11, 2000

No change

May 25, 2000

No change

June 8, 2000

Increase 50 basis points

June 21, 2000

No change

July 20, 2000

No change

August 31, 2000

Increase 25 basis points

October 5, 2000

Increase 25 basis points

As expected; many economists expect ECB to raise rate to 4% before summer recess (FT, 4–14–00) Increase was in line with expectations in financial markets (FT, 4–27–00) Decision had been widely expected (FT, 12–5–00) Decision strengthened expectations that next rise will come in June (Ft, 5–26–00); most economists expected the ECB to leave rates on hold (FT, 5–26–00) ECB surprised financial markets; most analysts had expected rise by 0.25 percentage points (FT, 6–9–00) Few people were surprised when ECB elected to keep rate unchanged (FT, 6–26–00) Decision confirmed expectations that next move is not likely before September (FT, 7–21–00) Money market and some economists had been looking for a bolder move (FT, 9–1–00) Biggest surprise in the bond market was ECB’s decision to raise rates; move caught many in the financial markets by surprise (FT, 10–6–00)

FAZ report

Exchanges were not surprised (FAZ, 4–28–00)

Traders surprised, rise of only 25 points had been expected (FAZ, 6–9–00)

Decision was expected (FAZ, 7–21–00)

Decision was widely expected (FAZ, 9–1–00)

ECB increases rates surprisingly (FAZ, 10–6–00)

36

Chapter 3

Table 3.2 (continued) Council meeting

Decision

FT report

October 19, 2000

No change

November 2, 2000

No change

November 16, 2000

No change

November 30, 2000

No change

Move welcomed by economists (FT, 10–20–00) A repetition of the 0.25 percentage point rise had been deemed highly unlikely (FT, 10–19–00) Financial markets had expected decision (FT, 11–3–00) Decision was in line with market expectations (FT, 11–17–00) Step widely anticipated in financial markets (FT, 12–1–00)

FAZ report

According to this analysis the policy of the ECB during 1999 was in line with financial market participants’ expectations. As the Financial Times (December 3, 1999) reported: . . . the ECB is starting to earn a reputation as a central bank that does not want to “fine-tune” the euro-zone economy by making frequent, often unexpected, interest rate changes. If predictability is the key to a central bank’s credibility, as Mr. Duisenberg recently asserted, the ECB is moving in the right direction.

In 2000 the situation changed. In contrast to our analysis of forward rates, the decision of October 2000 caught many financial markets by surprise according to the newspaper reports summarized in table 3.2. Still, this analysis supports the conclusion that, by and large, the ECB policy moves were in line with financial markets’ expectations. 3.3 Inflation Objective The initial ECB definition of price stability has been severely criticized in the literature as being asymmetric and ambiguous. It has been suggested that the ECB should make its objective more precise, for instance, by announcing a lower bound in the definition or by specifying the objective in terms of a point inflation rate. The discussion on

The ECB’s Strategy: An Assessment

37

the definition of price stability is summarized in section 3.3.1. Section 3.3.2 deals with the debate on the question of whether an inflation objective of “below but close to 2 percent” is too low. 3.3.1 Definition The price index used to calculate inflation should be credible. A price measure constructed by the central bank itself may therefore lack credibility, as the central bank may be tempted to manipulate that measure to achieve its target. As we pointed out before, the ECB defines inflation on the basis of the HICP, which is calculated by Eurostat. The HICP is a Laspeyres-type index of consumer spending, similar to the US consumer price index (CPI) and the UK retail prices index, excluding mortgage interest payments (RPIX). The HICP index is published monthly and in a timely manner. Furthermore it is very rarely revised, and it can be aggregated across countries (Camba-Mendez 2003). Some critics of the ECB have argued that a more transparent definition would formulate price stability in terms of core inflation (e.g., see Gros et al. 2000). Although the ECB also publishes HICP inflation excluding certain items like unprocessed food stuff and energy cost (see figure 3.3), the ECB rejects a definition of price stability based on core inflation as it, no doubt, would invite considerable criticism because of the arbitrary nature of deriving such a measure (ECB 2003a). We therefore concur that HICP inflation is an appropriate measure to be used for the definition of price stability.3 In providing a definition of price stability, the ECB has specified its medium-term inflation objective in more precise terms than some other central banks, such as the US Federal Reserve or the Bank of Japan, which do not offer quantitative definitions of their targets. However, in comparison to the objectives of central banks with an inflationtargeting strategy, the ECB definition of price stability is less precise. Some inflation targeting central banks have adopted point targets (Canada, Sweden, and the United Kingdom) while others have adopted target ranges (Australia and New Zealand).4 For those banks that have adopted ranges, the lower bound is explicit, and the midpoint of the range is generally understood to represent the preferred inflation outcome. The policy horizon for most inflation-targeting central banks is one and a half to two years (Kieler 2003). As is pointed out by Castelnuovo et al. (2003), the announcement of a range permits the central bank to clearly signal the uncertainty surrounding future price developments and the imperfect controllability of inflation,

38

Chapter 3

particularly at short horizons. Moreover a range can accommodate moderate and gradual variations in the optimal inflation rate over time. A possible drawback of a range objective is that its bounds could be seen as “hard edges,” as thresholds values. For that reason it is preferable to use a point objective. A point objective can also signal to economic agents a precise inflation target. The ECB’s definition of price stability provides a less clear-cut demarcation of the bank’s inflation preferences than an inflation target would. First, it does not follow from the definition of price stability that the ECB is indifferent between all inflation rates in the 0 to 2 percent range, or that it aims for the midpoint of that range. Second, the “medium-term” horizon over which price stability is to be maintained is not specified. The ambiguity about the ECB’s objective may hamper communication and understanding of its policies (Kieler 2003). The “clarification” that price stability means an inflation rate “below but close to 2 percent” is considered to be an improvement by most observers, although many argue that some ambiguity remains. The argument by the ECB that this “clarification” is not a policy change is also contested. At the press conference on the outcomes of the evaluation of the monetary policy strategy of the ECB, Issing declared that the close to 2 percent objective is not a change: “it is a clarification of what we have done so far, what we have achieved—namely inflation expectations remaining in a narrow range of between roughly 1.7 percent and 1.9 percent—and what we intend to do in our forwardlooking monetary policy.” However, as pointed out by Svensson (2003), in earlier statements the ECB called an inflation rate of about 1 percent in line with the objective. No matter whether it is a continuation or a change of policy, some ambiguities remain even after the “clarification.” For one thing, it is still not fully clear what “close to” 2 percent means. Furthermore it remains unclear to what time frame the medium term refers. For instance, if a shock pushes inflation above the 2 percent level, how long will it take the ECB to restore price stability? This question is quite relevant as inflation in the euro area was often above the 2 percent level in 2001 to 2002 (see figure 3.3). The ECB refused to give a clear position on this issue because of the uncertainty about the monetary transmission mechanism. As figure 3.3 shows, the ECB was successful in keeping inflation at a low level in the euro area. Nevertheless, inflation was sticky at a level above the ECB’s medium-term objective. The European Commission

The ECB’s Strategy: An Assessment

39

3.5 3 2.5 2 1.5 1 HICP 0.5

Jul-02

Mar-02

May-02

Jan-02

Nov-01

Sep-01

Jul-01

Mar-01

May-01

Jan-01

Nov-00

Sep-00

Jul-00

Mar-00

May-00

Jan-00

Nov-99

Sep-99

Jul-99

Mar-99

May-99

Jan-99

0

HICP excluding energy and unprocessed food

Figure 3.3 Inflation in the euro area, 1999 to 2002. (Source: Data provided by De Nederlandsche Bank)

(2002) interpreted this to be due to some extent to steady inflationary pressures in the services sectors. Since the service sector is the single most important component of the euro HICP basket, at a weight of close to 40 percent, its precipitous rise put upward pressure on euro area core inflation and made it difficult for inflation to work its way down.5 3.3.2 Is the Below 2 Percent Inflation Objective Too Low? Figure 3.3 also brings us to the question of whether the ECB has been too deflationary. Is a maximum inflation rate of 2 percent (even if it refers to the medium term) too low? The following arguments have been put forth in defense of setting a higher maximum rate of inflation: •

There can be an upward measurement bias in inflation.

•

There is less risk of deflation.

The Balassa-Samuelson effect calls for a higher ECB inflation objective.

•

•

It can create more room to stabilize the economy.

•

Inflation tends to “grease” the economy.

We will briefly summarize these arguments.

40

Chapter 3

It has long been recognized that due to measurement bias, measured inflation is likely to overstate actual inflation. As pointed out by Sibert (2003), there are several reasons for this. First, price indexes are calculated by comparing the price of a basket of goods consumed in a base year with the price of the same basket consumed in the current year. This overstates inflation because it does not take into account that consumers may substitute goods if prices rise. Second, improvements in the quality of goods may cause price changes to be overestimated. The quality of goods typically increases over time, and this, for example, makes computers bought at two different points in time not directly comparable. Ignoring this quality change will induce a measurement error in the price index. Third, prices of new goods often fall rapidly in the first few years right after their introduction. It may even be several years before goods are included in the basket of goods used to calculate the price index, and thus the fall in their prices may be missed. An important problem in handling new goods is also that there is no price in the base year that can serve as a reference. Finally there may be an outlet bias resulting from shifts in market shares among retailers that are not reflected in the composition of the sample outlets used in the survey. The publication of the so-called Boskin Report in 1996 gave a new impetus to this issue. The Boskin-commission concluded that the US CPI had been overstating the annual rise in the cost of living (COL) by about 1.1 percentage points. A recent study by Lebow and Rudd (2003) concludes that the size of the upward measurement bias is about 0.9 percentage points for the United States, with a plausible range of about 0.3 to 1.4 percentage points. These measurement problems led Alan Greenspan to conclude that a specific numeric inflation target for monetary policy would be unhelpful and misrepresentative.6 In contrast, the Governing Council of the ECB believes that the quality of the HICP makes it possible to provide a precise definition of price stability in the euro area (Issing 2001b). There are no broad studies of the accuracy of European consumer price indexes comparable to those conducted in the United States by the Boskin commission. As pointed out by Wynne and Rodríguez-Panzuela (2002), after the appearance of the Boskin Report, several studies were conducted within European countries to assess mis-measurements in their national CPIs. The conclusions differed across countries, but they reinforced those of the Boskin commission, though with effects that were

The ECB’s Strategy: An Assessment

41

generally smaller than for the United States.7 Putting a precise number on the degree of measurement error in the HICP is premature, according to Wynne and Rodríguez-Panzuela (2002). Despite the growth in the number of products that was studied for possible measurement error, there was still not enough information for setting a point estimate. Interestingly, at a recent ECB-CEPR workshop on this topic, evidence was presented that suggests the measurement bias for certain items to be downward and not upward, contrary to the widely held belief that measurement errors always overstate the true rate of inflation (see Camba-Mendez et al. 2002). If the measurement bias in the euro were the same as in the United States, an inflation objective of “close to 2 percent” would be sufficient to prevent deflation in the euro area. This is important because deflation may be very harmful (see Buiter 2003; Sibert 2003).8 Deflation involves many of the costs of inflation. First, there are the “menu costs” associated with changing prices that occur whether prices go up or down. Second, like inflation, deflation exacerbates the distortions inherent in tax and benefit systems considerably (Feldstein 1999). Third, unanticipated price changes redistribute wealth. Inflation redistributes it from creditors to debtors; deflation redistributes it from debtors to creditors. Sibert (2003) argues that the redistribution due to unexpected deflation may be more costly than the redistribution due to unanticipated inflation. Defaults may occur, and the resulting bankruptcies and restructurings destroy real wealth. The deterioration in debtors’ balance sheets brought about by unexpected deflation can thus lower both consumption and investment demand. According to Mishkin (1997), deflation has been a key factor in the promotion of financial instability in industrial countries. In the case of persistent deflation, there is the risk of a deflationary spiral of falling prices, output, profits, and employment. Aggregate demand-induced deflation can cause unemployment if nominal wages are downward rigid. With sticky wages, as prices decline, real wages rise, profit margins fall, and employment is cut back. A deflationary cycle is set off. Akerlof et al. (1996) show that at an annual rate of deflation of 1 percent and downward rigidities in nominal wages, unemployment in the United States could rise from a long-run equilibrium rate of 5.8 percent to 10.0 percent (see below for a discussion of how inflation can “grease” the economy). It should be obvious that it is better to prevent deflation than to try to cure it. Even if the inflation objective of “close to 2 percent” is high

42

Chapter 3

enough for the euro area as a whole to avoid deflation, deflation can still occur in a member state, depending on how large the inflation differentials are in the Union. If significant inflation differentials within the euro area should occur, some countries would experience deflation, whereas the euro area as a whole would not. In fact, according to the International Monetary Fund the risk of a mild deflation in Germany is “considerable” (see Kumar et al. 2003). Inflation differentials could be due to both temporary and structural factors. Temporary factors include differences in national fiscal policies, differences in pass-through patterns of external price and costs shocks (see section 3.6.1), and different cyclical positions of the various economies (see section 5.4). Temporary inflation differentials in a monetary union do not pose a real problem for monetary policy making. A structural factor could be the Balassa-Samuelson effect. Due to higher productivity growth the countries catching up will have a structural higher rate of increase of prices than more advanced countries. The differentials in sectoral productivity growth that result from the process of catching up and convergence of living standards across countries are an equilibrium phenomenon that does not require corrective policy actions. However, if substantial inflation differentials exist due to the Balassa-Samuelson effect, a stable aggregate price level for all countries taken together, may mean a deflation in the more advanced countries. Likewise a constant price level in the more advanced countries may mean inflation in the aggregate. According to Sinn and Reutter (2001), the inflation differentials resulting from the Balassa-Samuelson effect suggest the need for a reconsideration of the close to 2 percent inflation objective announced by the ECB (see section 5.5 for a further discussion of inflation differentials in the euro area). These authors argue that the ECB should tolerate a substantially higher inflation rate in Europe than the Bundesbank aimed for in Germany. In their words (Sinn and Reutter, 2001, p. 2): Euroland is bigger and more diverse than Germany, and for the time being substantial relative price changes have to be accounted for. These relative price changes require an inflation rate for Euroland that is higher than the one appropriate for any single European country if the risk of local deflation is to be avoided. If, say, Ireland needs a price increase relative to Germany, then it makes little sense to prescribe Germany’s previous idea of price stability to the average European price index, because it might require deflation for Germany or at least a dangerously low inflation rate. Thus, the Bundesbank’s goals should not be taken over by the ECB on a one-to-one basis.

The ECB’s Strategy: An Assessment

43

This conclusion crucially depends on how important the BalassaSamuelson effect is in the euro area. As we will show in section 5.5, there is little consensus in the literature on this issue. Furthermore, as pointed out by the ECB (2003b), the size of the Balassa-Samuelson effect for countries currently in the euro area is likely to diminish in the future, given that there has been substantial convergence among countries in terms of per-capita GDP. It may be more important for the acceding countries (see section 6.3). As for the fourth argument, Summers (1991) frames it as follows: since nominal interest rates cannot fall below zero, monetary policy faces a trade-off between a very low level of inflation and stabilizing the economy. If the economy is faced with a recession when inflation is zero, the monetary authority is constrained in its ability to engineer a negative shortterm real interest rate to counter the output loss. This constraint reflects the fact that the nominal short-term interest rate cannot be lowered below zero, the so-called zero interest rate bound.9 The zero bound constraint renders the risks of deviating from an inflation rate of zero asymmetric. As Alan Greenspan has noted, “. . . deflation can be detrimental for reasons that go beyond those that are also associated with inflation. Nominal interest rates are bounded at zero, hence deflation raises the possibility of potentially significant increases in real interest rates.”10 Japan is often mentioned as an example of a country where this problem has occurred. Krugman (1998) and Svensson (2000) have warned the United States may experience the same situation. Several studies have tried with small macro models to assess the likelihood of nominal interest rates hitting the zero lower bound for various levels of inflation objectives. While results in this area are highly uncertain and depend on a number of specific assumptions, most available studies indicate that the likelihood decreases to very low levels when the objective of the central bank is set at an inflation rate above 1 percent.11 Coenen et al. (2003) find that if the economy is subject to stochastic shocks similar in magnitude to those experienced in the United States over the 1980s and 1990s, the consequences of the zero bound are negligible for target inflation rates as low as 2 percent. The effects of the constraint are nonlinear with respect to the inflation target and produce a quantitatively significant deterioration of the performance of the economy with targets between 0 and 1 percent. This conclusion

44

Chapter 3

is based on stochastic simulations of a small structural rational expectations model, with monetary policy modeled as a Taylor type of interest rule. Reifschneider and Williams (2000) show that the zero bound can constrain policy in cases of very low inflation where the policy is based on the Taylor rule. Their simulations show that the nominal funds rate can become stuck at zero over 10 percent of the time. Nevertheless, a large reduction in the unfavorable effects of the zero bound can be realized by a simple modification to the Taylor rule, by incorporating a response to past constraints on policy. Viñals (2001) has compared the chances of the euro area hitting the zero lower bound with those of the United States. His results suggest that the euro area faces a smaller chance of hitting the zero lower bound because of its structural characteristics. Even if the zero bound were hit, monetary policy would not be impaired because there are other mechanisms by which monetary policy operates in the economy. First, as Krugman (1998) emphasizes, there is the importance of lifting expected inflation to lower the market real interest rate at the zero lower bound. In announcing an inflation target in a deflationary environment, the central bank can raise inflationary expectations, but the central bank cannot force higher expected inflation.12 Second, as Meltzer (2001) and others have argued, transmission mechanisms other than the nominal interest rate are effective at the zero lower bound. Bernanke and Gertler (1995), for example, show that the credit channel mechanism is not hampered by a lower zero bound on interest rates. Third, as Buiter and Panigirtzoglou (1999) and Goodfriend (2000) have proposed, a carry tax on money (both currency and bank reserves) can be introduced to circumvent the interest floor. The possible success of this suggestion depends on the availability of technologies that make it feasible without high costs. Fourth, as Svensson (2001) has shown, several approaches might be combined to escape the liquidity trap if and when, as Ullersma (2002) argues, the central bank can devaluate the exchange rate of the domestic currency without limit. Akerlof et al. (1996, 2000) argue, for different reasons, that a moderate level of inflation provides “grease” to the price and wage setting process.13 The economic adjustment of relative prices to shocks can become sluggish in the presence of downward nominal rigidities in wages and prices. For instance, at a zero inflation rate, individual firms facing an

The ECB’s Strategy: An Assessment

45

adverse firm-specific shock will not be able to secure real wage reductions in the presence of downward nominal wage rigidity, so they will, instead, lay off workers. Likewise, at low levels of inflation, a significant part of the price and wage setters may ignore or underweigh anticipated inflation in setting future prices. A moderate level of inflation provides for some real wage flexibility, which reduces the natural, or long-run, rate of unemployment.14 Akerlof et al. conclude that large permanent reductions in unemployment can be obtained by moving from either a very high or very low rate of inflation to a moderate rate, which they estimate for the United States to be 2 to 4 percent. However, Gordon (1996) argues that the prediction that a lower rate of inflation would imply a higher permanent level of equilibrium unemployment is not confirmed by the recent historical evidence of the United States, nor by cross-country analysis of the relationship between inflation and unemployment. Furthermore there may also be a “sand effect” on the natural rate of unemployment. When inflation rises, money illusion dissipates and the burden of price uncertainty rises (Groshen and Schweitzer 1999). So the long-run Phillips-curve is not vertical; the natural rate increases with inflation and not necessarily monotonously. As pointed out by Wyplosz (2001), the grease and sand effects are not mutually exclusive: at very low levels of inflation the grease will dominate, and the sand will set in only when inflation becomes more variable. Wyplosz (2001) examines the relationship between steady state unemployment and inflation in four European countries: France, Germany, the Netherlands, and Switzerland. He concludes that there is evidence for the grease effect to increase the natural rate of unemployment by some to 2 to 4 percentage points in the middle of the ECB’s inflation target range. To reduce that effect, the ECB should aim at an inflation rate of more than 5 percent.15 According to the ECB (2003b), however, the empirical evidence on the importance of downward nominal rigidities for the euro area is not conclusive. From the distribution of changes in the euro area price indexes, nominal price cuts do not appear to be as uncommon as often believed. For instance, the fraction of the euro area HICP that displayed negative year-on-year price changes in March 2002 was 11 percent. At the time inflation was 2.5 percent. In December 1998, when inflation stood at 0.8 percent, 20 percent of all categories exhibited negative price changes (Kieler 2003).

46

Chapter 3

In regard to wage-setting behavior, most studies have found a concentration of wage changes around the zero mark. Yet micro-based studies show a substantial proportion of wage earners to have experienced wage cuts. Beissinger and Knoppik (2001) document in their panel data, for example, that a large portion of wage cuts in Germany affected up to 20 to 30 percent of the blue-collar workers. Similarly, using the German socioeconomic panel data, Decressin and Decressin (2002) find that about one-fifth of the workers retaining their jobs experienced nominal cuts in wages. In these cases the level of inflation affected the distribution of real wage changes. Decressin and Decressin conclude that “the results for Germany, which has experienced inflation at about 2 percent per year on average over the 1990s, suggest that insufficient wage flexibility does not make a compelling case for the ECB to adopt a higher inflation target. Nonetheless, some ‘sand’ thwarts the functioning of the wage setting mechanism in Germany. . . . Specifically, the presence of a nominal rigidity at the zero mark for base wages suggests that pushing inflation much below 2 percent could bear risks” (p. 30). In contrast, using German micro data (IBABeschäftigtenstichprobe), Knoppik and Beissinger (2003) report a high degree of downward nominal wage rigidity. They show that a very low inflation policy can have expensive side effects. The high downward nominal wage rigidity is detrimental for individual expected wage growth, the aggregate wage level, and equilibrium unemployment when inflation rates are lower than 3 percent. When inflation rates are above 4 percent, these effects are negligible. At the macro level the presence of downward nominal rigidities has been usually associated with nonlinearities in the Phillips curve at low inflation (Ball et al. 1988). Downward nominal wage rigidities can give rise to nonlinearities in the Phillips curve at near-zero inflation. Because a larger fraction of workers is bound by the downward rigidity, the aggregate wage responds less to a negative shock to labor demand than to a positive shock of equal magnitude. The slope of the Phillips curve also becomes flatter at inflation rates close to zero, implying that a larger change in unemployment is needed to produce a certain change in wage inflation. Although there is some evidence for nonlinearities, there are as many findings to the contrary (see Kieler 2003 for a discussion). However, because of the lack of prolonged periods of very low inflation, the analysis of this issue is considerably clouded by the scarcity of evidence. It should additionally be noted that a positive trend in productivity growth permits firms to reduce labor costs per

The ECB’s Strategy: An Assessment

47

unit of output without necessarily cutting nominal wages. Finally, even if downward nominal rigidities were pervasive, there remains the question whether “accommodating” them with a higher inflation rate makes this undesirable structural feature of some economies not even more “entrenched” (ECB 2003a). Coenen (2003) investigates the importance of the existence of downward nominal wage rigidity, using stochastic simulations of a small estimated model of the euro area in which wage contracts are staggered. To establish downward nominal wage rigidity, it is assumed that the changes in the nominal wage contracts are nonnegative. Furthermore, monetary policy makers are assumed to follow a standard Taylor rule. The results suggest that the asymmetry of wage adjustments causes a nonvertical long-run Phillips curve in which output falls increasingly short of potential with lower inflation targets. The output loss, which depends on the degree of inflation persistence, is in the order of one-eighths to one-fourths of a percentage point even when the inflation target is set at zero. 3.4 Role of the First Pillar Jaeger (2003) argues that initial ECB comments on its strategy suggested that the first pillar would not only be the prominent but provide the dominant input into policy decisions. When ECB President Duisenberg was asked on the relative importance of the two pillars, he noted that “. . . it is not a coincidence that I have used the words that money will play a prominent role. So if you call it the two pillars, one pillar is thicker than the other is, or stronger than the other, but how much I couldn’t tell you.” With M3 growth persistently exceeding its reference value, increasing stress was put on the need to undertake a “comprehensive monetary analysis” that goes beyond “a comparison between M3 growth and the reference value.” The first pillar in the monetary policy strategy of the ECB has been severely criticized. One reason why the use of money growth has been criticized is that the outcome of the analysis of the monetary analysis is often at odds with the results of the economic analysis.16 No doubt, this is true for more recent years. However, in the early years of the ECB this conflict was not apparent. Table 3.3 summarizes the main comments of the president of the ECB referring to the growth rate of M3 and its components when he explained the ECB’s monetary policy decisions in the years 1999 to 2001. It also shows the actual decision.

48

Chapter 3

Table 3.3 Role of M3 growth in monetary policy decisions of the ECB, 1999–2001 Council meeting

Statements on M3 growth

December 22, 1998 (no change)

Latest three-month moving average annual growth rate of M3 stood at 4.5%. This was exactly in line with the reference value set by the Governing Council. Actual situation, characterized by monetary growth compatible with continued price stability justifies maintaining the current stance of monetary policy. The 12-month growth rate of M3 decreased from 5% in October to 4.5% in November 1998. The latest three-month moving average of M3 growth was approximately 4.7% and has remained close to the reference value. The Council has not altered its assessment regarding the outlook for price developments. Council confirmed its earlier assessment that the outlook for price stability remains favorable. Latest 3-month moving average of annual M3 growth was 4.7% and has remained very close to the reference value. As monthly data on monetary aggregates can be volatile, M3 growth is monitored on the basis of three-month moving average of the 12-month growth rate of M3, which stood at 4.9%. In view of the uncertainty relating to special factors pertaining to the changeover to stage 3, the Council does not consider the acceleration as a signal of future inflationary pressures. However, close monitoring remains necessary to give more conclusive evidence of the underlying causes and the permanent or temporary nature of the rise in M3 growth. The 12-month growth rate of M3 declined from 5.6% in January to 5.2% in February, largely reflecting a slowdown in growth of overnight deposits. Three-month moving average increased by 0.2 percentage point to 5.1%. The Council does not regard current monetary trends as constituting a signal of future inflationary pressures as it is still close to the reference value and considering that it may to some extent mirror the specific environment related to the start of stage 3. Monetary developments do not signal dangers to price stability over the short and medium term. The 12-month rate of growth of M3 remained at 5.1%. The latest figure for the 3-month moving average was 5.2%, which was 0.2 percentage points higher. Trends should not be seen as warning signal with regard to future evolution of inflation given special circumstances and the fact that growth remains relatively close to reference value. Twelve-month growth rate of M3 decreased from 5.2 to 4.9% in April, mainly due to decline in short-term deposits. Three-month moving average fell to 5.0 from 5.3%. Current monetary developments are in line with price stability over the medium term.

January 7, 1999 (no change)

February 4, 1999 (no change)

March 4, 1999 (no change)

April 8, 1999 (reduction 50 basis points)

May 6, 1999 (no change)

June 2, 1999 (no change)

The ECB’s Strategy: An Assessment

49

Table 3.3 (continued) Council meeting

Statements on M3 growth

July 15, 1999 (no change)

Three-month moving average of M3 growth increased slightly to 5.2%. Major contribution came from overnight deposits. Annual growth rate of total credit rose from 7.3 to 7.9% in May. Growth of loans to private sector rose to 10%. Situation not seen as threatening price stability, but reassessment may be appropriate if money and credit growth increase further. Twelve-month growth rate increased from 5.6 to 5.3%. Latest three-month average rose from 5.3 to 5.4%. Growth rate has been moving away from reference value. Upward trend may be explained by very low opportunity costs of holding monetary assets but perhaps also by gradually improving economic conditions. Annual growth rate of credit to private sector remained high, although it declined to 10.4%. Three-month growth rate increased to 5.6% from 5.5%. Twelvemonth growth rate displayed upward trend during 1999. Credit to private sector expanded at 10.7%. On balance, monetary conditions signal rather generous liquidity situation in the euro area. Main argument for increase of rates by 50 basis points was that balance of risks for price stability has gradually moved to the upside. Several indicators, including monetary growth, suggest that there is ample liquidity in the euro area. M3 growth is on rising trend. Three-month growth rate is 5.9%. Strong growth of most liquid components of M3 is noteworthy, suggesting that low interest rates favored strong growth of money. The pickup in economic activity is likely to have further stimulated M3 growth. Credit to private sector remained strong expanding at rate above 10%. Data confirmed picture of generous liquidity situation. Annual growth rate of M3 in October was 6%, compared to 6.2% in previous month. Three-month average was unchanged at 6%. Growth of most liquid components of M3 remained very dynamic. Annual growth rate of credit to private sector continued to exceed 10%. Developments are driven by low level of interest rates and pick-up of economic activity. Three-month growth rate of M3 was 6%. Current liquidity conditions continue to be generous. This is confirmed by continued strong growth of loans to the private sector (10.1% in November compared to 9.9% in October). Both monetary and credit growth continue to signal generous liquidity conditions. Monetary growth has been consistently above reference value and credit growth to the private sector is continuing to grow at more than 10%. Three-month average growth rate rose to 6.1%. At 10.5% annual growth rate of credit to the private sector continued to be high. Monetary and credit developments remain important factors contributing to upside risk to price stability.

September 9, 1999 (no change)

October 7, 1999 (no change)

November 4, 1999 (increase 50 basis points)

December 2, 1999 (no change)

January 5, 2000 (no change)

February 3, 2000 (increase 25 basis points)

50

Chapter 3

Table 3.3 (continued) Council meeting

Statements on M3 growth

March 2, 2000 (no change)

Three-month growth was 5.7 compared with 6% in last quarter. Reduction was determined by a base effect as monthly increase in January 1999 was exceptionally strong. Private sector credit growth remained high at 9.5%. Overall prolonged deviation from reference value and dynamic credit growth indicate that liquidity conditions remain generous. Prolonged deviation of M3 growth from reference value points to existence of ample liquidity, especially when seen in conjunction with continued strong growth of credit to the private sector. Annual growth rate of M3 was 6.2%, implying a small increase in three-month moving average growth rate. Annual growth of M3 was 6.2%, which implied small increase in 3-month moving average to 5.9% from 5.8%. Picture of generous liquidity conditions is confirmed by strong annual growth of loans to private sector at 10.5%. Three-month moving average rose to 6% from 5.9%. Annual increase in credit to the private sector rose to 10.9 from 10.4%. These figures confirm earlier assessment that liquidity continued to be ample in the euro area in early 2000. Strong growth of money and credit throughout 1999 and the pronounced expansion of money and credit aggregates over the first four months of 2000 have strengthened the view that liquidity conditions are ample. Three-month moving average growth rate of M3 was 6.3%. Expansion of credit to the private sector exceeded 11% in April 2000. Three-month moving average stood at 6.3%, unchanged, but substantially above reference value. This, combined with the strong growth of both M1 and credit to the private sector, indicates that liquidity conditions continued to be ample in the euro are through May. Increase in interest rates of June exerts moderating influence on money and credit growth. Although measures taken since November 1999 are gradually feeding their way through, latest available information indicates that M3 growth has continued to deviate on the upside from the reference value. At the same time the expansion of credit to the private sector suggests that households and firms regard financing in the euro area as being very favorable. In the context of the robust expansion of economic activity, a continuation of ample liquidity conditions would constitute a risk to price stability.

March 16, 2000 (increase 25 basis points) March 30, 2000 (no change) April 13, 2000 (no change)

May 11, 2000 (no change)

June 8, 2000 (increase 50 basis points)

July 6, 2000 (no change)

August 31, 2000 (increase 25 basis points)

The ECB’s Strategy: An Assessment

51

Table 3.3 (continued) Council meeting

Statements on M3 growth

September 14, 2000 (no change)

M3 growth has been showing some signs of moderation over recent months, mainly as consequence of lower growth in the most liquid components of M3, which bear little or no remuneration, indicating that the policy measures taken since November 1999 are gradually feeding their way through. At the same time growth of M3 at 5.5 continued to deviate on the upside from the reference value. Pronounced expansion of credit to the private sector suggest that households and firms continue to regard financing conditions as favorable, which constitutes a risk to price stability. M3 growth declined to 5.3% from 5.5%. German UMTS auction appeared to have a temporary upward impact on M3 growth. Short-term developments in monetary aggregates seem therefore to be more moderate than August data would suggest. Still, liquidity conditions continue to be ample given protracted deviation of M3 growth from reference value, coupled with still relatively high growth of loans to private sector. M3 growth stood unchanged at 5.4% While remaining above reference value, M3 growth has shown signs of moderation. Growth rate of credit to private sector increased to 10.8%, but this was influenced by financing of UMTS licenses. Growth rate of MFI credit to euro area residents remained broadly unchanged. M3 growth rates have shown signs of moderation over the past few months. However, taking into account protracted upward deviation from reference value and still robust growth of growth of credit to private sector, caution continues to be warranted with regard to upside risks to price stability. M3 growth was slightly lower at 5.0%. Therefore, after recent slowdown in monetary growth, the risks to price stability from the monetary side have become increasingly balanced. Caution still needs to be exercised, however, given the continued dynamics in credit growth to the private sector and the upward deviation of M3 form the reference value in the past. Moderation of the growth of monetary aggregates continued. Three-month average of M3 growth stood at 5.0%, which was slightly lower. M3 growth has been on gradual downward trend since spring 2000. Growth rate reached 4.8%. Growth in credit aggregates has also been less buoyant over the past few months. All in all, information from the first pillar is signaling that upwards risks to price stability have diminished over the past few months.

October 5, 2000 (increase 25 basis points)

November 2, 2000 (no change)

December 14, 2000 (no change)

February 1, 2001 (no change)

March 1, 2001 (no change) April 11, 2001 (no change)

52

Chapter 3

Table 3.3 (continued) Council meeting

Statements on M3 growth

May 10, 2001 (decrease 25 basis points)

M3 growth has been on gradual downward trend since spring 2000. Growth rate stood at 4.8%, close to reference value. Slowdown in credit aggregates has also been visible in past few months. In addition there have been indications that the monetary aggregates are distorted upward by non–euro area residents’ purchases of negotiable paper included in M3. Distortion has become more sizable over recent months and amounts to half a percentage point. There is an additional distortion due to nonresidents holdings of other marketable paper included in M3. Taking into account these upward distortions, as well as all information from the first pillar, it can be concluded that there is no longer a risk to price stability over the medium term emanating from the first pillar. M3 growth has been on gradual downward trend since spring 2000, driven predominantly by its most liquid components. M3 growth has stabilized. M3 adjusted stood at 4.6%. Growth rate of loans to the private sector continued to moderate over recent months. M3 adjusted stood at 4.9%. As indicated earlier, M3 is currently distorted by holdings of money market paper and short-term debt securities of non–euro area residents. As a result three-month moving average was broadly in line with reference value. In addition growth of credit continued to be moderate. Three-month average growth of M3 was 5.9%, but this figure needs to be corrected for holdings of money market paper and short-term debt securities by nonresidents of the euro area, which according to preliminary estimates have contributed around threequarters of a percentage point to annual M3 growth. Increase of M3 growth reflects relatively flat yield curve and recent weakness in stock markets which make holdings of short-term assets attractive. Growth may be transitory and does not necessarily have implications for price stability in the medium term. Loans to private sector declined over recent months. Strong increase in M3 growth over the past few months. However, a number of temporary factors play an important role (flat yield curve and uncertainty in stock markets). Furthermore growth of credit to private sector continued to slowdown. Therefore monetary developments do not signal risks to price stability at this juncture. M3 growth increased further in September but needs to be seen as reflection of an increased preference of investors. In addition growth of credit to private sector has continued to decline over recent months. Overall, current monetary developments do not signal risks to price stability in the medium term.

June 7, 2001 (no change)

July 5, 2001 (no change)

August 30, 2001 (decrease 25 basis points)

October 11, 2001 (no change)

November 8, 2001 (reduction 50 basis points)

The ECB’s Strategy: An Assessment

53

Table 3.3 (continued) Council meeting

Statements on M3 growth

December 6, 2001 (no change)

M3 growth rose to 6.8%, significantly above reference value. Investors have shifted their portfolio toward liquid and relatively safe short-term assets included in M3. Such shifts should be temporary and not indicating future inflationary pressures. Assessment underpinned by the fact that credit growth to private sector has been continuously falling over recent months.

In 1999 the developments in the first pillar were broadly in line with decisions taken. In April the twelve-month growth rate of M3 declined, and the three-month moving average growth rate was close to the reference value.17 In November the opposite pattern can be discerned. Apart from these growth rates of M3, the ECB also often referred to the growth rate of credit to the private sector when it reported on developments under the first pillar. Basically the same pattern emerged for the interest rate increases in the first half of 2000. M3 growth remained substantially above the reference value and increased. This situation, in conjunction with the high growth rate of credit to the private sector, contributed to the decisions taken by the Governing Council. In the latter part of 2000 the situation became less clear-cut. Although M3 growth remained above the reference value, it started to decline. This is, of course, the kind of situation where the decision-making process is the most difficult. Nevertheless, given that M3 growth was still substantially above the level that the ECB considers consistent with price stability and that the growth rate of credit to the private sector was still very high, the increase in interest rates in October 2000 should not be a real surprise. In 2001 M3 growth declined further, getting close to the reference value. So from this perspective the cut in interest rates in May 2001 was in line with this development. Still there was quite some confusion, mainly caused by the statements about the distortion of M3 growth due to the holdings of assets by non-euro area residents. This caused the impression that the ECB was looking for some kind of argument to ensure that developments under the first pillar were in line with the decision taken. Careful reading of the previous announcements of Duisenberg shows that the ECB had announced a number of times before that M3 growth was on a downward trend. The decision to cut interest rates further in August and November 2001 was harder to

54

Chapter 3

understand from this perspective, as M3 growth was increasing again. According to the ECB this was not worrisome, given the distortions that were estimated to be three-quarters of a percentage point, and temporary factors. During this period it is easy to understand why observers started to talk about smokescreens when referring to the first pillar. And in more recent periods it became questionable whether money should play a separate role in the monetary policy strategy of the ECB. Obviously a target that is specified in terms of the final objective is more comprehensible than a target with an uncertain relation to the ultimate objective. Hence it is a more effective communication device between the ECB and the public (Cukierman 2001a). Still, according to the ECB, there are arguments in favor of maintaining the M3 reference value (Angeloni et al. 1999). First, since in the long run money is a major, if not the most important, determinant of inflation, it is reasonable to take inflation into account in deciding about monetary policy. Although the long-run correlation between money growth and inflation in historical data is high, the correlation for shorter horizons of around one to three years is more relevant for monetary policy purposes. This correlation is much lower than the long-run correlation. In the short run real money growth is more variable, causing a relatively low short-run correlation between nominal money growth and inflation. This low correlation in the short run throws considerable doubt on the use by the ECB of the money growth indicator, namely M3 growth relative to the reference value. According to the ECB, substantial or long-lasting deviations of monetary growth from the reference value could signal risks to mediumterm price stability under normal circumstances. However, the growth rate of money in the euro area has signaled an acceleration of the inflation rate for many years, although inflation has not increased. Second, given the uncertainty about the transmission mechanism of monetary policy decisions, it may be better not to discard money growth. However, this is not a reason why money should be given a “prominent role”; monetary developments could be taken into account in the second pillar. Finally, since the Bundesbank, which is generally regarded as highly credible (see chapter 5), announced targets for M3 growth for many years, the continuation of this practice may help the ECB to inherit some of the credibility of its predecessor. However, the German central

The ECB’s Strategy: An Assessment

55

bank missed its money growth targets more often than it hit them. Furthermore the ECB has not literally followed the German practice of monetary targeting. As was pointed out before, the ECB analyzes the relationship between actual money growth and the pre-announced reference value. To the extent that deviations of money growth from the reference value indicate a threat to price stability, they will lead to a monetary policy response. The ECB never used the term “intermediate target,” but employs the term reference value to indicate that it does not change its interest rates in a mechanistic fashion. What are the conditions to be met when the money stock is to be used as an intermediate target of monetary policy? First, the money stock must serve as a leading indicator for the inflation rate. Second, with its operating target(s) the central bank should be able to control the money stock. Finally, the demand for money must be sufficiently stable. The euro area money demand function is relatively stable in comparison with the demand for money in other economies such as those of the United States, the United Kingdom, and Japan. Calza and Sousa (2003) give some explanations for this. First, countryspecific factors have influenced money demand in other economies. Second, the effect of financial innovation on money demand has been weaker in the euro area. Finally, aggregating data across countries is beneficial for the stability of the euro area money demand. It is expected that these aggregation gains will remain because the crosscountry differences will always be present. In the future money demand stability could decrease in the euro area if the share of wealth held in financial assets increases, because in that case portfolio shifts to and from bond and stock mutual funds would increase. The stability of the demand for money is necessary but not sufficient for monetary targeting. If the operating targets of the ECB do not have a systematic influence on the money growth rates, monetary targeting becomes inapplicable. This problem is thought to exist because the money stock has shown little reaction to changes in the short-term nominal interest rates in the euro area (Bofinger 1999). Rudebusch and Svensson (2002) show that money growth targeting to stabilize money growth in the euro area would be a very bad monetary strategy. This is because it can cause very high variability of both inflation and the output gap. Current inflation and the output gap are much better predictors of future inflation at horizons of interest to monetary policy.

56

Chapter 3

In conclusion, we feel that the ECB should abandon the monetary analysis as a dominant navigation system for its monetary policy but use the relevant information in the economic analysis. 3.5 Inflation Forecasts Most academic observers favor some kind of inflation targeting. The second pillar has been criticized because it is not a real inflationtargeting strategy. Initially the ECB was even reluctant to publish inflation forecasts. According to the European Monetary Institute (EMI 1997, p. 16), the predecessor of the ECB, . . . there are conceptual difficulties in formulating forecasts conditioned on unchanged policies since some financial market indicators which provide input to these forecasts are typically influenced by the markets’ anticipated stance of monetary policy over the forecasting horizon. Furthermore, there is a risk that publishing inflation forecasts may, at times, have adverse effects on financial markets and wage and price setting and that the credibility of the ESCB could be damaged in the medium term if the conditional nature of the forecasts is not well explained.

However, if the meaning of forecasts is properly communicated— which is not easy—publication of the inflation forecasts may help the public understand and evaluate ECB policy decisions. It is important that forecasts be properly explained, as they may otherwise become self-fulfilling prophecies. Just publishing an inflation forecast is not a substitute for explaining the reasoning and assumptions underlying monetary policy. In December 2001 the ECB published for the first time its so-called projections. These projections are prepared by the staff of the Eurosystem. They are conditional on the assumption of unchanged short-term interest rates and are formulated in terms of ranges rather than point values; there are no mean or modal values published. The results are presented to the Governing Council for its meetings in June and December. According to Issing (2001b), the projections provide a counterfactual scenario that aims to facilitate the discussion in the Governing Council by pointing out the consequences of a lack of monetary policy actions over the project horizon. Therefore the projections do not represent the ultimate synthesis of the Governing Council’s assessment of the euro area inflationary outlook. They are just one of the inputs for the Governing Council in coming to a decision on the proper monetary policy. In this respect, the situation is very different from the United

The ECB’s Strategy: An Assessment

57

Kingdom, where the projections are made under the responsibility of the Monetary Policy Committee of the Bank of England. In our view it would be better if the two-pillar strategy of the ECB were replaced by explicit inflation targeting. A broad assessment of all relevant information variables regarding expected inflation, including broad and narrow money growth, is incorporated in an inflationtargeting strategy. According to Svensson (1999), there are three main other characteristics of inflation targeting. First, there exists an explicit quantitative inflation target. Second, policy decisions are based on conditional inflation forecasts. Finally, the strategy takes place in an institutional environment with a high degree of central bank accountability and transparency (see chapter 4 for a further discussion). 3.6 External Value of the Euro With the creation of the monetary union (most of) the EU economy has become relatively closed like that of the United States. Most trade flows consist of intra-EU trade. Most observers would therefore probably agree with Eichengreen (1998) that under these circumstances, the theory of optimum currency areas suggests that it is optimal to let the euro float. Still, even if the euro floats vis-à-vis most other major currencies, this does not imply a policy of benign neglect by the ECB. First, as explained in chapter 2, the euro–dollar exchange rate is one of the indicators of future inflationary developments in the monetary policy strategy of the ECB. After the start of the monetary union, the euro–dollar exchange rate declined; in 2002 an upward movement started (see figure 3.4). How important is the euro–dollar exchange rate for inflation in the euro area? This issue will be discussed in section 3.6.1. A second reason why the ECB closely follows exchange rates is that there is always a risk that actual exchange rates may be out of line with “fundamentals,” whatever they may be. Indeed, the concerted intervention of the ECB and other central banks in September 2000, was— according to an ECB press release—motivated by the “shared concern about the potential implications of recent movements in the euro exchange rate for the world economy.” In November 2001 the ECB intervened again, but now the central banks of the United States and Japan did not join in. In section 3.6.2 the usefulness of this kind of interventions will be discussed.

58

Chapter 3

1.30 1.20 1.10 1.00 0.90 0.80 0.70 1-1-04

7-1-03

10-1-03

4-1-03

1-1-03

7-1-02

10-1-02

4-1-02

1-1-02

7-1-01

10-1-01

4-1-01

1-1-01

10-1-00

7-1-00

4-1-00

1-1-00

10-1-99

7-1-99

4-1-99

1-1-99

0.60

Figure 3.4 Dollar–euro exchange rate, monthly data 1999 to 2003. (Source: http://research. stlouisfed.org/fred2/series/EXUSEU/95)

Table 3.4 Estimates of the effects of a 10 percent depreciation of the euro on euro area inflation Effect on inflation Model/paper

Price index

Short run

Medium run

Long run

INSEE (Chaney, 2003) Hüfner and Schröder (2002) Campa and Gonzáles Mínguez (2002) ECB-area wide model (Fagan et al. 2001)

HICP Import prices HICP

1% 4%

0.4% 5.5% 0.45%

0.8% 6%

HICP

1.6%

Source: Wyplosz (2003a). Note: Short run is up to three months, medium run is from six to twelve months, and long run is two years and more.

3.6.1 Why the ECB Cares about the Euro–Dollar Exchange Rate: Pass-through to Inflation The ECB has at various occasions made it clear that it closely monitors the euro–dollar exchange rate mainly because the initial decline of the euro since the start of EMU has affected inflation in the euro area.18 Table 3.4 presents some estimates of the effect of 10 percent depreciation of the euro on the inflation rate in the euro area as summarized by Wyplosz (2003a). By and large, the effect of a depreciation on HICP inflation is small and imprecisely known. However, the effect is far

The ECB’s Strategy: An Assessment

59

from negligible. The table also suggests that the effect on import prices is much larger than on HICP. The remainder of this section reviews the literature on the exchange rate pass-through (ERPT), meaning the degree to which exchange rate changes are passed through in (domestic) prices.19 Goldberg and Knetter (1997) define ERPT as the percentage change of import prices (in local currency) resulting from a 1-percent change in the exchange rate between the exporting and importing countries. Firms may choose to pass exchange rate alterations fully into their selling prices (complete ERPT). At the other extreme, firms may decide to absorb the shock by reducing their profit margins, so the selling prices will be unchanged (no ERPT). Incomplete ERPT may be caused by market power (see Feenstra et al. 1993). The relationship between exchange rates and prices is given by the following equation: pt = a + d X1 + g Et + y Zt + e r .

(3.1)

In equation (3.1), p is the local currency import price, and X is a measure of the export costs. Z may include import demand-shifting factors, such as competing prices or income, and E is the exchange rate (importer’s currency per unit of exporter currency). Taking the logarithm of both sides, yields the elasticity of ERPT with which g is referred to as the pass-through coefficient. Greater openness of the economy means more competition, and this puts downward pressure on the ERPT (see McCarthy 1999). However, according to Knetter (1993), the variations between industries are more important than the variations between countries. The market structure and trade restrictions differ across industries and influence the degree of ERPT. According to Feenstra et al. (1993), the ERPT will be one when the market share of a firm is zero. When the market share increases, the degree of ERPT will first decrease and then increase. When the market share is one the ERPT will be complete (g = 1). This theory is described in figure 3.5.20 The intuition behind figure 3.5 is the following: In a situation with strong competition (i.e., many firms having low market shares) changes in exchange rates are largely reflected in prices, because firms have low profit margins. When less competition is present, ERPT decreases because profit margins are higher. Exchange rate changes are therefore not necessary fully reflected in price changes. Firms want to keep their customers and therefore accept lower profits instead of increasing

ERPT elasticity

60

Chapter 3

1 0.8 0.6 0.4 0.2 0 0

0.2

0.4

0.6

0.8 1 Market share

Figure 3.5 Relationship between market share and ERPT-elasticity. (Source: Feenstra et al. 1993)

prices. When the market share is high, ERPT rises. Firms pass exchange rate movements through to prices, instead of decreasing profits. This is because the loss of customers is small when competition intensity is low. Corsetti and Dedola (2001) argue that incomplete ERPT may exist because of the difference in distribution and transportation costs between domestic markets and markets abroad. Even if imported goods are perfect substitutes for the domestically produced goods, they are not consumed in large amounts because their costs are higher than the costs of domestic goods. Consequently a change in the exchange rate will have a small effect on the consumer price index (a lower degree of ERPT). Another possibility is that the low degree of ERPT does not primarily reflect the stickiness of prices but optimal price discrimination. Imported goods can be regarded as intermediate products for which there are domestic substitutes available. So local producers can combine intermediate goods with local goods to produce a final product for the consumers. While the price for consumers is determined in local prices, the intermediate imports are fixed in the currency of the producer of that intermediate good. The importer is able to switch between the imported intermediate good and the locally produced alternative when the exchange rate changes (expenditure switching effect). According to Devereux et al. (1999), there are still many substitution possibilities with the domestically produced goods. Dornbusch (1987) used industrial organization models to express the relationship between exchange rate fluctuations and the domestic price level in terms of market concentration, import penetration, and the substitution possibilities of the imported and domestic products. The ERPT

The ECB’s Strategy: An Assessment

61

on domestic producer prices appears larger in industries that are characterized by a lower degree of concentration and a lower degree of import penetration. Pass-through is smaller in more segmented economies, which can participate in price discrimination. Countries with a larger import share should have a larger ERPT (see Goldberg and Knetter 1997). Most of the empirical literature on ERPT published in the 1980s focuses on the degree of pass-through in the United States. The growing importance of imperfect competition and strategic trade theory stimulated researchers to estimate the degree of ERPT at the level of industries. For instance, Feenstra et al. (1993) analyze the automobile industry and find that the degree of ERPT varies substantially between different importing and exporting countries. The majority of the empirical studies about ERPT of the last twenty years shows that ERPT is not complete. According to Goldberg and Knetter (1997), the average degree of pass-through for the United States Table 3.5 Exchange rate pass-through into import prices

Full sample pass-through elasticities

Change in pass-through elasticities, 1999 versus 1989

Country

Short run

Long run

Short run

Long run

Austria Belgium Germany Denmark Spain Finland France United Kingdom Greece Ireland Italy Netherlands Portugal Sweden

1.22a 0.16b 0.59a,b 0.56a,b 0.66a,b 0.69a,b 0.53a,b 0.39a,b 0.40b 0.79a 0.67a,b 0.75a,b 0.60a,b 0.67a,b

1.25 0.71 0.79a 0.68a 0.56a,b 0.82a 1.21a 0.47a,b -0.02b 1.37a 0.62a 0.77a 0.88a 0.59a,b

0.68 1.02a -0.32 -0.42 -0.40 -0.15 -0.82a 0.11 -0.61 -0.03 -0.52 -0.18 -0.04 -0.62a

0.24 0.18 -0.12 -0.80a -0.94a -0.02 -1.00 0.11 -1.24 -1.17 -0.65 -0.17 -0.26 -0.45

Source: Campa and Goldberg (2002). a. Significantly different from zero at 5 percent level. b. Significantly different from one at 5 percent level.

62

Chapter 3

is 50 percent. For illustrative purposes, table 3.5 reproduces the results of Campa and Goldberg (2002) for the EU countries in their sample. In general, the ERPT is higher in the long run than in the short run, but for all countries the ERPT is less than one both in the short and long run. Note that there are huge differences across the countries of the euro area. Campa and Goldberg (2002) compare the elasticities estimated over the first half of the sample, 1977 through 1989, with those for the full sample. The last two columns of table 3.5 show the results of this split sample approach. In only 3 of the 14 countries there was an increase in ERPT. In most of the countries the long-run and shortrun ERPT decreased, but the decline in ERPT was significant for only 3 of the 11 countries. The monetary union may effect ERPT in Europe. It is likely that the international role of the euro will become more important in the medium and long run. This implies that the euro will be accepted more often in international payments, in particular as a vehicle currency. Table 3.6 shows the influence of the currency of pricing on the ERPT to the retailer and the consumer. Devereux et al. (1999) consider that firms will always set their wholesales prices in their own currency, for sales to both the domestic and foreign retailers. This has not changed by the introduction of the euro. Table 3.6 Currency of pricing Wholesale prices before and after the euro Region of production

Region of consumption

Currency

Pass-through to retailer

United States Europe

Europe United States

US dollar Euro

Complete Complete

Retail prices before the euro Region of production

Region of consumption

Currency

Pass-through to consumer

United States Europe

Europe United States

US dollar US dollar

Complete Zero

Region of production

Region of consumption

Currency

Pass-through to consumer

United States Europe

Europe United States

Euro US dollar

Zero Zero

Retail prices after the euro

Source: Devereux et al. (1999).

The ECB’s Strategy: An Assessment

63

In contrast, retailers can set the retail prices either in the consumers’ currency or in the currency used by the firm. In the first case, retailers bear the exchange rate risk, while in the second case, the exchange rate risk is passed through to the consumers. It is assumed that after the introduction of the euro the European retailers set retail prices of imported goods in euro instead of dollar terms (the euro becomes a vehicle currency). Consequently the sensitivity of the European consumer prices to fluctuations in the exchange rate is reduced. So the ERPT will then be lower. Some recent studies have suggested that ERPT may influence optimal monetary policy rules. For instance, Corsetti and Pesenti (2002) show that the degree of ERPT and exchange rate exposure in markets at home and abroad are key elements determining the optimal monetary rules. Economies with a high degree of ERPT should, ceteris paribus, have a monetary policy focused on internal conditions. We conclude that according to most empirical studies, ERPT is incomplete in the short and in the long run for most countries, although ERPT in the long run exceeds short run ERPT. Exchange rate movements influence prices more in countries with a larger degree of ERPT. In countries with a high degree of ERPT exchange rate movements therefore contain information as to future inflation. Especially in the long run, exchange rate movements can lead to price changes, so they are important to predicting inflation. The introduction of the euro has probably reduced the ERPT to consumers because the European retailer sets the retail prices of imported goods in euro terms instead of dollar terms (the euro becomes a vehicle currency). This means that the sensitivity of the European consumer prices to fluctuations in the exchange rate has been reduced. The importance of exchange rate movements in predicting the level of inflation in the euro area might therefore decrease somewhat. 3.6.2 The Euro–Dollar Exchange Rate: Do Interventions Help? At the ECB’s initiative, the monetary authorities of Canada, Japan, the United Kingdom, and the United States joined to buy euros on September 22, 2000. According to Koen et al. (2001) a total of around 6 billion euro was probably bought, of which around 3 billion by the ECB. Within a few hours the euro jumped from 85 to 90 US dollar cents. It then settled around 88 dollar cents for about a week. However, depreciation resumed and by mid-October the euro had dropped below its previous lows.

64

Chapter 3

A subsequent round of unilateral interventions was undertaken in early November 2000, spread over three trading days (November 3, 6, and 9). This time only the national central banks of the Eurosystem stepped in. According to Koen et al. (2001) the amount bought was slightly less than in September. The euro reacted less markedly this time. It depreciated anew until late November and then appreciated for some time. The bad performance of the euro after the start of the monetary union was not expected. During most of the period after the start of EMU, the European economy was in good shape; inflation was relatively low, production growth was rising, and consumer confidence was high. Indeed, most studies of the equilibrium value of the euro lead to the conclusion that the euro was undervalued during 2000 to 2001 (see table 3.7). Various papers have tried to link the euro–dollar exchange rate to news about recent developments in order to explain the unexpected weakness of the euro, but with little success (see sidebar 4). The Eurosystem can intervene either multilaterally or bilaterally. Unlike the United States and Japan, the Eurosystem does not have to wait for specific governmental instructions if it wants to intervene. An unannounced intervention may counteract market feelings as existing at the time, that the euro was a one-way bet (Koen et al. 2001). However, many academics doubt whether sterilized interventions at foreign exchange markets by central banks are very effective. However, in his testimony to the US Senate Bergsten (2002) argues: There are of course those who doubt the effectiveness of sterilized intervention in the currency markets. Such a view ignores the fact that all three cases of intervention by the Rubin-Summers Treasury worked in textbook fashion. Joint US–Japan intervention stopped and reversed the excessive strengthening of the yen in 1995. Similar intervention stopped and sharply reversed the excessive weakening of the yen in 1998. Joint US–EU intervention in late 2000 stopped the slide of the euro and prompted a 10 percent rebound.

Whether or not official exchange intervention is effective in influencing exchange rates is of crucial policy importance. There is a vast literature on this topic. In this section we review the recent literature on (sterilized) foreign exchange interventions, drawing heavily on Sarno and Taylor (2001) and Frenkel et al. (2001). We define an exchange market intervention as a sale or a purchase of foreign currencies by the monetary authorities with the aim of changing the exchange rate of their own currency vis-à-vis one or more foreign currencies. Coordinated (or concerted) official intervention

The ECB’s Strategy: An Assessment

65

Table 3.7 Estimates of the equilibrium value of the euro Study Wren-Lewis and Driver (1998) Alberola et al. (1999)

Reference period 2000

Methodology

Borowski and Couharde (2000) Van Aarle et al. (2000) Gern et al. (2000)

2000:II 2000:I

Clostermann and Schnatz (2000)

Winter 1999– 2000

Teïletche (2000)

June 2000

Hansen and Roeger (2000)

1999:III

Equilibrium FEER model Internal/external balance model Equilibrium FEER model Monetary model BEER, uncovered interest parity Real long-term yield spread, oil price, government spending, relative price of traded to nontraded goods Relative productivity, government spending, real long-term yield spread, M1, industrial production PEER

Lorenzen and Thygessen (2000)

1999–2000

BEER

IMF (2000b)

Summer 2000

Saving-investment

Schulmeister (2000) Stein (2001) Duval (2001)

Mid 2000 2001:I 2000:III

Alberola et al. (2002)

End 1999

PPP for tradables NATREX Natrex and Balassa Samuelson Internal–external balance model

Alquist and Chinn (2001)

June 2000

End 1998 First half of 1999

Monetary model (M1, GDP, interest differentials) and relative productivity growth

Equilibrium rate (against $) 1.19–1.45 1.26 1.23–1.31 1.07 1.03 Short run 1.20 Medium run 1.13

1.09

Undervation: 15% against trading partners Long run 1.28; medium run 1.19; Short run 1.09 Undervaluation 30% 0.87 1.17 1.15 Undervaluation 12.4% against trading partners Medium run 1.17–1.24

66

Chapter 3

Table 3.7 (continued) Study

Reference period

Methodology

Koen et al. (2001)

Second half 2000

BEER

Maeso-Fernandez et al. (2001) Detken et al. (2002)

2000:IV

BEER/PEER

End 2000

Four different models, including structural VAR and NATREX

Equilibrium rate (against $) Undervaluation 9% Undervaluation of 3 to 20% Undervaluation differs widely, ranging from 5 to 27%

Source: Koen et al. (2001), European Economic Advisory Group at CESifo (2002), ECB (2002) and own update. Note: The BEER (behavioral equilibrium exchange rate) approach is not based on a specific structural model but encompasses several variables. Often those fundamentals have medium-run effects but wash out in the longer run. The PEER (permanent equilibrium exchange rate) methodology builds on the BEER methodology and decomposes the variables into permanent and transitory components. The FEER (fundamental equilibrium exchange rate) methodology places more structure on the computation of the equilibrium exchange rate. The NATREX (natural real exchange rate) approach is founded on a more rigorous modeling of the stock-flow interaction in a macroeconomic growth model. A distinction is made between a medium-term equilibrium, where external and internal balance prevails (equivalent to the FEER approach), and the long-run equilibrium, where net foreign debt is constant and the capital stock is at its steady state level. Internal balance is obtained when a country has full employment and low inflation, while external balance corresponds to a sustainable current account position.

4 Impact of News on the Euro–Dollar Exchange Rate A number of recent papers have examined the impact of so-called news on the euro–dollar exchange rate. These studies have the assumption that foreign exchange markets are efficient in common, meaning that all information relevant for the pricing is instantaneously reflected in the exchange rate. The exchange rate reacts to new information that may be relevant for exchange rate pricing, just as news on economic developments or on monetary policy. Galati and Ho (2001) investigate how daily news reports on macroeconomic conditions in the United States and in the euro area (e.g., unemployment conditions, industrial production reports, consumer prices) affect the euro–dollar exchange rate. They find that during the first two years of EMU, “bad” macroeconomic news tended to depreciate the euro but “good” macroeconomic news did not similarly always lead to an appreciation of the euro. Often the market seemed to ignore good news

The ECB’s Strategy: An Assessment

from the euro area, in line with the views of De Grauwe (2000). De Grauwe (2000) explained the weakness of the euro since 1999 in terms of how markets form views about the factors driving exchange rates. First, De Grauwe identified some fundamental variables that determine the euro–dollar rate. Unexpected changes in these variables, namely the news component, were assumed to influence the exchange rate. De Grauwe found that in most cases the news on the US fundamentals was less favorable than the news on the European fundamentals. De Grauwe argued that because of the great uncertainty about equilibrium levels of exchange rates, short-run movements tend to be driven by technical and chartist analysis. Sustained movements in one direction or another then lead to a search for fundamentals that explain these developments (framing). When the exchange rate changes, a search starts for fundamentals that can explain the observed change. This way a self-enforcing process can evolve: a declining euro is seen as evidence that there are problems in the economy of the euro area. These problems then reinforce the downward movement of the euro. So the causality is reversed: it is not the news about fundamentals that drive the exchange rate but the exchange rate that determines the way the fundamentals are perceived. De Grauwe saw the results of Corsetti and Pesenti (1999) and Corsetti (2000) as additional proof for his theory. Since there was great optimism about the future growth of the US economy, financial market analysts focused solely on the variable that supported these beliefs: the growth rate. Little attention was paid to other fundamentals. Once the direction of change becomes different, the search for new “fundamentals” that can explain this development begins. Fatum and Hutchison (2002) examined the impact of certain news paper reports. The financial press have reported numerous rumors of intervention, official statements expressing views regarding the value of the euro as well as statements on the usefulness of intervention. These authors investigated whether statements and rumors of this nature reported in the press are systematically related to the euro–dollar exchange rate changes. The reports were separated into four categories: • • • •

Rumors and speculation of euro support intervention Statements by officials in support of the euro Statements by officials not supportive of the euro Firm reports of intervention

In their model for the euro–dollar exchange rate, statements by ECB officials in support of the euro and reports of ECB intervention in support of the euro are not statistically significant. However, official statements denying ECB intervention have worked to depreciate the value of the euro. This effect is persistent over a week at least. Rumors and speculation of intervention in support of the euro are associated with euro appreciation, but the effect is not persistent.

67

68

Chapter 3

Jansen and De Haan (2003) also examined the impact of statements of ECB and other European central banks officials, using reports in Bloomberg. In contrast to Fatum and Hutchison (2002), they took a much broader array of statements into account. Over the period January 1, 1999, to May 17, 2002, they identified 936 news reports on monetary policy or the euro. They concluded that most of these statements had no effect on the exchange rate. The only positive effect stems from statements suggesting that the ECB may intervene at the foreign exchange market. Jansen and De Haan further examined whether there are systematic differences between the effects of statements of different central bankers. Table 3.8 is reproduced from this study. It follows that there is a systematic difference in terms of the impact: statements of Bundesbank officials had some short-run impact in contrast to statements from other central bankers, including ECB President Duisenberg.

Table 3.8 Effect of statements of central bankers on the euro–dollar exchange rate Variables

Coefficient

p-Value

EB+ (1) EB+ (0) EB+ (-1) EB+ (-2) EB0 (1) EB0 (0) EB0 (-1) EB0 (-2) EB- (1)*** EB- (0) EB- (-1) EB- (-2) NCB+ (1) NCB+ (0) NCB+ (-1) NCB+ (-2) NCB0 (1) NCB0 (0) NCB0 (-1) NCB0 (-2) NCB- (1) NCB- (0) NCB- (-1)

-1.84 5.33 -3.79 -2.14 -9.89 -1.88 3.16 -8.17 -26.51 3.20 13.29 -8.85 -1.10 -3.16 2.85 -7.77 1.09 -7.68 12.05 -2.95 -4.47 -9.83 7.25

0.75 0.34 0.54 0.73 0.20 0.82 0.70 0.27 0.01 0.77 0.23 0.45 0.84 0.56 0.60 0.15 0.89 0.28 0.12 0.72 0.59 0.22 0.41

The ECB’s Strategy: An Assessment

69

Table 3.8 (continued) Variables

Coefficient

p-Value

NCB- (-2) BuBa+ (1) BuBa+ (0) BuBa+ (-1)** BuBa+ (-2)*** BuBa0 (1)** BuBa0 (0) BuBa0 (-1)* BuBa0 (-2) BuBa- (1) BuB- (0)** BuBa- (-1) BuBa- (-2) R2

-5.19 -9.86 9.79 -23.03 -30.18 -20.73 6.19 -18.16 9.06 -29.37 31.23 17.79 18.42

0.48 0.37 0.27 0.04 0.01 0.05 0.56 0.08 0.37 0.16 0.03 0.17 0.24

0.06 0.00 2.00

Adjusted R2 DW Source: Jansen and De Haan (2003). Note: The estimated model is 5

Rt = b 0 +

K

2

Â b R + ÂÂ b i

i =1

t -i

ki

Dk ,t -i + b id D(itUS - itEU )

k =1 i =-1

+ b m Dm + b tu Dtu + b w Dw + b th Dth + e t , e t ~ (0, ht ), ht = d 0 + d 1e t2-1 + d 2 ht -1 + d m Dm + d tu Dtu + d w Dw + d th Dth . Here Rt represents the change in the natural logarithm of the exchange rate, Dk represents dummies for the ECB statements (EB denotes Executive Board, NCB is national central banker, and BuBa is Bundesbank). The second equation also has dummies for the weekdays (Dm . . . Dth) and the change in the interest differential between the United States and the euro zone (iUS - iEU) as control variables. The error term has a zero mean and an conditional variance ht. A GARCH(1,1) specification is used. The weekday dummies also enter the variance equation. Only the results for the statement dummies are shown. A plus means a positive statement, a zero indicates a neutral statement, while a minus indicates a negative statement. The statement dummies enter the equation with one lead, the current value and two lags. All coefficients are multiplied with a factor 10,000. */**/*** denotes significance at the 10/5/1% level.

70

Chapter 3

occurs when two or more central banks intervene simultaneously in the market in support of the currency, according to an explicit or implicit agreement. The purchase (sale) of foreign currency by the central bank leads to an increase (decrease) in the net foreign assets at the balance sheet of the monetary authorities, and an equivalent increase (decrease) in the monetary base (M). So interventions at the foreign exchange market have basically the same effect on M as an open market operation. There is no debate that this will influence exchange rates. The debate is on the question of whether sterilized interventions will have an effect. If intervention is sterilized, the effects of the intervention on M are offset one for one by the central bank. Normally this is done through sales or purchases of domestic-currency bills or bonds, in other words, through an offsetting open market operation. By far the larger part of exchange market intervention is carried out in the spot market. The reason for this seems to be that an intervention operation derives a great deal of its effect from the announcement of the operation itself. Highly visible spot market operations confirm the announcement. Sterilized purchases and sales of foreign exchange by a central bank can, in theory, have an impact on the exchange rate. Sarno and Taylor (2001) distinguish the portfolio balance channel and the signaling channel. Central in the portfolio balance channel is that after the intervention the composition of the agents’ portfolios has changed, since the central banks will have bought or sold domestic assets in their sterilization operations. When agents try to rebalance their portfolios by buying or selling foreign assets, the spot exchange rate will shift. A crucial assumption in the portfolio balance channel is that agents regard domestic and foreign assets as imperfect substitutes, meaning that there is a risk premium between foreign and domestic assets. If they were perfect substitutes, agents do not have any incentive to rebalance their portfolio. Even if domestic and foreign assets are perfect substitutes, sterilized intervention may have impact on exchange rates through the signaling or expectations channel. If the foreign exchange markets are efficient, all information relevant for the pricing is instantaneously reflected in the exchange rate. So why can an intervention effect exchange rates? The basic idea of this channel is that the central bank may have an informational advantage over private agents. The central bank may have access to information that is not (or only after a certain time lag) available to market participants. It is assumed that the monetary authori-

The ECB’s Strategy: An Assessment

71

ties have superior information to other market participants and that they are willing to reveal this information through their actions in the foreign exchange market. This information may refer to fundamentals (including future policy). So agents may process the new information and adjust their expectations regarding the future evolution of fundamentals. For instance, they may view the intervention as a signal about the future stance of monetary policy. Another channel through which sterilized central bank interventions may affect exchange rate has been discussed in the strand of the literature concerned with the noise trader model of exchange rate determination (Frenkel et al. 2001). Noise traders are financial market participants whose demand for a financial security is not influenced by economic fundamentals alone. Instead, they employ techniques provided by, for example, technical chart analysis to extract buy and sell signals from historical exchange rate trajectories. In noise trader models of exchange rate determination it is assumed that at least in the short and medium run, the group of noise traders influences the dynamics of the exchange rate so that the price of this asset can depart from its fundamental value. If the central bank succeeds in affecting the exchange rate and noise traders respond to these changes, the price impact of noise traders’ orders may then account for persistent exchange rate effects of sterilized foreign exchange market interventions.21 In general, empirical research on the effectiveness of exchange market interventions was for a long time hampered by a lack of sufficiently detailed data. For example, the German and Swiss central banks swapped foreign exchange with commercial banks. As these swaps are included in the changes in official reserves, the latter does not accurately reflect official interventions. Furthermore, estimating the portfolio balance channel requires proxies for the risk premium between domestic and foreign assets. According to Sarno and Taylor (2001), the portfolio balance channel has not attracted a large empirical literature relative to other models of exchange rate. Two types of test have been conducted. The first type is based on estimating a reduced form solution of the portfolio balance model (PBM) in order to measure its explanatory power. The second type focuses on solving the PBM for the risk premium and testing for perfect substitutability of bonds, where the risk premium is measured by deviations from uncovered interest parity. The empirical studies on testing the PBM published in the late 1970s and in the 1980s suggested that sterilized intervention is effective at most in the very short term. Furthermore the effects of interventions through the portfolio balance

72

Chapter 3

channel are very small in size (Sarno and Taylor 2001). Data on official exchange rate intervention and survey data on exchange rate expectations became available for utilization during the late 1980s. Especially the work by Dominguez and Dominguez and Frankel has led to the conclusion that interventions may have some impact through the portfolio balance channel and the signaling channel. Below we summarize this literature, while sidebar 5 discusses more recent evidence based on the so-called event study approach.22 Dominguez and Frankel (1993b) estimate a portfolio-balance equation using survey data on US dollar–German mark and US dollar–Swiss franc exchange rate expectations in the mid-1980s to construct measures of the risk premium as the deviation from uncovered interest rate parity. They find that intervention variables have statistically significant explanatory power in a regression for the risk premium. There is strong support in favor of a significant portfolio balance effect.

5 Are Interventions Effective After All? Some Recent Evidence Exchange rates are highly volatile on a daily basis and interventions come generally in sporadic clusters. Traditional time series modeling may therefore not be the most appropriate tool for analysis. Fatum and Hutchison (2003) therefore apply the event study approach used in the finance literature to data on Bundesbank and Fed interventions over the period September 1985 (Plaza Agreement) to December 1995. The starting point for an event study is to define the event and to identify the period over which the security price is examined. Interventions often comprise more than one day. Therefore Fatum and Hutchison (2003) define the event as a period of days with official intervention in one direction conducted by either the Bundesbank, the Fed, or both, possibly with a number of days (with a maximum of 2, 5, 10, or 15 days) with no intervention. Fatum and Hutchison employ three criteria to judge whether an intervention has been successful. First, is the direction of change the same as the direction in which the central bank was intervening (direction criterion)? Second, is the intervention associated with a smoothing exchange rate movement (smoothing criterion)? Finally, has the intervention led to a change of the exchange rate movement (reversal criterion)? The null hypothesis of no link between the intervention event and the subsequent short run exchange rate movements is rejected for all the criteria for success. The authors conclude therefore that interventions

The ECB’s Strategy: An Assessment

appear to influence exchange rates in the short run, although the methodology does not allow to identify the channel though which these interventions have an effect. Frenkel et al. (2001) also apply the “event study” approach, using tenminute exchange rate returns for the dollar–euro spot exchange rate on the days on which the ECB intervened in the foreign exchange markets. These authors find that the interventions of the ECB resulted in an immediate and statistically significant appreciation of the euro vis-à-vis the US dollar. This effect persisted to a certain extent on the intervention day. However, in all cases of foreign exchange market intervention, the euro depreciated on the day following the intervention suggesting that the effect of the interventions of the ECB was not persistent over a longer horizon. Fatum and Hutchison (2002) apply the event study methodology to the firm newspaper reports of euro intervention that they identified (see sidebar 4). Two separate events emerge: a one-day event on September 22, 2000, and a three-day event spanning from November 3 through 9, 2000. The authors examine windows of 2, 5, 10, and 15 days. The first event appears successful according to all three criteria outlined above when the analysis is based on the shorter window lengths. However, when the analysis is based on the longer window lengths, the September 22 event appears unsuccessful with respect to the “direction” and “reversal” criteria. The November 3 through 9 events appears successful when the analysis is based on the shorter window lengths and the window of 15 days. When the analysis is based on the 10-day window, the event appears unsuccessful. Fatum and Hutchison (2002) conclude that their evidence provides some support for the short-run efficacy of ECB intervention operations, but the effects are not long lasting. They also find that “negative” statements by ECB officials questioning the efficacy of intervention, denying past intervention, and ruling out future intervention have systematically depreciated the euro and the effect appears persistent. Finally, Payne and Vitale (2003) study the effects of sterilized intervention operations of the Swiss National Bank (SNB). They use a transaction-based data set of the activities of the SNB between 1986 and 1995 and combine this information with exchange rate quotes documented in news-wire reports. Payne and Vitale use the event study approach and quantify the effects of single intervention operations on the US dollar–Swiss franc rate at a 15 minutes sampling frequency. They find that even after accounting for the interventions of other central banks, the effects of the SNB interventions on the exchange rate are strong and persistent. The effectiveness of coordinated interventions is higher, while the impact of interventions is larger when the interventions follow the current trend rather than opposing it.

73

74

Chapter 3

Dominguez (1990) investigates whether ex post one-day, thirty-day, and ninety-day excess returns in the US dollar–German mark and US dollar–yen market are related to unilateral and coordinated intervention by the Bundesbank, the Federal Reserve System, and the Bank of Japan. She finds mixed results for the various subperiods distinguished. Also “coordinated intervention had a significantly different and longer term influence on market expectations than did unilateral intervention over the three-year period examined” (p. 158). Dominguez and Frankel (1993a) estimate a two-equation system where one of the equations defines the expectations formation mechanism and the other equation is an inverted portfolio balance equation that allows for mean-variance optimization. The authors report strong statistical evidence supportive of the effectiveness of sterilized intervention through both the portfolio balance channel and the signaling channel. They conclude that “intervention can be effective, especially if it is publicly announced and concerted. It may be that sterilized intervention can only have effects in the short term. But if ‘short-term effects’ include the bursting of a nine-month bubble earlier than it would otherwise have burst, then such an effect may be all that is needed” (p. 140). Dominguez (2003) investigates the influence of Fed intervention operations on both dollar–Dmark and dollar–yen intra-day returns and volatility. She finds evidence supporting the hypothesis that central bank interventions influence intra-daily foreign exchange returns and volatility. In order to get the largest effects the central bank should time intervene when the trading volume is high. Furthermore, the interventions should occur in the aftermath of the release of other macroeconomic news and when other central banks are also intervening in the market (coordinated interventions). 3.7 Asset Price Inflation Several reasons exist why central banks may worry about asset price inflation. First, assets are claims on future services, so asset prices are a proxy for the prices of future consumption. An effective measure of inflation should therefore include asset prices.23 As Goodhart (2001, p. 3) puts it: My dictionary defines inflation as a fall in the value of money, not as a rise in the consumer price index. If I spend my money now on obtaining a claim on future housing services by buying a house, or on future dividends by buying

The ECB’s Strategy: An Assessment

75

an equity, and the price of that claim on housing or on dividends goes up, why is that not just as much inflation as when the price of current goods and services rises?

However, Gilchrist and Leahy (2002) point out that asset prices change for many reasons and not all of them are related to the cost of future consumption. Asset prices may rise when expected profits rise while interest rates may remain unchanged. This implies that changes in asset prices reflect changes in the quantity of future consumption rather than changes in the price of future consumption (Camba-Mendez 2003). Second, if rising asset prices spill over into excess demand, asset prices can be leading indicators of CPI inflation. Consumption will rise through an increase in wealth and rising share prices reduce the cost of capital, so firms invest more. Third, asset price bubbles will distort price signals and cause a misallocation of resources. Firms will, for example, overinvest in risky projects if the cost of capital is artificially low. Finally, a bursting bubble may cause severe economic and financial harm. The earlier the bubble is pricked, the less pain it may cause. The first point is an argument against the definition of inflation as currently used by the ECB, and many other central banks. The other arguments are basically a plea to take asset prices into account in the reaction function for the monetary authorities, either directly (as an objective) or indirectly (as an indicator for other objectives, e.g., inflation). As explained in section 2.3, the ECB uses asset price developments in its broadly based assessment of the risks for price stability. However, the ECB does not explicitly target asset prices. While identifying financial imbalances ex ante can be difficult, Borio and Lowe (2002) suggest that it is not impossible. Sustained rapid credit growth combined with large increases in asset prices appears to increase the probability of an episode of financial instability. Low and stable inflation promotes financial stability and it also increases the likelihood that excess demand pressures show up first in credit aggregates and asset prices rather than in goods and services prices. A monetary response to credit and asset market developments may be useful to preserve both monetary and financial stability. Likewise Cecchetti et al. (2000) argue that if the authorities target asset prices explicitly, it is possible to stop bubbles from getting too far out of hand, thereby avoiding “boom–bust” cycles. The main argument against this view is that it is impossible to distinguish ex ante between fundamental causes that raise asset prices and the existence of a bubble in asset prices. Central banks are able to make better judgments only if they have

76

Chapter 3

superior information in comparison to the private sector. As Mishkin (2001, pp. 15–16) puts it: Without an informational advantage, the central bank is as likely to mis-predict the presence of a bubble as the private market and thus will frequently be mistaken, thus frequently pursuing the wrong monetary policy.

Furthermore an objective for asset prices assumes that central banks are able to influence asset prices. However, most empirical evidence suggests that the link between stock prices and monetary policy is weak at best. Most fluctuations in stock prices occur for reasons unrelated to monetary policy. Therefore Bernanke and Gertler (2000) argue that a central bank dedicated to price stability should pay no attention to asset prices per se, except insofar as they are signals of changes in expected inflation. According to these authors it is problematic to stabilize asset prices because it is nearly impossible to know for sure whether a given change in asset prices results from fundamental factors, nonfundamental factors, or both. What about real estate prices? Boom–bust episodes seem to be more frequent in real property prices than in stock prices (Bordo and Jeanne 2002). In most countries changes in house prices seem to have a bigger impact on consumer spending than do equity prices. Investing in property may lead to a capital gain and also yield rental income. Buying a house is particularly attractive because of the tax treatment of housing. A housing bubble is associated with more debt than a stock market bubble. This makes a housing bubble more dangerous. When the house prices fall sharply, this can harm the economy more than a similar fall in share prices. Still Vickers (1999) argues that house prices should not be an independent concern of monetary policy. The prices matter to the extent that they contain information about inflation prospects. The relationship between house prices and inflation are complex and imperfectly understood, according to Vickers (1999). He argues that the effects of higher house prices on consumer spending are far from unambiguous. The foregoing analysis suggests that a central bank should not apply an objective for asset and house prices. But what about a broader definition of price stability (see sidebar 6)? The inclusion of equity prices in the definition of price stability has actually some other major drawbacks of an empirical nature (Camba-Mendez 2003). First, the large volatility and/or potential for bubbles in equity prices can add too much noise to a measure of inflation, thereby making it more difficult

The ECB’s Strategy: An Assessment

77

to extract the true signal from price developments. Second, it is very risky for a central bank to try to smooth large changes in asset prices. If, for example, a central bank tries to correct for sharp falls in asset prices, this can ultimately lead to larger asset price bubbles as investors may count on the central bank to come to their rescue. Third, the empirical evidence supports a link between real estate prices and output, but this link is much weaker for equity prices (see Cecchetti et al. 2000; Goodhart and Hofmann 2000). Goodhart (2001) was led to suggest that a measure of inflation should accord some weight to housing prices but not to equity prices.

6 Implications of a Broader Definition of Price Stability Arnold et al. (2003) examine a broader definition of price stability for ECB’s policies by way of the euromon model. In their definition of price stability, they use various weights for asset prices, following Bryan et al. (2001), to observe the effects of the inclusion of stock prices and housing prices. Their counterfactual simulation starts in 1994 and ends in 2001 for the case where the objective of the ECB (inflation not exceeding 2 percent) remains in place. The standard reaction function assumed in the euromon model is based on the Taylor rule. The broad inflation rate p(t) is defined as: p (t) = j cpp cp (t) + j hp h (t) + j eqp eq (t )

with j cp + j h + j eq = 1,

where pcp is the consumer price inflation, ph is the price increase of houses, and peq is the rate of increase of equity prices. The weights for housing prices are 10 and 20 percent and for equity prices the weights are 0.5, 2, and 4 percent. Table 3.9 shows the simulation results. The table shows the mean (M) and the standard deviation (SD) of the annual rate of growth of real GDP ( y), consumer price inflation, the policy rate (RS) housing price inflation, and equity price inflation for the period 1994 to 2001. Both the actual (realized) and the simulated outcomes according to euromon are shown for a broad definition of price stability by the ECB. Columns 2 and 3 of the table show that the inclusion of housing prices implies a higher interest rate. Consequently output growth is lower if the ECB employs a broader definition of price stability, whereas CPI inflation is higher. If equity prices are included in the definition of price stability, interest rates are, on average, only slightly higher, and output growth slightly lower (columns 4–7 of table 3.9). If both housing and equity prices are included (column 8), interest rates slightly rise and output falls compared to the base scenario. Nevertheless, the analysis suggests that, all in all, the effects of a broader definition of price stability would be small for the period under consideration.

78

Chapter 3

Table 3.9 Actual and simulated macroeconomic developments in EMU, 1994–2001 Realized

(1) My SD y M pcp SD pcp M RS SD RS M ph SD ph M peq SD peq

Counterfactual with broad definition of inflation jcp = 0.90 jcp = 0.80 jcp = 0.995 jcp = 0.98 jcp = 0.96 jh = 0.10 jh = 0.20 jh = 0.00 jh = 0.00 jh = 0.00 jeq = 0.00 jeq = 0.00 jeq = 0.005 jeq = 0.02 jeq = 0.04 (2) (3) (4) (5) (6)

jcp = 0.96 jh = 0.00 jeq = 0.04 (7)

jcp = 0.88 jh = 0.10 jeq = 0.02 (8)

2.22 0.84

2.14 0.81

2.06 0.81

2.20 0.81

2.12 0.75

2.02 0.69

2.01 0.79

2.06 0.73

2.23 0.74 4.68 1.24 3.07 2.05 11.84 21.05

2.32 0.68 4.84 0.93 3.09 1.99 10.92 20.30

2.41 0.63 4.98 0.74 3.10 1.95 10.12 19.86

2.21 0.74 4.69 1.24 3.01 2.01 12.26 19.91

2.17 0.74 4.75 1.32 2.84 1.91 13.32 17.00

2.12 0.76 4.88 1.47 2.62 1.77 14.24 14.22

2.18 0.74 5.02 1.10 2.75 1.77 11.63 19.07

2.26 0.68 4.89 1.02 2.88 1.89 12.25 16.72

Source: Arnold et al. (2003).

3.8 Interest Stepping versus Smoothing Should interest rates be changed gradually or not? So far the ECB has changed interest rates either by 50 basis points (7 times) or 25 basis points (8 times). Arguments have been put forward for both bigger (interest rate stepping) and smaller steps (interest rate smoothing). Interest rate stepping can be defined as the tendency of interest rates to remain fixed for a while in an environment that is being continually hit by shocks to the determinants of future inflation and growth. As a result the interest rate is shifted in discrete jumps while the determinants evolve smoothly (i.e., continuously) over time. This practice should be clearly distinguished from the practice of interest rate smoothing, which denotes a situation where a required interest rate change is implemented by a series of small steps in the same direction rather than by a single step at once. There are several rationales for not changing the interest rate continuously. According to Eijffinger and Huizinga (1999), central banks can maintain a lower rate of inflation by a commitment not to change the interest rate in every period. Since the costs of a sudden inflation are permanent—and will induce an increase in inflationary expectations— while the benefits are temporary, such a commitment will induce the

The ECB’s Strategy: An Assessment

79

central bank to be more aware of the costs of an expansionary policy. The result will be an increase in the central bank’s credibility for fighting inflation. Hence especially for the ECB, which does not have a historical record for fighting inflation, a commitment to interest rate stepping is likely to be beneficial. Another good reason for stepping is provided by Eijffinger et al. (1999) and is based on the fact that interest rate changes are costly to the central bank. The cost is loss of reputation in the financial markets. Too many steps—and especially too many reversals—lead to a perception of a central bank’s incompetence. Too many steps can even impede the signaling value of a change in monetary policy, especially in an environment where there is a lot of uncertainty. Because monetary policy takes place in a dynamic framework, the central bank has to take into account that it has the option to wait and see whether the economy will move back toward the inflation target of its own accord. In particular, the higher the uncertainty, the more the central bank should not attempt to change the interest rate. A third reason, due to Goodfriend (1991), argues that by implementing a policy of interest rate stepping central banks can obtain more leverage over longer term interest rates. This is beneficial because these rates are generally agreed to affect aggregate demand. There are also some good arguments for smoothing (Goodhart 1996). The primary one concerns the presence of multiplier uncertainty (Brainard 1967). If the central bank is uncertain about the impact of a change in the interest rate on output and inflation, and if it cares about the variances of inflation and output around, respectively, the inflation target and the natural rate, there will be a tendency for caution. The reason is that large changes in the monetary policy stance (with the aim setting the conditional mean rate of inflation equal to the target) will increase the expected variability of output and inflation. A related argument is that if central banks implement small changes in the presence of uncertainty, then they can assess the effect of each change, and subsequently whether or not a further change is needed. However, this can be dangerous because there are long and variable lags in monetary policy. So the central bank will risk falling behind the curve (i.e., doing too little too late). Empirical evidence by Eijffinger and Huizinga (1999) suggests that a relative activist monetary policy does not result in higher output growth but rather in more costs because of a higher average and

80

Chapter 3

variability of inflation in the long run. The experience with the ECB’s interest rate policy balancing bigger steps (50 basis points) and smaller changes (25 basis points) indicates that the ECB is still undecided on whether to go with interest rate stepping or smoothing. 3.9 Conclusions What do we make out of all this? In our view the ECB has been quite successful in its monetary policy so far. Inflation in the euro area has remained low, although it has exceeded 2 percent almost continuously since mid-2000, which is for most of the ECB’s history. Still this has not led to increases in expected inflation. Furthermore the ECB cannot reasonably be said to have been overly aggressive in pursuing its primary objective to the detriment of other considerations. So critics who argue in favor of an inflation objective above 2 percent should be quite content with actual ECB policies. In addition most of the policy decisions have been in line with market expectations, although there were some surprises and sometimes the market expected a different timing of the steps taken by the ECB. Despite this favorable assessment, we believe that the ECB can improve the transparency of its monetary policy strategy. First, it should drop altogether the first pillar and concentrate, exclusively, on a more explicit inflation targeting strategy. Second, the ECB should provide a midpoint inflation target that includes a symmetric band or range around it. On the basis of the various arguments put forward in this chapter, we favor a midpoint inflation target of 2 percent with a range of 1.5 to 2.5 percent. As to the role of the first pillar, most of the academic literature favors less emphasis on the role of money in the monetary strategy of the ECB and a (gradual) move toward inflation targeting. The main problem with the role of money in the strategy of the ECB is that it is widely misunderstood by many observers and therefore creates the impression that the ECB lacks transparency (see also chapter 4). We therefore conclude that the ECB should abandon the monetary analysis as a dominant navigation system for its monetary policy. The abolishment of the first pillar would avoid confusion and increase transparency. However, it is unrealistic to think that focusing on the present second pillar will do away entirely with the problem. Just keeping the second pillar in place will lead to similar problems, as it contains many indicators that are unlikely to point in the same

The ECB’s Strategy: An Assessment

81

direction at the same time. Therefore an inflation-targeting strategy is to be preferred. As to the inflation objective, in our view the arguments put forward for raising the inflation objective substantially are not compelling. Our reading of the literature leads us to the conclusion that the consequences of the zero bound are negligible for target inflation rates as low as 2 percent. Likewise the upward measurement bias in inflation in the euro area seems to be less of a problem than in the United States. Also the risk of deflation in an individual country in the euro area seems to be small. The evidence on the Balassa-Samuelson effect is so mixed that we do not consider this to be a convincing argument for a higher ECB inflation objective (see also chapter 5). The ECB cannot target the external and the internal value of the euro at the same time. Because the ECB’s mandate is for price stability, the internal value of the euro takes priority. However, this does not imply that the euro–dollar exchange rate is not important as the euro-dollar exchange rate affects inflation. The ECB has occasionally also expressed concern that the exchange rate may not be in line with “fundamentals.” Whether interventions are needed to bring the exchange rate in line with these fundamentals is doubtful. Although there is some evidence of short-run effectiveness of sterilized foreign exchange intervention, there is up until now too little evidence of its persistence over a longer horizon. Some academics have argued that central banks should either broaden their definition of inflation by including asset prices or directly target asset prices. The ECB rightly takes the information that asset prices may contain with respect to future inflation into account under its second pillar. To take asset prices into account in defining inflation is a step we would not endorse, since asset prices change for many reasons and not all of them are related to the cost of future consumption. As to the wealth effects and the pass-through of asset and property inflation to goods and services inflation, there is at present no conclusive empirical evidence of their relevance. Furthermore the inclusion of asset prices into the definition of price stability would add too much noise to a measure of inflation because of the large volatility and/or potential bubbles in equity prices. This makes it more difficult to extract the true signal from price developments.

4

Transparency, Accountability, and Credibility of the ECB

4.1 Introduction For a long time central banks have been associated with secrecy. Recently, however, various central banks, including the Bank of England and the Reserve Bank of New Zealand, have embraced openness. A comprehensive survey of 94 central banks by Fry et al. (2000) reveals that 74 percent of the respondents consider transparency a vital or important component of their monetary policy framework. There exist differences in the literature as to the definition of transparency. Basically two kind of definitions can be distinguished. Sometimes transparency refers to the activities of the central bank in providing information. In other cases transparency relates to the public’s understanding of monetary policy. When a central bank is open about its policies, the general public can get a good understanding of the decisions taken by the monetary authorities and the reasoning behind them. In line with this Winkler (2000) defines transparency as “the degree of genuine understanding of the monetary policy process and policy decisions by the public.” Promoting public understanding involves monetary authorities being active in developing and conveying information, using the various means available to disclose information, and tailoring the message according to the needs of particular audiences. We will use the term disclosure whenever we refer to the activities of the central bank to enhance the understanding of the public of its policies (see also Siklos 2002).1 The concept of transparency as defined by Winkler (2000) and the activities of the central bank to promote this understanding (disclosure) are related. In principle, it is possible that the public does have a good understanding of the policies of the central bank (or claims that it has),

84

Chapter 4

although the central bank hardly provides information on the decisions taken and the reasoning behind them. Alternatively, the central bank may be very active in providing information, while the public feels that it does not fully understand what the central bank does. Whether the activities of the central bank lead to a better understanding depends on many factors, some of which are discussed in the present chapter. One of the crucial factors, of course, is the quality of the information provided. No matter how often a central bank publishes information, if the receivers of it feel that this information is hard to understand, the central bank may not be transparent. In this chapter we first discuss issues of disclosure and transparency of central banks. In the next section we start by reviewing the literature on disclosure and transparency distinguishing between various types of disclosure. In section 4.3 we discuss recently developed indicators of central bank disclosure. We also discuss in this section the transparency of the ECB as perceived by financial markets. According to Issing (2001a), transparency of monetary policy making is thought to bring various benefits. First, democratic accountability of central banks requires disclosure. Whatever other arrangements concerning democratic accountability may exist, their scope is limited without proper information concerning the behavior of central banks as this is crucial for the evaluation of its performance.2 In section 4.4 we consider the accountability of the ECB. Second, transparency of monetary policy making may enhance the effectiveness and credibility of monetary policy. By providing the public with adequate information about its activities, the central bank can establish a mechanism for strengthening its credibility by matching its actions to its public statements (IMF 2000a). Although economists often base their models on the notion of credibility, little is actually known about what that means. For instance, how does a central bank become credible? Should a central bank be independent to be credible? Blinder (2000) held a survey among central bankers on the credibility of central banks (section 4.5). Blinder’s respondents considered credibility important “to keep inflation low.” The best way for a central bank to earn credibility is to “have a history of doing what it says it will do.” We have done a similar survey among professional economists from a broad sample of countries (section 4.6). Because we ask the same questions as Blinder did, we are able to examine whether private sector economists share the views of central

Transparency, Accountability, and Credibility of the ECB

85

bankers on these matters. In this section we also examine the credibility of the ECB. The final section offers some concluding comments. 4.2 Transparency and Disclosure: Theory and Evidence 4.2.1 Various Facets of Disclosure To understand the role that transparency plays in monetary policy, it is important to examine the interrelationship between the disclosure of a central bank and transparency of monetary policy. Figure 4.1 is a stylized account of how information provided by the central bank (disclosure) will affect the public’s understanding of the monetary policy process (transparency). The top panel depicts the central bank’s internal activity of information processing, analysis, and decision making. On the basis of the monetary policy strategy, the central bank reaches a decision about monetary policy that is communicated to the outside world. In this process the monetary policy strategy is, again, crucial as the central bank will use it in explaining the reasons for reaching a certain decision. The public receive most of the information through print and broadcast media.3 The effect of monetary policy on inflation and output growth (outcome) is determined not only by monetary policy decisions but also by the expectations and the behavior of the

Objectives Ø Input (raw data) Æ

Central bank strategy

Æ

Output (decisions)

Ø Communication about strategy (disclosure) (activities, contents) Ø

Outcomes

Reporting about strategy by intermediaries Ø Public’s understanding (transparency) Æ Behavior Figure 4.1 Monetary policy strategy and communication. (Source: Adapted from Winkler 2000)

86

Chapter 4

public, based on their understanding of the bank’s strategy (transparency). The outcome arguably influences the raw data that in turn influence the bank’s decisions. Disclosure and transparency are multifaceted concepts that could pertain to any aspect of economic policy making.4 Thus it seems natural to use a conceptual framework that reflects the different stages of the decision-making process. Following Geraats (2000, 2002), we distinguish five aspects of disclosure: political, economic, procedural, policy, and operational disclosure. Each may give rise to different motives for disclosure. Political disclosure refers to openness about policy objectives. This comprises a statement of the formal objectives of monetary policy, including an explicit prioritization in case of potentially conflicting goals, and quantitative targets. Political disclosure is enhanced by institutional arrangements, like central bank independence and central bank contracts, because they ensure that there is no undue influence or political pressure to deviate from stated objectives. Economic disclosure focuses on the economic information that is used for monetary policy. This includes the economic data the central bank uses, the policy models it employs to construct economic forecasts or evaluate the impact of its decisions, and the internal forecasts the central bank relies on. The latter are particularly important since monetary policy actions are known to take effect only after substantial lags. So the central bank’s actions are likely to reflect anticipated developments. Procedural disclosure is about the way monetary policy decisions are taken. It involves an explicit monetary policy rule or strategy that describes the monetary policy framework, and an account of the actual policy deliberations and how the policy decision was reached, which may be achieved by, for instance, the release of minutes and voting records. Policy disclosure means a prompt announcement of policy decisions. In addition it includes an explanation of the decision and a policy inclination or indication of likely future policy actions. The latter is relevant because monetary policy actions are typically made in discrete steps; a central bank may be inclined to change the policy instrument but decide to wait until further evidence warrants taking a full step. Operational disclosure concerns the implementation of the central bank’s policy actions. It involves a discussion of control errors in achieving the operating targets of monetary policy and (unanticipated)

Transparency, Accountability, and Credibility of the ECB

87

macroeconomic disturbances that affect the transmission of monetary policy. The index for central bank disclosure discussed in section 4.3 attempts to quantify each of these five aspects. 4.2.2 Disclosure in Theory Although there seems to be an unambiguous trend toward greater openness in monetary policy, the theoretical literature on the desirability of central bank disclosure and transparency is less equivocal.5 Since the motives and consequences of disclosure may differ by aspect, we will use the conceptual framework described above to provide an overview of the theoretical findings. In addition there are some interesting informal discussions on the desirability of central bank openness, for instance, by Buiter (1999), Issing (1999b), Goodfriend (1986), and Winkler (2000). Buiter (1999) is an advocate of the publication of inflation forecasts because this makes it possible to evaluate the quality of monetary policy and it increases accountability. In addition Buiter favors the publication of individual voting records because it would increase accountability. However, Issing (1999b) opposes the publication of inflation forecasts and individual voting records. In his view, the publication of inflation forecasts could be misleading because the public would attach more value to them than they have in the decision-making process. Issing poses that the publication of individual voting records is not needed to communicate relevant information to the public. Goodfriend (1986) argues that the central bank should avoid revealing too much information to the markets that would induce extra volatility. Winkler (2000) poses that greater openness (more information) does not always increase the clarity of the central bank in communication. Political Disclosure Any uncertainty about policy makers’ preferences is probably reduced if the central bank provides information about formal objectives and quantitative targets. This could be beneficial. Eijffinger and Geraats (2002) argue, for instance, that a quantitative target, which reduces the preference uncertainty of the central bank, could reduce the inflation bias. Likewise Nolan and Schaling (1996) show that in a static monetary policy game a reduction in uncertainty about the central bank’s preference parameter for inflation stabilization reduces the inflation bias; see also Cukierman (2001b).

88

Chapter 4

According to Eijffinger et al. (2000), more uncertainty about preferences increases the inflation bias because wage setters will see the central bankers as less conservative. In addition more monetary policy uncertainty increases the variance of the inflation rate. However, Eijffinger et al. (2000) also show that uncertainty about preferences could be beneficial when a country has a large flexibility problem relative to the credibility problem because uncertainty could decrease the variance of output. So a country facing large supply shocks may be better off with a central banker with uncertain preferences. Eijffinger et al. (2000) therefore conclude that optimal central bank secrecy involves a trade-off between the negative and beneficial effect of uncertainty about monetary policy. However, Beetsma and Jensen (2003) show that it may not be true that uncertainty about preferences is beneficial, even when it is assumed that a country has a relatively large flexibility problem and faces large supply shocks. Other arrangements that do not involve uncertainty about the preferences of the central bankers (e.g., an inflation contract or an inflation target) dominate policy with preference uncertainty. In addition Beetsma and Jensen (2003) argue that the way in which Eijffinger et al. (2000) modeled preference uncertainty causes arbitrary effects on monetary policy. Using their earlier preference specification (Beetsma and Jensen 1998), Beetsma and Jensen (2003) show that preference uncertainty is always harmful. Quantitative targets could generate losses to monetary policy makers when these targets are missed. Walsh (1999) shows that an explicit inflation target could reduce the inflation bias for this reason. Interestingly, this even holds if the target is announced by the central bank and is not perfectly credible. In fact Walsh (1999) finds that such selfannounced targets may be preferable to an external, fixed target because it gives the central bank greater flexibility to respond to private information it has about economic disturbances. Thornton (2002) agrees with Freedman (2002), who states that being transparent and explicit about the inflation objective and being credibly committed to it is a defense against the economic and political forces that could create an inflation bias. Hughes Hallett and Viegi (2001) argue that a lack of central bank transparency could lead to a lower mean inflation rate as a consequence of lower inflation expectations but also less emphasis on output stabilization. To counteract this effect, a less “conservative” government that attaches more weight to output stabilization could be elected. Depending on which effect dom-

Transparency, Accountability, and Credibility of the ECB

89

inates, the total effect of a lack of political central bank transparency on inflation and output is therefore unclear. Institutional arrangements like central bank independence, central bank contracts, and explicit override mechanisms also contribute to political transparency because they clarify the relationship between the government and the central bank. The theoretical motivation for central bank independence refers to the benefits of the appointment of “conservative” central bankers. The seminal paper by Rogoff (1985) shows that central bankers that attach a greater weight to inflation stabilization than socially optimal, reduce the inflation bias of monetary policy, albeit at the cost of greater output fluctuations. The latter side effect could be overcome by the appointment of conservative central bankers with a lower inflation target (Svensson 1997), or by “responsible” central bankers who do not attempt to stimulate output beyond the natural rate (Blinder 1997). Central bank contracts could not only provide quantitative targets but also direct penalties for missing them, like fines or dismissal of the central banker. Walsh (1995) shows that central bank contracts could eliminate the inflation bias without compromising output stabilization. However, when there is uncertainty about the central bank’s preferences, Beetsma and Jensen (1998) and Muscatelli (1998) find that the optimal institutional setting in the form of inflation targets and contracts may involve a trade-off between credibility in the form of a reduction of the inflation bias, and flexibility to stabilize output in response to supply shocks. Such a credibility–flexibility trade-off is also present in the optimal override mechanism derived by Lohmann (1992). Economic Disclosure Most of the literature on economic transparency focuses on the disclosure of economic shocks and/or central bank forecasts. Svensson (1999) and Buiter (1999a) favor the publication of forecasts to enhance transparency and enable an evaluation of monetary policy. When there is (mutual) uncertainty about expectations of the private sector and the central bank, Tarkka and Mayes (1999) argue that the release of central bank forecasts could help and make monetary policy more predictable which would decrease the variability of output. Furthermore Geraats (2000) shows that the publication of central bank forecasts reduces the inflation bias and facilitates reputation building when there exists uncertainty about

90

Chapter 4

the preferences of the central bank. This way the central bank also acquires greater flexibility to stabilize economic shocks. But Geraats finds that inflation forecasts typically do not suffice to reap these benefits. When a central bank uses the interest rate as its policy instrument, central bank forecasts for both inflation and output are needed. In addition she finds that similar benefits can be obtained when the central bank releases the economic model(s) it uses for policy analysis. In cases where there is no preference uncertainty and a Lucas-type transmission mechanism is present, Gersbach (1998) and Cukierman (2001a) show that the premature disclosure of economic disturbances can hamper their stabilization. Jensen (2001) further finds a negative stabilization effect using a new-Keynesian Phillips curve and assuming preference uncertainty. Another reason against economic transparency is that it could lead to greater political pressures when the central bank lacks independence or a clear political mandate (Geraats 2001). Procedural Disclosure The only formal models that analyze procedural disclosure pertain to the release of individual voting records. Gersbach and Hahn (2001b) show that this is beneficial when central bankers’ preferences may differ from the socially optimal objectives. On the other hand, Gersbach and Hahn (2001a) argue that the disclosure of attributed voting records can be harmful when central bankers differ in their degree of competence. To the best of our knowledge, there are no models on the desirability of an explicit monetary policy strategy or the publication of minutes. In defense of the latter, Buiter (1999a) strongly argues in favor of a “culture of openness and accountability” such that “all information is automatically in the public domain, unless there are overriding public interest reasons for not releasing a particular item.” In this regard he promotes the release of nonattributed minutes, since attributed, verbatim transcripts are likely to discourage open discussion during the monetary policy meetings. Still the importance of minutes in enhancing transparency is disputed (e.g., see Issing 1999b). For one thing, minutes do not reveal much because they tend to be very brief and do not aim to make clear how the discussion evolved. The publication of minutes can also undermine free discussion among the monetary policy board members, and where central banks have a clear collective responsibility, the usefulness of making voting behavior public may be

Transparency, Accountability, and Credibility of the ECB

91

limited. Publication can even undermine the credibility of a decision taken by only a slight majority and put pressure on presidents of national central banks. Policy Disclosure Several papers analyze the effects of a prompt announcement of the policy decision, and they all focus on (nonborrowed) reserves targeting. Tabellini (1987) shows that when there is uncertainty about the average reserves target, secrecy about the shortterm reserves target increases volatility of the federal funds rate, which could be detrimental to the achievement of monetary objectives. In contrast, Dotsey (1987) argues that secrecy about the short-term monetary target reduces variability of the federal funds rate when the average money target is perfectly known. In addition Rudin (1988) finds that such policy secrecy could increase the predictability of the federal funds rate when some private sector agents engage in Fed watching. Finally Cosimano and Van Huyck (1993) find that secrecy about policy directives for reserve targets is beneficial when the central bank’s trading desk has an incentive to manipulate reserves to reduce the federal funds rate. The consequences of immediate policy explanations and indications of policy inclination have, as far as we know, not been formally modeled. Operational Disclosure An influential precursor to the transparency literature is Cukierman and Meltzer (1986) on the optimal degree of ambiguity in monetary policy through control errors when the central bank’s preferences are uncertain and change over time. Faust and Svensson (2001) extend their model and distinguish between imperfect monetary control and (operational) transparency. In their model high transparency generally reduces the inflation bias and improves social welfare. It is assumed that the central bank controls inflation imperfectly and that the central bank has an employment target which varies over time according to an idiosyncratic component. By revealing the control error over inflation, the central bank renders its intentions for inflation observable, which results in lower inflation as it increases the sensitivity of a central bank’s reputation to its actions, making it more costly for the central bank to pursue a high inflation policy. In the Faust and Svensson (2001) model there is also a regime (called the extreme transparency regime) in which both the central bank’s employment goal and its inflation intentions are observable. In this situation the

92

Chapter 4

central bank’s reputation is no longer affected by its actions, and an inflationary bias brings inflation to a higher level. Faust and Svensson (2000) argue that minimum transparency is likely to occur in practice when the degree of transparency is a choice variable for the central bank. Jensen (2001) finds that greater operational transparency can be beneficial when central banks suffer from low credibility, but that it limits the ability to stabilize economic disturbances in the case of a newKeynesian Phillips curve. In his model, openness about control errors worsens the current inflation–output trade-off. The central bank attempts to create surprise inflation when the output goal is high. The scope for this is reduced by greater operational transparency because of the increase in forward-looking inflation expectations. The assumption made is that information about these control errors or shocks is disclosed before current inflation expectations are set so that it affects the current inflation–output trade-off. This leaves us with the question of whether central bank disclosure is desirable. Clearly, it depends on which aspect of disclosure is considered. But even then, there is a wide variety of theoretical arguments regarding the economic consequences, depending on the specific assumptions used. Ultimately the answer may be that it depends on the type of disclosure considered, the monetary framework and the structure of the economy. Nevertheless, from the perspective of democratic accountability of monetary policy, which is especially desirable in the case of independent central banks, some degree of disclosure is simply necessary. In this respect, possible economic drawbacks of disclosure could be considered as the price that society may need to pay for accountability. 4.2.3 Effects of Disclosure: Empirical Evidence The issue of the economic consequences of disclosure and transparency has received only scant attention in the empirical literature (see Posen 2003 for a discussion). Many policy makers argue that disclosure (or, as they often call it, transparency) can strengthen the effectiveness of monetary policy making. According to Issing (2001a, p. 13), “transparency imposes discipline on policy makers and is a means to ensure a general understanding of the monetary policy strategy. In turn, this may add to the credibility, and thereby the effectiveness, of monetary policy, hence facilitating the central bank’s effort to attain its statutory objective.”

Transparency, Accountability, and Credibility of the ECB

93

Indeed, Chortareas et al. (2002) find for a sample of 87 countries that their index of disclosure (based on data taken from Fry et al. 2000), which is based on central banks’ published forecasts and ranges from zero to four, is negatively related to average inflation, also if various control variables are taken up. However, using the Eijffinger-Geraats disclosure index for 9 OECD countries, as discussed in the next section, Demertzis and Hughes Hallett (2002) conclude that disclosure does not affect the levels of inflation and output, although it does affect their variability. Another reason that transparency matters is that communication by the central bank about its long-term inflation goal allows the bank to be more flexible in response to shocks in the short run. The trust in the central bank resulting from communication implies that deviations from the target do not indicate a lack of commitment (King 1997). If the central bank builds trust by communicating its long-term inflation objective, inflation persistence will decline since there is a strong belief that inflation will return to a target level. There is evidence in support of this view. Kuttner and Posen (1999) examine the behavior of bond markets (proxying for inflation expectations) in Canada, New Zealand, and the United Kingdom before and after the central banks in these countries adopted inflation targeting. They find that interest rates decreased consistent with the view that the adoption of inflation targeting increases flexibility. Kutner and Posen (2001) report similar results for a broader range of countries: inflation targeting reduces inflation persistence, in contrast to other elements of the monetary framework, like central bank independence. However, Ball and Sheridan (2003), who compare 7 OECD countries that adopted inflation targeting in the early 1990s to 13 that did not, report no supportive evidence. They find that after the early 1990s, performance improved in both the targeting and nontargeting countries. Where targeters improved by more than nontargetters this is explained by the fact that targeters performed worse than nontargeters before the early 1990s, and there is regression to the mean. Once regression to the mean is taken up, there is no evidence that inflation targeting improves performance. Chortareas et al. (2003) investigate the effect of central bank transparency on the costs of disinflation. They assume that disinflation efforts imply a higher sacrifice ratio when the public is not fully convinced that the central bank will reduce inflation. Chortareas et al. estimate sacrifice ratios for 21 OECD countries and use disclosure indexes

94

Chapter 4

relating to the detail with which forecasts are published and the means by which policy decisions are explained. Their results suggest that when the degree of central bank disclosure is higher, the sacrifice ratio is lower. This result is robust to alternative estimation methods and periods considered. A final reason why transparency may matter is that communication may remove noise from markets (Posen 2003). Greater disclosure will lead to greater predictability of central bank actions. The results reported by Kuttner (2001) offer support for this point of view. Changes in the Federal Reserve’s disclosure policy have reduced market volatility and increased predictability. 4.3 Transparency of the ECB Various indicators for central bank disclosure have been suggested in the literature. The indicator of Eijffinger and Geraats (2002) is discussed in some detail in this section. The central bank disclosure index of Eijffinger and Geraats (2002) is the (nonweighted) sum of the scores for the answers to a number of questions discussed below. The minimum score is 0; the maximum score is 15. The index consists of 5 classifications of disclosure, while each classification includes 3 items. Note that all questions pertain to published information that is freely available in English as of June 2001. The index consists of the scores for nine central banks: the Reserve Bank of Australia, the Reserve Bank of Canada, the European Central Bank, the Bank of Japan, the Reserve Bank of New Zealand, the Sveriges Riksbank, the Swiss National Bank, the Bank of England, and the Federal Reserve System. The index is based on the various aspects of disclosure as distinguished in the previous section. Table 4.1 presents the scores for the various central banks. In the table political disclosure refers to openness about policy objectives. This comprises a formal statement of objectives, including an explicit prioritization in case of multiple goals, a quantification of the primary objective(s), and explicit institutional arrangements. It is quantified by the following questions: Is there a formal statement of the objective(s) of monetary policy, with an explicit prioritization in case of multiple objectives?

•

•

Is there a quantification of the primary objective(s)?

Australia

Canada

Euro area

Japan

New Zealand

Sweden

Switzerland

United Kingdom

United States

3 1 1 1

3 1 1 1

3 1 1 1

1.5 0.5 0 1

3 1 1 1

3 1 1 1

2.5 0.5 1 1

3 1 1 1

1 0.5 0 0.5

Economic disclosure Economic data Policy models Central bank forecasts

1 0.5 0 0.5

2.5 1 1 0.5

2.5 1 1 0.5

1.5 1 0 0.5

2.5 0.5 1 1

2 1 0 1

1.5 1 0 0.5

2.5 0.5 1 1

2.5 1 1 0.5

Procedural disclosure Explicit strategy Minutes Voting records

1 1 0 0

1 1 0 0

1 1 0 0

2 0 1 1

3 1 1 1

2 1 1 0

1 1 0 0

3 1 1 1

2 0 1 1

Policy disclosure Prompt announcement Policy explanation Policy inclination

1.5 1 0.5 0

2 1 1 0

1.5 1 0.5 0

1.5 1 0.5 0

3 1 1 1

2 1 1 0

2 1 1 0

1.5 1 0.5 0

3 1 1 1

Operational disclosure Control errors Transmission disturbances Evaluation policy outcome Total

1.5 1 0.5 0 8

2 1 0.5 0.5 10.5

2 1 0.5 0.5 10

1.5 0.5 0.5 0.5 8

2 1 0.5 0.5 13.5

3 1 1 1 12

2.5 1 1 0.5 12.5

1.5 1 0 0.5 10

0.5 0.5 0 0 7.5

95

Political disclosure Formal objectives Quantitative targets Institutional arrangements

Transparency, Accountability, and Credibility of the ECB

Table 4.1 Eijffinger-Geraats index of central bank disclosure

96

Chapter 4

Are there explicit institutional arrangements or contracts between the monetary authorities and the government?

•

As follows from table 4.2, almost all central banks have three points. The exceptions are the Bank of Japan, the Swiss National Bank, and the Federal Reserve. These central banks don’t earn the full score on the first question, since they have multiple objectives of monetary policy without prioritization. Furthermore the Bank of Japan and the Federal Reserve have no quantification of the primary objectives of monetary policy, so they earn no points under the second question. Finally, the Fed does not have explicit instrument independence or a contract and therefore it receives only 0.5 point under the third question. Economic disclosure refers to the economic information used for monetary policy. This includes economic data, the model of the economy that the central bank employs to construct forecasts or evaluate the impact of its decisions, and the internal forecasts (model based or judgmental) that the central bank relies on. It is quantified by the following questions: Is the basic economic data relevant for the conduct of monetary policy publicly available? The focus is on the following five variables: money supply, inflation, GDP, unemployment rate, and capacity utilization.

•

Does the central bank disclose the formal macroeconomic model(s) it uses for policy analysis?6

•

Does the central bank regularly publish its own macroeconomic forecasts?

•

The results for economic disclosure are very mixed. No central bank earns the full score. Five central banks have the highest score, that is 2.5 points (the Bank of Canada, the ECB, the Reserve Bank of New Zealand, the Bank of England, and the Federal Reserve). These central banks loose 0.5 point on either publishing only quarterly time series for three or four out of the five variables requested under the first question above or publishing numerical central bank forecasts for inflation and/or output at less than quarterly frequency. The Reserve Bank of Australia and the Sveriges Riksbank both have the lowest score, which is 1 point. The Reserve Bank of Australia reaches this score since it publishes only quarterly time series for four out of the five variables requested under the first question; therefore it receives only 0.5 point.

Transparency, Accountability, and Credibility of the ECB

97

Furthermore it is unclear whether the Reserve Bank of Australia uses a formal macroeconomic model for policy analysis. Finally the Reserve Bank of Australia only publishes a rough short-term inflation projection; therefore it earns only 0.5 point under the first question. Procedural disclosure is about the way monetary policy decisions are taken. It involves an explicit monetary policy rule or strategy that describes the monetary policy framework, an account of policy deliberations and how the policy decision was reached. It is quantified by the following questions: Does the central bank provide an explicit policy rule or strategy that describes its monetary policy framework?

•

Does the central bank give a comprehensive account of policy deliberations (or explanations in case of a single central banker) within a reasonable amount of time?

•

Does the central bank disclose how each decision on the level of its main operating instrument or target was reached?

•

Only the Reserve Bank of New Zealand and the Bank of England get the full score for procedural disclosure. The Bank of Japan, the Sveriges Riksbank, and the Federal Reserve Bank all earn 2 points. Both the Bank of Japan and the Federal Reserve loose 1 point under the first question above, since they don’t provide an explicit policy rule or strategy that describes its monetary policy framework. The Sveriges Riksbank earns no points under the last question above, since it has no actual voting records available. Policy disclosure means prompt disclosure of policy decisions. In addition it includes an explanation of the decision, and an explicit policy inclination or indication of likely future policy actions. It is quantified by the following questions: Are decisions about adjustments to the main operating instrument or target promptly announced?

•

Does the central bank provide an explanation when it announces policy decisions?

•

Does the central bank disclose an explicit policy inclination after every policy meeting or an explicit indication of likely future policy actions (at least quarterly)?

•

The results for policy disclosure show that only the Reserve Bank of New Zealand and the Federal Reserve earn 3 points. The Bank of

98

Chapter 4

Canada, the Sveriges Riksbank and the Swiss National Bank get 2 points, making them second best. All three central banks earn 0 points under the third question, since they do not provide an explicit indication of their policy inclination or likely future policy actions, although the Swiss National Bank gives its expectations for the future. The remaining central banks loose another 0.5 point, since they only provide an explanation for policy decisions when policy decisions change. Their scores are 1.5 points. Operational disclosure concerns the implementation of the central bank’s policy actions. It involves a discussion of control errors in achieving operating targets and (unanticipated) macroeconomic disturbances that affect the transmission of monetary policy. Furthermore the evaluation of the macroeconomic outcomes of monetary policy in light of its objectives is included here as well. It is quantified by the following questions: Does the central bank regularly evaluate to what extent its main policy operating targets (if any) have been achieved?

•

Does the central bank regularly provide information on (unanticipated) macroeconomic disturbances that affect the policy transmission process?

•

Does the central bank regularly provide an evaluation of the policy outcome in light of its macroeconomic objectives?

•

The results for operational disclosure are very mixed. Only the Sveriges Riksbank earns the full score. The Bank of England is second best with 2.5 points. It looses 0.5 point under the first question above, since it provides an evaluation of the policy outcome only casually and without providing explanations for deviations. The Swiss National Bank has the worst score, only 0.5 points, earned under the same question. It has an operational target range of 100 basis points for threemonth LIBOR rate, but it does not provide an explanation for the significant fluctuations within the target. Furthermore it has only a brief abstract of the analysis of macroeconomic disturbances that affect the policy transmission process available in English. Finally, the Swiss National Bank does not account for discrepancies between policy outcome and target. The disclosure index of Eijffinger-Geraats consists of the overall score for each central bank. The Reserve Bank of New Zealand has the highest rating, that is 13.5 points, while the Sveriges Riksbank is second

Transparency, Accountability, and Credibility of the ECB

99

best with 12 points. The central bank with the lowest score is the Swiss National Bank, with only 7.5 points. It follows from table 4.2 that in terms of information being provided and activities undertaken to distribute this information, the ECB scores comparatively high. In fact the ECB is as open as the Federal Reserve. One obvious objection that is often raised against indictors like this is that they are rather arbitrary and—to a certain extent—subjective. To check whether our conclusion on ECB disclosure is sensitive with respect to the measure chosen, we show in sidebar 7 similar indicators of Gros and Bini-Smaghi (2001), Siklos (2002), and of De Haan and Amtenbrink (2002). We also show the results for the indicator of Fry et al. (2000), even though this indicator is not available for the ECB. It clearly follows that most indicators come up with a very similar ranking of the ECB. The exception is Siklos (2002), who gives the ECB the lowest ranking of the banks under consideration here. That outcome is remarkable because Siklos (2002) takes many of the same issues into account as the other authors and uses similar indicators. On closer inspection it turns out that the low score for the ECB on the Siklos indicator is the result of the relatively high weight of the items “publication of minutes of central bank meetings” and “publication of committee voting record” (on which the ECB scores zero) and a low weight on items like publication of reports, regular speeches on which the ECB gets the highest score possible.7 So far the analysis suggests that the ECB has a relatively high score on disclosure.8 There are, however, indications that financial markets do not have a good understanding of the ECB’s strategy and face considerable uncertainty over monetary policy. Because the policies of the ECB are often considered to be unclear, Issing (2000) concludes that the ECB faces a “communication gap”: On the one hand, few observers contest the success and credibility of the ECB in delivering on its primary objective and on the appropriateness of most of its policy actions in this regard. . . . On the other hand, however, the overall perception of the ECB by the public, academics, financial analysts, market participants, and not least, journalists continues to remain—at best—rather mixed.

The ECB also did not perform well, in a survey by Goldman Sachs taken in February 2000. In this survey financial market participants were asked to rate on a scale of 1 to 5 how well they understood the reasoning behind monetary policy decisions of four central banks (a

7 Other Indicators of Central Bank Disclosure This sidebar summarizes the scores of various central banks according to some other disclosure indicators. We start with the index of De Haan and Amtenbrink (2002). This indicator consists of 14 questions, which, if answered in the affirmative, yield either 1 or 2 points (see table 4.2). The first group of criteria relates to the objectives of monetary policy. If the law and/or central bank documents clearly stipulate the objectives of monetary policy a score of 1 is assigned. If the law or some central bank document identifies clear priorities with regard to different objectives a score of 1 is given. Likewise, when there is a clear definition and time horizon and when a quantification of the objective(s) is given (either in some legal document or some official document from the bank), a score of 1 is given for criteria 3 to 5. The second group of criteria relates to the strategy of the central bank to reach the ultimate objective(s). If the strategy is clearly announced, a score of 2 is assigned. If the decision of the central bank on interest rates is immediately announced and if there is, as a rule, a supporting statement explaining the decision on the basis of the central bank’s strategy, the score is 2 on criterion 7. A value of 2 is given for criterion 8 for clear quantified forecasts of the decision-making unit of the central bank, and a value 1 for “projections” because these are not underwritten by the central bank’s decision-making unit. The final group of criteria refers to the procedures followed by the central bank to effectively communicate with the public. For criterion 9 a score of 2 is assigned if the governor of the central bank appears before the Parliament at least three times a year; the score is 1 in case of one to three annual hearings before Parliament, and zero otherwise. When the central bank publishes a report on a monthly basis, a value of 2 is assigned for criterion 10, a value of 1 in case of quarterly reports, and 0 otherwise. If the schedule of meetings of the policy-making unit is publicly available in time so that it is clear when policy decisions will be taken, De Haan and Amtenbrink (2002) assign a value of 1 for criterion 11. Two points are given for criterion 12 if press conferences are held in regular intervals (monthly) and regular press releases are issued. A value of 1 is assigned if only one of the two is featured and a value of 0, if neither exists. A value of 1 is assigned to criterion 13 when the minutes of the policy-making bodies are released within a reasonable time. Similarly a value of 1 is given for criterion 14 if votes of members of the decisionmaking body are released (with a certain lag). Both the subcategories and the total score with and without the issues related to minutes and information on voting behavior are given. The indicator of De Haan and Amtenbrink (2002) is inspired by a similar index of Gros and Bini-Smaghi (2001). The total score of this index is also shown in table 4.2. Finally, the index of Fry et al. (2000) is shown, although it is not available for the ECB. This index is based on a survey of central banks that provides estimates of many characteristics of disclosure. This indicator ranges from zero (minimum disclosure) to hundred.

Transparency, Accountability, and Credibility of the ECB

101

Table 4.2 Disclosure indicator of De Haan and Amtenbrink compared with other indicators

Objectives 1 2 3 4

Clear objectives Clear priorities Clear definition Clear time horizon 5 Quantification Strategy 6 Announcement of strategy 7 Interest rate decision immediately announced and always explained 8 Inflation forecast Communication strategy 9 Parliamentary hearings 10 Frequency of reports 11 Meeting schedule 12 Press conferences/ press releases 13 Publication of minutes 14 Publication of individual votes Total Subtotal (1–12) Gros and Bini-Smaghi index Siklos index Index of Fry et al. Eijffinger and Geraats index

ECB

Reserve Federal Bank Reserve New Bank of Bank of System Zealand Canada England

Deutsche Bundesbank

4

1

3

5

5

5

1 1 1 0

1 0 0 0

1 1 1 1

1 1 1 1

1 1 1 1

1 1 1 0

1 5 2

0 3 0

1 6 2

1 6 2

1 5 2

0 3 2

2

2

2

2

1

1

1 7

1 7

2 6

2 6

2 8

0 4

2

1

2

2

2

0

2

2

1

1

1

2

1 2

1 1

1 2

1 2

1 2

1 1

0

1

NA1

0

1

0

0

1

NA

0

1

0

16 16 19 (19)

11 9 16 (14)

17/19 17 n.a.

17 17 15 (15)

18 16 24 (20)

10 10 13 (13)

0.52 NA 10

0.87 95 10

0.83 92 13.5

0.83 79 10.5

0.91 94 12.5

0.70 70 NA

102

Chapter 4

higher grade indicates a better understanding). Table 4.3 shows the results. The respondents gave the US Federal Reserve a top rating of 4.3. Furthermore votes were relatively highly concentrated on the two top ratings. Right behind the Fed come the German and British central banks with average scores of 3.5 and 3.3, respectively. Note that in these cases there was quite some dispersion around the mean. The ECB ranked lowest with an average score of just 2.2. Of course, one objection that can be raised against the results reported in table 4.3, is that at the time of the survey the ECB had just been created. For this reason, we re-examined the issue, using more recent information. As part of a survey, which is explained in more detail in section 4.6 and which primarily focuses on the credibility of central banks, we asked our respondents (professional economists) to rank seven central banks according to their transparency. The results are presented in table 4.4. Table 4.3 Transparency ratings of central banks according to a survey by Goldman Sachs Distribution of ratings (in %)

Federal Reserve Bank of England Bundesbanka ECB

1

2

3

4

5

Average

0.0 5.5 2.7 21.2

1.8 10.0 10.0 43.4

7.1 42.7 35.5 27.4

48.7 33.6 38.2 7.1

42.5 8.2 13.6 0.9

4.3 3.3 3.5 2.2

Source: Goldman Sachs, as reported in Gros et al. (2000). a. Until 1999.

Table 4.4 Rankings of central banks in terms of their transparency Central Bank

Transparency

Banca d’Italia Bank of England Bank of Japan Banque de France Deutsche Bundesbank European Central Bank Federal Reserve

3.33 (1.38) 5.93 (1.24) 4.86 (1.04) 2.63 (1.24) 3.17 (1.50) 1.66 (1.21)

6.17 (0.94)

Transparency, Accountability, and Credibility of the ECB

103

Two conclusions can be drawn. First, the results on (perceived) transparency by our respondents are broadly in line with the results presented in table 4.3. The Federal Reserve and the former Deutsche Bundesbank were clearly perceived as more transparent than the ECB. Second, the Bank of England—which according to the disclosure indicators of Eijffinger and Geraats (2002) and De Haan and Amtenbrink (2002) was expected to be very transparent—had a similar score to that of the ECB in both surveys. How can the discrepancy between the high degree of disclosure and the public perception of the ECB’s transparency be explained?9 A possible cause for the low score on perceived transparency may be found in the quality of the information being provided. As pointed out by the IMF (2000a), transparency requires more than just making information available about policy objectives, responsibilities, policy decisions, and performance results. The content of disclosure is critical for the efficient functioning of markets, and its importance will only increase with the evolving changes in international trading and financing arrangements and sophistication of markets. Failure to present public statements and reports on monetary policy issues with appropriate content can undermine the credibility of central banks and result in corresponding behavior by the financial markets, thereby negatively influencing the outcome of monetary policy (see figure 4.1). The focus of disclosure should therefore be on the materiality and relevance of the information that is being provided to the public. The objective of transparency would, for example, not be met by the release of reports that offer contradictory assessments. The same holds true for contradicting public statements by the members of the decision-making organs of the central bank, something that the ECB has been criticized for in the past and sometimes rightly so (see also sidebar 8). For instance, at about the same time that ECB President Duisenberg announced in the European Parliament that the ECB would publish inflation forecasts, Issing (1999b) argued that the ECB would not do so.10 Similarly, not applying disclosure consistently (e.g., reversals of previously applied practices when developments are unfavorable), would defy the spirit and intent of transparency and could moreover weaken credibility. Often the low degree of ECB transparency is related to the monetary policy strategy of the ECB. Begg et al. (2000, p. 25) find, for example, that “Our observations of ECB’s deeds and words in 1999 . . . suggest that much remains to be done to communicate the precise meaning

104

Chapter 4

8 Communicating about the Euro Why would ECB officials communicate about the euro–dollar exchange rate? We see at least three reasons. First, there has been confusion about the role of the exchange rate in the monetary policy strategy of the ECB. Sometimes it is thought that the ECB has an exchange rate objective, which it does not. However, the ECB closely watches the external value of the euro under what it now called the economic analysis (the second pillar) because exchange rate shocks can cause inflation in the euro area (see section 3.6). A second reason why the ECB closely follows exchange rates is that there is always a risk that exchange rates will fall out of sync with “fundamentals,” whatever they may be. By communicating these worries about developments in the exchange rate, the ECB can try to affect the external value of the euro. A third reason why ECB officials have made exchange rate projections is to quell the anxieties of the public about the decline of the euro after the start of the monetary union. Politicians and central bankers had declared that the euro was to be a strong currency, but the external weakness of the euro was often seen to contrast with this claim. Therefore ECB officials kept on re-assuring the public that the decline of the euro was no sign of a weak currency. ECB officials have not always been clear in their communication about the role of the euro in the ECB policies. For instance, in April 1999 Duisenberg said to the European Parliament: “Not having an exchange rate policy, and we have no policy, does not mean that there is benign or malign neglect. For the time being, there is neglect.” Eight days later Duisenberg tried to undo the damage, stating that the ECB does not neglect the euro exchange rate. These and other such instances led Sims and Wessel (2000) to the conclusion that “The European Central Bank Can’t Master Communication.” The frequency of official communications by ECB officials is also quite remarkable. Table 4.5 is reproduced from Jansen and De Haan (2003). It indicates the extent to which some central bankers go to seek publicity. Notably the comments of the president of the Bundesbank and other officials of the German central bank are more often given coverage in the news than those of other central banks.

Transparency, Accountability, and Credibility of the ECB

105

Table 4.5 Look who is talking about the euro

Groups Executive Board National central bank presidents BundesBank officials (excluding president) Total Individuals Duisenberg BundesBank president Tietmeyera Weltekeb Trichet

Positive

Neutral

Negative

Total

149 203 34 386

72 69 30 171

35 64 18 117

256 336 82 674

48 99

27 45

17 35

92 179

17 82 72

6 39 13

1 34 10

24 155 95

Source: Jansen and De Haan (2003). Note: Positive/negative means a statement expected to have an appreciating/ depeciating effect on the euro (e.g., a hint that interest rate may rise/decline). a. January 4, 1999, until October 31, 1999. b. September 1, 1999, May 17, 2002.

of the announced monetary policy strategy. Unless the ECB clarifies its intentions it will take time—possibly a lot of time—for outside observers to form a clear view.” We have discussed this issue already in chapter 3, where we concluded that another monetary policy strategy is likely to enhance the transparency of the ECB. Another factor that has received scant attention in the literature so far is the role of media.11 As shown in figure 4.1, the public depends to a large extent on the information they receive through the media. It becomes thus interesting to see how the media portray ECB policy decisions. In particular, the media may not necessarily fully reflect the decisions taken by the central bank nor delve into the reasoning behind them. Sidebar 9 reports some preliminary outcomes of a project in which reports on monetary policy decisions by two major European newspapers are critically evaluated. Clearly, part of the gap between the ECB’s efforts to provide transparency and the public perception of its inadequacy in this regard may be due to the selective way that media reporters cover monetary policy issues. Differences in reporting style on monetary policy issues by the newspapers are notable. The explanation seems to lie not only in differences of political orientation but, more important, in a different understanding of the role of the ECB.

106

Chapter 4

9 Are Newspaper Reports on the ECB Biased? De Haan and Amtenbrink (2002) examine newspaper reports published one day after the Governing Council of the ECB took a certain policy decision in 1999 and likewise in 2000 in the Financial Times (FT) and in the Frankfurter Allgemeine Zeitung (FAZ). First, with a few exceptions, the reports in the FT and FAZ are correct in their content. Second, relatively little attention is paid to money growth in the FT, in sharp contrast to the FAZ. In instances where something is written about the first pillar of the ECB monetary policy strategy, the FT tends to be critical while the FAZ clearly supports the idea that money should be given a prominent role in the strategy of the ECB. An excellent example occurred in December 1999 when the ECB announced that it would keep the reference value for M3 growth at 4.5 percent. At the press conference where this was announced, President Duisenberg of the ECB criticized the German government for supporting the German construction company Phillipp Holzmann, which at the time was at the brink of bankruptcy; the remark was made within the context of a weakening euro. The FT had three reports on the ECB on December 3, 1999, but did not pay much attention to the bank’s decision on the reference value for M3; it instead focused on Duisenberg’s remarks on the German support for Holzmann. This stands in sharp contrast to the way in which the FAZ reported on the press conference. The article in the FAZ started with the decision by the ECB Governing Council to keep the reference value at 4.5 percent and the reasoning behind this decision. Only at the very end of the report, after a description of what Duisenberg said about the euro and foreign exchange intervention, and within the proper context, did the FAZ mention Duisenberg’s criticism of the German government. This example also illustrates a third characteristic of FT reporting on ECB matters: a strong focus on (alleged) disagreements with governments, the OECD, IMF, or—even more strongly—within the ECB. Very often decisions are disclosed in terms of victory for certain members of the Governing Council. A good illustration in this respect is the FT report on November 5, 1999, in which it is stated that the Bundesbank was opposed to an early rise of interest rates because Germany’s recovery was less strong than in other euro-zone countries, thereby implying that national instead of euro-areawide considerations are a dominant factor in the Governing Council decision-making process. The FT reports almost always express whether the decision by the ECB was in line with financial markets’ expectations. Any hint for future changes of interest rates are spelled out at length. The FAZ pays somewhat less attention to these points. A strong focus on the euro is another characteristic of FT reports on policy decisions. The consequences for the euro–dollar exchange rate of

Transparency, Accountability, and Credibility of the ECB

107

decisions (not) to change interest rates are stressed. In fact it is often suggested—and sometimes stated explicitly—that policy decisions are taken in view of the weak position of the euro. In this respect the FT reports, and particularly the headlines, have been unbalanced and even wrong. For example, in the headline as well in the interior part of the report on February 4, 2001, there was stated that the ECB increased rates to support the euro. The official ECB press release rather stated that: “[T]hese decisions were taken on the basis of an assessment of the risks to price stability in the medium term in the context of the Eurosystem’s monetary policy strategy. Considering the first pillar of this strategy, both monetary and credit growth continue to signal generous liquidity conditions. . . . As for the second pillar of the monetary policy strategy, price and cost increases—including oil and nonenergy commodity prices as well as producer prices—have been larger and more protracted than foreseen earlier and hence indicate the risk of “second round” effects. . . . The depreciation of the euro which we have witnessed is contributing to increases in import prices.” In other words, the rise was motivated by price stability considerations and not by the weakness of the euro, as suggested by the report in the FT. Again, a comparison with the report in the FAZ shows major differences. The FAZ reported that interest rates were increased as the risks for price stability increased stronger than expected, suggesting that: “[I]n this context the weakness of the euro also plays a role.” In effect, the FT is more critical in reporting on the ECB than the FAZ. This is not to say that the FAZ is not ever critical. It has especially criticized severely the lack of a transparent policy with respect to the external value of the euro, but over all, on a good number of occasions the FAZ has demonstated full support for the two-pillar strategy of the ECB. This includes the ECB’s strong focus on money growth, which the FT does not support. Also the differences, especially in newspaper editorials, on the importance of wage moderation, labor market reform, and fiscal policy are remarkable. Whereas the FAZ does not criticize the ECB’s views on these issues, the FT is more critical, and hones in if the ECB goes beyond a plea for deficit reduction and instead indicates how consolidation should be done.

108

Chapter 4

4.4 Accountability of the ECB The European Central Bank (ECB) is widely considered to be a politically independent institution. One potential objection toward a completely independent central bank is lack of democratic accountability (e.g., see Stiglitz 1998). In this section we first outline the concept of democratic accountability. We then compare the legal accountability of the ECB with those of some other central banks (Bank of Canada, Bank of Japan, Bank of England, and the Federal Reserve System).12 The Oxford English Dictionary defines accountable as “obliged to give a reckoning or explanation for one’s actions; responsible.” So, contrary to what Issing (1999b) seems to say, accountability refers not only to the question of whether a central bank has achieved its mandate but also to what the central bank has done in attempting to achieve this mandate. It does not make sense to state that accountability refers to deeds and transparency to words. Since interest rate changes (deeds) are easily observed, what matters is an explanation to the entities to which a central bank is accountable why a certain decision is made (words). The latter refers to accountability. How can this general concept be made operational in relation to central bank accountability? In our view the concept of central bank accountability has three main features: 1. Decisions about the explicit definition and ranking of objectives of monetary policy. 2. Disclosure of actual monetary policy. 3. Who bears final responsibility with respect to monetary policy. In a democratic society it is up to elected politicians to decide on the explicit definition and ranking of objectives of monetary policy. Why should decisions on the objectives of this form of economic policy be treated differently than those of other forms of economic policy, such as fiscal policy? Indeed, it is widely believed that only elected representatives should decide on tax laws and government spending. It is therefore questionable whether it is legitimate in a democratic system to leave the decisions on the objectives of monetary policy in the hands of an independent institution that is not subject to elections or ministerial responsibility. Clearly, monetary policy objectives should be defined. As we showed in chapter 2, the primary objective of the ECB (i.e., to maintain price

Transparency, Accountability, and Credibility of the ECB

109

stability) is no where precisely specified in primary Community law. In the current setting it is left to the ECB to provide an operational expression of this primary objective. Any institution—be it the European Parliament or some other entity—charged with holding the central bank accountable is therefore strictly not provided with an effective statutory yardstick by which to evaluate the performance of the ECB, and thus cannot hold the bank accountable for its conduct of monetary policy. The choice of a single objective, of course, simplifies the monitoring of central bank performance. The announcement of a single goal (or a primary goal), rather than several unranked goals, enables authorities and public opinion to effectively follow an institution’s performance. In this sense the Treaty on European Union (TEU) and the Statute of the European System of Central Banks and of the European Central Bank are very clear: they provide a hierarchy of goals. This is, in our view, a sensible thing to do. In contrast, the Federal Reserve System faces multiple conflicting objectives (maximum employment, stable prices, and moderate long-interest rates), and neither the Federal Reserve Act nor any other law provides any hierarchy. A good example of a central bank with a clear prescription of objectives is the Reserve Bank of New Zealand, which has the pursuit of price stability as its primary objective. The governor of the Reserve Bank of New Zealand has to agree with the government a tight target range for inflation. The document that defines the target range for the inflation rate is called a Policy Target Agreement (PTA). We advocate this system with its clearly defined a target range for the inflation rate for European monetary policy arrangements as well. Disclosure is also an important component of accountability. Central bank officials can be held accountable for their behavior where the reasoning behind and strength of opinion supporting certain monetary policy decisions are clear. A central bank should be required to report at regular intervals on its past performance and future plans for monetary policy in accordance with some monetary objective. This is especially necessary where no clear monetary objective exists because then the central bank must be judged on the basis of its statements alone. As disclosure should not be at the discretion of the central bank, the law must prescribe certain procedures about explaining monetary policy. There are various ways to do this, ranging from reports, to minutes, to other communication devices. Legally the ECB is compelled to publish a quarterly statement of its activities.

110

Chapter 4

Whether and to what extent these statements include details of past performance, or future monetary policy plans is, again, left to the ECB to decide. The TEU and ECB statute do not specify the contents of these reports. Concerning the final responsibility for monetary policy, we think that three issues are crucial: the ECB’s relationship with the European Parliament, the introduction of an override mechanism, and a dismissal procedure for the European Central Bank’s governor. The relationship between a central bank and parliament has to play a major role in any evaluation of the democratic accountability of a central bank. There should be a legal requirement for the central bank to report and/or explain its policy actions to parliament. Parliament should have the opportunity to review the performance of the central bank on a regular basis so that the central bank can at the same time explain and justify its operations accordingly. Apart from the obligatory presentation of an annual report by the president of the ECB, the European Parliament could ask the members of the Executive Board to appear before Parliament as well. The ECB has gone some way in this direction as the president of the ECB has expressed his willingness to appear before the European Parliament at least four times a year in addition to his presentation of the annual report. We fully agree with Buiter that it is essential that “the European Parliament act as an effective watchdog over the ECB. The legitimacy of the ECB will depend on the extent to which it is effectively accountable to the European Parliament” (Buiter 1999a, p. 200). We would even like to go a step further. It has been argued in the context of national central banks in general that parliament always holds the ultimate responsibility for monetary policy because it can change the legal basis of the central bank. Indeed, the mere threat of a change of law can ensure that independent central banks (e.g., the Bundesbank) will behave in general in accordance with the wishes of elected politicians (Gormley and De Haan, 1996). However, the power of national parliaments with respect to the legal basis of the ECB is quite limited. Any change must require an amendment to the primary Community law to which, by implication, all countries have to agree. We believe that the European Parliament should have the final say in the statute of the ESCB and thus act in a real parliament any way. In general, a central bank is not only directly accountable to a parliament but also to a government, which is itself accountable to the parliament. So it is important that a government be able to influence its

Transparency, Accountability, and Credibility of the ECB

111

central bank’s behavior. Without such instruments, accountability cannot go beyond mere reporting by government to parliament of central bank policies, for which government in that case cannot be held responsible. An override mechanism for government would be an instrument to change central bank policy. If the government does not interfere, it apparently agrees with central bank policies and can be held accountable for this by parliament. Such a mechanism existed in the Netherlands before the latest change of the Dutch central bank law. It was at one time also considered for the ECB but rejected for various reasons. (One reason was that at the European level there is no equivalent to a national minister of finance, nor is there at the European level a full equivalent to a national parliament.) If an override mechanism exists, it is critical that the conditions by which it can be applied be laid down in full. The mechanism must not be allowed to be used as a tool for undesired political influence. The procedure for the application for the override mechanism must be transparent. The decision to apply the override mechanism must be made public. Furthermore, the procedure to apply an override must provide for a review (e.g., allow the central bank to appeal) that ensures that the override is used judiciously. Finally, the dismissal procedure for a central banker can account to a mechanism of ex post accountability if a central bank official can be dismissed for bad performance, that is, for not meeting the stated objectives. This is the case for the Reserve Bank of New Zealand where the PTA between the governor of the central bank and the minister of finance lays down the bank’s policy targets. Inadequate performance can result in dismissal of the governor. In the case of the president of the ECB, in contrast, dismissal results if he can no longer fulfill the general conditions required for his performance or if there is serious misconduct. Presumably a mechanism like the one in New Zealand is based on the principle of individual responsibility for monetary policy making. Therefore this mechanism might not be appropriate since the ECB Governing Council is collectively responsible for policymaking (Issing 1999b). How does the accountability of the ECB compare to that of other central banks? De Haan et al. (1999) provide an indicator for democratic accountability based on the general definition we gave above.13 Table 4.6 is reproduced from this study. For the 13 categories of accountability the number of positive answers to the questions determines the score for the bank. The accountability indicator is based on

112

Chapter 4

Table 4.6 Accountability of central banks

Various aspects of accountability 1. Does the central bank law stipulate the objectives of monetary policy? 2. Is there a clear prioritisation of objectives? 3. Are the objectives clearly defined? 4. Are the objectives quantified (in the law or based on document based on the law)? Subtotal on ultimate objectives of monetary policy 5. Must the central bank publish an inflation or monetary policy report of some kind, in addition to standard central bank bulletins/ report? 6. Are minutes of meetings of the governing board of the central bank made public within a reasonable time? 7. Must the central bank explain publicly to which extent it has been able to reach its objectives? Subtotal on transparency 8. Is the central bank subject to monitoring by Parliament (is there a requirement—apart from an annual report—to report to Parliament and/ or explain policy actions in Parliament)? 9. Has the government the right to give instructions? 10. Is there some kind of review in the procedure to apply the override mechanism? 11. Has the central bank possibility for an appeal in case of an instruction? 12. Can the central bank law be changed by a simple majority in Parliament?

Bank Bank of of Bank of Canada Japan England

Federal Reserve System ECB

Yes

Yes

Yes

Yes

Yes

No

No

Yes

No

Yes

No No

No No

Yes Yes

No No

No (yes) No (yes)

1

1

4

1

2 (4)

Yes

—

Yes

Yes

No (yes)

No

No

Yes

Yes

No

Yes

Yes

Yes

Yes

Yes

2 Yes

1 Yes

3 Yes

3 Yes

1 (2) Yes

Yes

Yes

Yes

No

No

Yes

Yes

Yes

Yes

No

No

No

No

No

No

Yes

Yes

Yes

Yes

No

Transparency, Accountability, and Credibility of the ECB

113

Table 4.6 (continued)

Various aspects of accountability 13. Is past performance a ground for dismissal of a central bank governor? Subtotal on final responsibility Total on accountability

Bank Bank of of Bank of Canada Japan England

Federal Reserve System ECB

No

No

No

No

No

4 7

4 6

4 11

2 6

1 4 (7)

Source: De Haan et al. (1999).

central bank laws, like most indicators for central bank independence (see chapter 6). As this index shows, the ECB has a low degree of (legal) democratic accountability. Recall from our discussion above that the ECB goes in some respects further than the law prescribes (operationalization of the objective, monthly report, the reasoning behind specific decisions is made public immediately, appearances before European Parliament). Therefore the final column of table 4.6 also provides (in parentheses) the score for the ECB if actual practice is taken into account. While the ECB has then a higher score, the ECB cannot be considered as “the most transparent and accountable central bank in the world” (Issing 1999b, p. 505). 4.5 Blinder’s Survey on Credibility Although credibility of policy makers is widely believed to be important, empirical evidence in support of this view is scarce. A recent interesting study is by Cecchetti and Krause (2002), who use the information provided by Fry et al. (2000) to examine the extent to which macroeconomic performance in their sample of 63 countries is related to credibility, transparency, independence, and accountability of central banks.14 They find that central bank credibility, and to a lesser extent, transparency, is tied to inflation. However, as Cechetti and Krause constructed their index of credibility on past inflation, their results are perhaps better interpreted as suggesting that inflation is persistent. Not only is little known about the effects of credibility, we also know little about the determinants of credibility. According to Issing

114

Chapter 4

(2001a), transparency of monetary policy can enhance credibility. By providing the public with adequate information about its activities, the central bank can establish a mechanism for strengthening its credibility if its actions match its public statements (IMF 2000a). In Jensen’s (2001) model increased transparency can increase the reputational costs of deviations from the inflation target and therefore increase the credibility of the central bank. However, the credibilityenhancing effect of transparency is redundant where central bank preferences are already public information. Chortareas et al. (2002) argue that a high degree of transparency is desirable for central banks with poor credibility but may be costly in terms of less flexibility for high-credibility banks. Whether transparency enhances the credibility of policy makers is, however, not extensively investigated in the empirical studies by Chortareas et al. (2002) and Cecchetti and Krause (2002). The latter authors only report that the correlation between their index of credibility and their index of transparency is 0.31, while credibility is virtually unrelated with their measures of accountability and independence. Recently some authors have presented the results of surveys among central bankers on transparency and credibility. The comprehensive survey by Fry et al. (2000), to which we referred above, showed 74 percent of the respondents to consider transparency a vital or very important component of their monetary policy framework. Blinder (2000) mailed a questionnaire to the heads of 127 central banks soliciting their opinions on a number of issues related to central bank credibility. The response rate was 66 percent. The respondents were asked to answer questions relating to the importance and determinants of central bank credibility. Table 4.7 summarizes Blinder’s findings. The first question (Q1) is: How important is credibility to a central bank? In this case the points on the five-point scale were as follows: 1 = unimportant 2 = of minor importance 3 = moderately important 4 = quite important 5 = of the utmost importance The average score to the first question is a stunning 4.83, with a standard deviation of only 0.37. Blinder (2000) asked the same question to

Transparency, Accountability, and Credibility of the ECB

115

Table 4.7 Blinder’s (2000) survey of central bankers about credibility Question

Issue

Average

Standard deviation

Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14

Importance Related to dedication to price stability Less costly disinflation To keep inflation low To change tactics To serve as lender of last resort To defend the currency Public servants should be truthful For support of independence Ranking of Q3 to Q9 Importance of CBI for credibility Importance of transparency for credibility Importance of history of honesty for credibility Importance of history of fighting inflation for credibility Importance of being constrained by a rule for credibility Importance of incentives (personal loss) for credibility Importance of small deficit and low debt ratio for credibility

4.83 4.10 4.13 (2) 4.39 (1) 4.38 (5) 4.12 (6) 4.29 (3) 4.00 (7) 4.34 (4) — 4.51 (2) 4.13 (4) 4.58 (1) 4.15 (3)

0.37 NA 0.78 0.60 0.54 0.77 0.70 0.84 0.75 — 0.63 0.71 0.52 0.67

2.89 (6)

1.01

2.15 (7)

1.10

3.92 (5)

0.93

Q15 Q16 Q17

Source: Blinder (2000). Note: Ranking of Q3–Q9 and Q11–Q17 is shown in parentheses.

a smilar-sized sample of aceademic economists. The average score for the first question in this group is somewhat lower (4.23) while the standard deviation is higher (0.85). The second question (Q2) is: How closely related are the concepts of (a) a central bank’s credibility and (b) a central bank’s dedication to price stability? In this case the five-point scale has the following meaning: 1 = unrelated 2 = slight related 3 = moderately related 4 = quite closely related 5 = virtually the same

116

Chapter 4

As follows from table 4.7, the average score on this question in Blinder’s survey is 4.10. The academic economists give a considerably lower score (3.31). The next seven questions focus on the issue of why credibility might be important to a central bank. Respondents were asked to express their agreement or disagreement on a five-point scale, ranging from strongly disagree to strongly agree. One argument—with a score of 4.13 in the survey of Blinder (Q3)— is that credibility will reduce disinflation costs. In the ranking it even gets the second place, also in the survey among academic economists. Surprisingly, given this high score, the academic literature has come up with discomforting outcomes as almost all studies report that CBI worsens the trade-off. For instance, Posen (1998) finds a positive correlation between CBI and the sacrifice ratio, i.e., the cumulative increase in unemployment that is due to the disinflation effort divided by the total decrease of inflation (see Eijffinger and De Haan 1996 and Berger et al. 2001 for extensive surveys of the literature on CBI).15 Still one might argue that what really matters for disinflation costs from a theoretical point of view is credibility, which may be influenced by actual (instead of legal) independence of a central bank. So far lack of indicators for credibility have made more direct testing impossible. Table 4.7 also summarizes the outcomes for the other reasons given by Blinder as to why credibility may be important. Interestingly the argument “to keep inflation low” receives the highest average score. Also the argument that a credible central banker may find it easier to change operating procedures—as for instance, the Federal Reserve did under Volcker in 1982—receives a very high average. Question 10 of Blinder’s survey lists the reasons indicated in the previous seven questions on the importance of credibility and asked the respondents to rank them. The ranking is shown in parentheses in table 4.7. Note that there are some inconsistencies in the average scores and the rankings. For instance, the average score for the answer that credibility will allow central banks to change tactics is 4.38, only slightly lower than the average score for the argument that credibility will help to keep inflation low (score 4.39). However, the changing-tactics argument is only ranked fifth in question 10. The argument that is it less costly to disinflate receives the fifth score in terms of the average score, while it is ranked second in question 10. The rankings of the academic economists differs somewhat from those of central bankers. The largest difference occurrs with respect

Transparency, Accountability, and Credibility of the ECB

117

to the support-for-independence argument, which is ranked seventh by the economists and fourth by the central bankers. However, the academic economists agree with central bankers in their number one and two rankings (i.e., to keep inflation low and less costly to disinflate). How can credibility be earned? Questions 11 to 17 in Blinder’s survey give various possible answers of which the scores are again shown in table 4.7. The top-rated method for a central bank to establish credibility, at least according to central bankers, is to “have a history of doing what it says it will do.” Although the economic literature is full of optimal contracts for central bankers (e.g., Walsh 1995) and incentive-compatible payment schemes (e.g., Svensson 1997), central bankers give these options a rather low rating, as do academic economists. 4.6 Credibility of the ECB We were able to survey economists participating in Ifo’s World Economic Survey (formerly known as Economic Survey International, ESI). The Ifo Institute has been running this survey since 1981 (Brand et al. 1997; Haupt and Waller 2000).16 The aim is to maintain an up-to-date quarterly picture of the global economic situation as well as forecasts for the important industrialized, emerging, and developing nations. The veracity of the World Economic Survey (WES) results was tested in a number of studies. The results can be summarized as follows (Brand et al. 1997): The survey results’ indicators were valuable in explaining global economic developments.

•

The results were suitable for use in forecasting economic developments, although forecasting abilities are inferior to their explanatory power.

•

The number of participants in the survey, which varied from country to country, did not seem to have much impact on the quality of the survey results.

•

Most of the questions that we asked were similar to those in the Blinder (2000) survey.17 Our first question corresponds to Q1 in table 4.7, How important is credibility to a central bank? Following Blinder (2000), we did

118

Chapter 4

not provide our respondents with a definition of credibility.18 Blinder (p. 1422) motivates this as follows: “I deliberately failed to provide a precise definition of credibility, allowing each respondent to attach his or her own preferred meaning to the term. In fact, there appears to be no generally agreed-upon definition.” We also did not ask our respondents to give their definition of credibility. Rather as our second question, we asked about the relationship between credibility and dedication to price stability. This question is almost the same as Q2 in table 4.7, How closely are the concepts of credibility and dedication to price stability related? The possible answers on a five point-scale ranged from unrelated (1) to virtually the same (5). Because we could only ask a small number of questions, we did not include Blinder’s questions Q3 to Q9. Instead, we asked the respondents to rank the seven reasons. Can you rank (from 1 to 7, where 1 is highest) the following reasons that are often considered as explanations why credibility may be important for a central bank? •

A credible central bank can reduce inflation at lower social cost.

A credible central bank is better able to maintain low inflation once low inflation has been achieved.

•

A credible central bank can more easily change tactics or operation procedures without upsetting markets or raising concern over its underlying objectives or resolve.

•

A credible central bank can more easily act as a lender of last resort in a financial crisis (e.g., during a market crash or bank run) without creating fears that it has lost its dedication to fighting inflation.

•

A credible central bank can more easily defend a country’s currency in a speculative attack.

•

Central bankers are public servants that have a duty to be open and truthful.

•

Credibility is important as a way to justify public support for an independent central bank.

•

The outcomes of this question compare with those of question Q10 in Blinder’s survey. Next we had our respondents rank the various possibilities given by Blinder as to how a central bank can build credibility. So Blinder’s questions Q11 to Q17 are combined to, Can you rank (from 1 to 7, where 1 is

Transparency, Accountability, and Credibility of the ECB

119

highest) the following means which have been suggested to establish or create central bank credibility? •

The central bank should have a high level of independence.

•

The central bank should be open and transparent.

The central bank should have a history of doing what it says it will do.

•

•

The central bank should have a history of fighting inflation.

The central bank should be bound (by law or by custom) to follow a prescribed rule that constrains decision making.

•

The central bank governor should take a personal loss (e.g., lower salary or step down) if inflation rises too high.

•

Absence of high fiscal deficit and debt ratio create central bank credibility.

•

The survey was held for the first time in May 2000. More than 200 people filled in the questionnaire. This gave a response rate of 45 percent. The survey respondents included economists from various countries, allowing us to differentiate between economists located in EMU countries and those located elsewhere. We could also differentiate between economists affiliated with financial institutions (banks, insurance companies, etc.) and those working for firms or research institutes. Last, the survey allowed us to determine whether the inflation experience of the respondent’s country was systematically related to the answers given. Table 4.8 summarizes the main findings for the first questions of our survey. Note that the figures for questions 3 and 4 are ranking averages and cannot be compared to the ratings of Blinder’s survey. As far as the importance of credibility is concerned (Q1), our respondents gave almost as high a mark as central bankers, although the standard deviation is somewhat higher. Also as far as the relationship between credibility of a central bank and its dedication to price stability is concerned (Q2) the average score of our respondents is very close to those of Blinder’s survey among central bankers. Professional economists apparently agreed more on this issue with central bankers than with academic economists. With respect to the question of why credibility is important (Q3), some notable differences between the two surveys show up. The strongest divergence of rankings is on the usefulness of credibility for

120

Chapter 4

Table 4.8 Our survey of private sector economists about credibility Question

Issue

Average

Standard deviation

Q1 Q2 Q3

Importance Related to dedication to price stability Less costly disinflation To keep inflation low To change tactics To serve as lender of last resort To defend the currency Public servants should be truthful For support of independence Importance of CBI for credibility Importance of transparency for credibility Importance of history of honesty for credibility Importance of history of fighting inflation for credibility Importance of being constrained by a rule for credibility Importance of incentives (personal loss) for credibility Importance of small deficit and low debt ratio for credibility

4.66 4.02 3.31 (2) 3.40 (3) 3.26 (1) 3.74 (4) 3.96 (5) 5.82 (7) 4.53 (6) 1.80 (1) 3.13 (3) 2.93 (2)

0.49 0.61 2.09 1.68 1.72 1.71 1.91 1.70 1.86 1.32 1.46 1.43

3.70 (4)

1.36

4.85 (5)

1.47

6.38 (7)

1.07

5.26 (6)

1.49

Q4

changing tactics. Our respondents give this reason for the importance of credibility the highest ranking, whereas the central bankers in Blinder’s survey placed this fifth. Another interesting result is the ranking of the importance of credibility for price stability in both surveys. Although central bankers put this on top of their list, our respondents ranked it third. The role of credibility in central bank independence is only minor according to our respondents. This result is in line with Blinder’s findings for academic economists, who ranked it seventh. However, central bankers give it rank 4. With respect to the question as to how credibility can be built (Q4) the rankings of our respondents are broadly in line with those of the central bankers (and academic economists) in Blinder’s survey. A history of honesty and central bank independence gets the highest ranking in both surveys. Personal incentives for central bankers are not regarded as adequate to earn credibility.

Transparency, Accountability, and Credibility of the ECB

121

Table 4.9 Our survey of private sector economists about credibility: Economists from EMU countries, financial institutions, and high-inflation countries

Question

Issue

Q1 Q2

Importance Related to dedication to price stability Less costly disinflation To keep inflation low To change tactics To serve as lender of last resort To defend the currency Public servants should be truthful For support of independence Importance of CBI for credibility Importance of transparency for credibility Importance of history of honesty for credibility Importance of history of fighting inflation for credibility Importance of being constrained by a rule for credibility Importance of incentives (personal loss) for credibility Importance of small deficit and low debt ratio for credibility

Q3

Q4

EMU based

Financial Institutions

Highinflation countries

4.65 4.10

4.78 4.15

4.67 4.03

3.29 (2) 3.38 (3) 3.41 (1) 3.62 (4) 3.99 (5) 5.75 (7) 4.57 (6) 1.89 (1) 3.05 (3)

2.82 (1) 3.49 (3) 3.12 (2) 3.81 (4) 4.28 (5) 5.99 (7) 4.54 (6) 1.75 (1) 3.09 (3)

3.43 (3) 3.75 (4) 3.33 (1) 3.85 (5) 3.40 (2) 5.71 (7) 4.57 (6) 1.75 (1) 3.36 (3)

3.01(2)

2.97 (2)

2.89 (2)

3.71 (4)

3.59(4)

3.88 (4)

4.79 (5)

5.18 (5)

4.75 (5)

6.43 (7)

6.19 (7)

6.36 (7)

5.11 (6)

5.35 (6)

5.05 (6)

Table 4.9 shows the outcomes of three subsamples of our respondents: economists from euro-area countries, economists affiliated with a financial institution, and economists based in countries that had relatively high inflation rates in the past.19 The answers from economists from banks and insurance companies as to the reasons why credibility is important and how it can be built turned out to be very much in line with the results for our total sample. The same was true for economists located in the euro area and in countries with relatively high inflation rates in the past. Last, we asked our respondents to rank seven central banks with respect to their credibility, Can you rank (from 1 to 7, where 1 is highest) the following central banks in terms of their credibility?

122

Chapter 4

Table 4.10 Rankings of central banks in terms of their credibility Central bank

Average

Standard deviation

Banca d’Italia Bank of England Bank of Japan Banque de France Deutsche Bundesbank European Central Bank Federal Reserve

6.42 (6.29) 3.74 (3.56) 5.48 (5.71) 4.86 (4.82) 1.89 (2.25) 3.86 (3.52) 1.79 (1.59)

1.02 1.19 1.29 1.14 1.12 1.39 1.10

Note: The figures in parentheses in the second column show the outcome of a second survey held in October 2001.

•

Banca d’Italia

•

Bank of England

•

Bank of Japan

•

Banque de France

•

Deutsche Bundesbank

•

European Central Bank

•

Federal Reserve

Table 4.10 presents our findings for the credibility marks of the various central banks. Our respondents gave the Federal Reserve the highest mark, closely followed by the Bundesbank. The credibilities of the Bank of England and the ECB were clearly higher than those of the Bank of Japan and the Banca d’Italia. Note that the ranking of central banks in terms of credibility is the same as the ranking in terms of transparency (table 4.4). One explanation may be that our respondents considered the concepts to be closely related. Alternatively, they could have simply ranked central banks on the basis of their recent performance. Finally, table 4.11 shows the scores for three different groups of respondents in our sample: economists based in an EMU country, economists affiliated with financial institutions, and economists from a high-inflation country. Interestingly economists from EMU countries gave the ECB a somewhat higher rating. The rating of the ECB by economists affiliated with financial institutions was a bit lower than in table 4.9, but the opposite can be seen for economists from high-inflation

Transparency, Accountability, and Credibility of the ECB

123

Table 4.11 Rankings of credibility of central banks: Economists from EMU countries, financial institutions, and high-inflation countries Central bank

EMU based

Financial Institutions

High-inflation countries

Banca d’Italia Bank of England Bank of Japan Banque de France Deutsche Bundesbank European Central Bank

6.30 3.95 5.53 4.83 2.03 3.56

6.24 3.71 5.80 4.76 1.94 4.03

6.23 3.65 5.68 4.90 1.94 3.70

Federal Reserve

1.81

1.50

1.90

countries. To examine whether over time the credibility of the ECB had improved, we repeated question 5 in another survey held in October 2001. In table 4.10 the results of this second survey are shown in parentheses. It is reassuring to see that the rank of the ECB had improved somewhat. So far we have only analyzed and compared the results of the total samples of both surveys. However, the samples do differ significantly in size—the sample of the first survey was almost twice as big as in the follow-up survey—and in composition. Some participants took part in the first survey but not in the second, and vice versa. The answers of experts who participated in both surveys allowed for a more in-depth analysis of the change of assessments and opinions over time. Unfortunately, this reduced the sample to 58 participants. It turned out that the results for this group were identical to those of the full sample (not shown). Respondents in EMU countries gave the ECB a higher rating in the follow-up survey. Although they generally considered the ECB a bit less credible than most other respondents, among respondents from financial institutions the ECB gained ground and surpassed the Banque de France. 4.7 Conclusions We distinguished among five aspects of openness of central banks: political, economic, procedural, policy, and operational disclosure. From the review of the theoretical literature on central bank disclosure and transparency we found no consensus over the desirability, from an

124

Chapter 4

economic point of view, of the various categories of disclosure and transparency. Clearly, more study needs to be done on the economic consequences of disclosure and transparency. According to the Eijffinger-Geraats index of disclosure, the ECB has a relatively good overall rating due to its good scores for political and economic disclosure. The ECB also has a high rating in terms of most other indicators for central bank disclosure. This contrasts with the findings of various surveys among private sector economists, including our own, in which the ECB’s operations were perceived to be opaque. Our indicator for accountability suggests that the ECB does not rank high, although the ECB goes further than required by law in providing information. As to perceived credibility, the respondents in our survey did not give the ECB a high rating. Still our results suggest that over time the credibility of the ECB has increased somewhat. The views of our respondents as to how important credibility is for monetary policy making and its determinants are broadly in line with those of central bankers.

5

Centralization or Decentralization

5.1 Introduction As we pointed out in chapter 1, within the Eurosystem the national central banks play an important role. Unlike other similar central banks, the ECB is very decentralized. This is not only true for the implementation of monetary policy (as discussed in chapter 2) but also for the decision-making process within the Governing Council of the ECB. In this chapter we consider first, in section 5.2, the decentralized nature of the Eurosystem. In section 5.3 we show how this decentralized setup can be problematic. A crucial factor is the extent to which the countries in the euro area are alike. Where countries (or sizable regions) are at different points in the business cycle or have diverging inflation rates, decision-making on the appropriate monetary policy stance becomes a difficult task. Furthermore, if the monetary policy transmission mechanism differs across the member countries of the monetary union, the appropriate size and timing of monetary policy decisions become difficult to assess. Various authors have argued that monetary policy transmission differs substantially among countries in the European monetary union because of differences in financial structures. In sections 5.4 to 5.6 we turn to analyze synchronization of the business cycles of the countries in the euro area, their diverging inflation rates, and the differences in monetary transmission across EMU countries. In section 5.7 we present our conclusions. 5.2 The Decentralized Eurosystem Regional central banks have a very strong voice in the decision making of the Eurosystem. Table 5.1 compares the ECB with the German and American central banks in terms of the power of the regional central

126

Chapter 5

Table 5.1 Distribution of voting power: Center versus regions Bank

Center

Regions

Center/region

Bundesbank before 1957 Bundesbank before 1992 Bundesbank before 1999 Federal Reserve System ECB 1999 ECB 2001

1 7 8 7 6 6

9 11 9 5 11 12

0.11 0.64 0.89 1.40 0.55 0.50

Source: Berger et al. (2004).

banks. It follows that the decision-making structure of the ECB is less centralized than that of comparable central banks. In terms of voting power, the position of the Executive Board within the Eurosystem is therefore relatively weak. Bini Smaghi and Gros (2000) argue that its powers are also limited in some other respects. First, the Board has the responsibility to set up meetings of the ECB Governing Council, but it does not have an exclusive power of initiative. Second, several committees have been created in all major areas of ECB competence to assist in the work of the Eurosystem. These committees are mandated by the Governing Council and also—generally via the Executive Board— report to the Council. Third, the Executive Board has no budgetary autonomy, since the Governing Council has to approve the budget. Finally, the ECB has no control over the many activities performed by national central banks. They are free to perform certain functions, unless the Governing Council decides by two-third majority that these activities interfere with the objectives and tasks of the Eurosystem. The ECB also differs considerably from the Federal Reserve and the Bundesbank in terms of economic size. Table 5.2 shows the distribution of the share in total GDP of the member central banks of the Bundesbank, the Federal Reserve, and the Eurosystem. Note that in the United States the regional central banks are close in size, whereas in the euro area the economic sizes are uneven. Germany is in-between: the central banks of Nordrhein–Westfalen (22 percent) and Bavaria (17 percent) have substantially more economic power than most of the other banks. As a consequence of the institutional setup of the Governing Council, political power and country size diverge. As shown in figure 5.1, for seven out of the current twelve countries in the monetary union the

Centralization or Decentralization

127

Table 5.2 Share in total GDP of central banks (distribution), 1999 Share in GDP (%)

Germanya

United States

EMU

0–5 5–10 10–15 15–20 20–30 >30 Total

0 5 2 1 1 0 9

1 10 0 1 0 0 12

7 2 0 1 1 1 12

Source: Berger and De Haan (2002). a. Germany: 1998.

LUX IRL PRT GRC FIN AUS Political weight (PW)

BEL NLD ESP ITA

Economic weight (EW)

FRA GER 0

5

10

15

20

25

30

35

Share (%) 1 Council seat Figure 5.1 Economic and political weights of central banks in the current euro area. (Source: Berger et al. 2004)

128

Chapter 5

political weight (i.e., 1/18 or about 6 percent of all votes in the Governing Council) exceeds economic weight (i.e., their share of the euro area GDP). Another important difference with the Bundesbank and the Fed is that the staff employed by the ECB is much smaller than that of the central banks of its system (see table 5.3). In time this imbalance may be corrected.1 As Bini Smaghi and Gros (2000) point out, any proposal by the Executive Board of the ECB to increase staff has to be approved by the Governing Council. An opposite situation exists in the United States, where the Board of Governors’ responsibility is to approve the budgets of district banks. Besides staff size, there is the relevance of staff qualifications. This has many dimensions, an important one being the staff’s research activities. As sidebar 10 shows, based on publications in international journals relative to the size of the staff involved in research, the position of the ECB is surprisingly strong. Table 5.3 Staff: Total size and size of the research departments, 1990–1999 Averages Central Bank of

Total staff

Research staff

Ratio

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden United Kingdom ECB

1,243 2,695

50 49 33 45 NA 69 118 23 212 22 95 129 253 42 116 110

0.041 0.018 0.057 0.055 NA 0.025 0.038 0.037 0.023 0.156 0.057 0.067 0.080 0.056 0.031 0.172

582

826 NA 2,761 3,240 621 9,229 141 1,655 1,966 3,175 807 4,050 633

Source: Eijffinger et al. (2002). Notes: Data for Luxembourg and the ECB refer to 1998–99; data for Sweden refer to 1992–98. For Portugal the data until 1996 include statistics staff.

Centralization or Decentralization

129

10 Research Output of European Central Banks From the perspective of a regional central bank in a decentralized system of central banks quality of research is important (Goodfriend 1999; Angeloni 1999). The diversification of research within a system of central banks brings a variety of analytical perspectives to policy deliberations that are invaluable in an increasingly complex economy. Moreover a system of regional banks harnesses competitive forces to encourage innovative thinking within the central bank. Eijffinger et al. (2002) surveyed the research output of the European central banks for the period 1990 to 1999. The questionnaire pinpointed the following concerns: Total staff. How many employees were working at your central bank during the last ten years? • Research staff size. How many employees were working at the economic and research departments, measured over the same period? • Publication record. How many scientific journal articles have been published by the staff of your economic and research departments in international, refereed journals and when were these articles published? •

Eijffinger et al. (2002) measure quality by counting the number of journal publications per employee in the research and economics departments. Measuring output for each employee seems natural because the various central banks differ a lot in size. As considerable quality differences exist between scientific journals, they apply a weighting scheme. All international journals in which central bank employees had published have been ranked into three classes: top journals, very good journals, and good journals. A top publication delivers three points, a very good publication two points and a good publication one point. Eijffinger et al. have calculated the research output per employee by multiplying the number of journal articles with the respective scores for the journal (3, 2, or 1) and dividing the resulting sum by the number of employees. Table 5.4 shows the results. The table shows that the ECB is in sixth place. However, if the output over the period 1994 to 1999 is also taken into account, the ECB occupies the fourth place, before the central banks of Austria and Portugal. This is a remarkable achievement, given the relatively short period for which data are available. In a more recent assessment of ECB research, Goodfriend et al. (2004, p. 46) similarly conclude that “Our overall assessment of ECB research is quite positive. In the short time since its creation, the ECB has managed to build up significant research capabilities, not only in DG Research but also in other business areas. This shows in the hiring of PhD economists, the production of a large number of working papers, the publication of research in leading journals, and the participation of ECB staff in many academic conferences and workshops.”

130

Chapter 5

Table 5.4 Weighted journal publications per employee of European central banks (including the ECB), 1990–1999

Central Bank of

Total number of journal publications

Quality-weighted number of articles

Quality-weighted number of articles per employee in research

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden United Kingdom ECB (1994–99) ECB (1998–99)

14 7 3 20 NA NA NA 3 89 NA 49 31 29 5 8 29 13

17 9 5 35 NA NA NA 4 137 NA 68 50 51 12 14 50 23

0.34 0.19 0.15 0.78 NA NA NA 0.17 0.64 NA 0.72 0.39 0.20 0.24 0.08 0.45 0.21

Notes: Figures for the United Kingdom and Sweden refer to 1998–99. Figures for the ECB refer to 1994–99 or 1998–99. In the first case the publications of the ECB research staff during 1994–97 were also counted. Information for the United Kingdom is incomplete, while data for Italy were made available after the publication of Eijffinger et al. (2002).

5.3 Risks of Decentralization If the Executive Board of the ECB is in a relatively weak position visà-vis the national central banks in the Eurosystem, what will be the consequences of this institutional setup? Various critics of EMU have raised concerns that a common monetary policy can be undermined by diverging economic developments in the countries in the eurozone.2 Indeed, the severity of the depression of the 1930s has been attributed to serious policy mistakes by the Federal Reserve, caused by its excessively decentralized decision-making structure, by Friedman and Schwartz (1963, pp. 415–16) in their seminal study of American monetary policy:

Centralization or Decentralization

131

There is more than a little element of truth in the jocular description of a committee as a group of people, no one of whom knows what should be done, who jointly decide that nothing can be done. And this is especially likely to be true of a group like the Open Market Policy Conference, consisting of independent persons from widely separated cities, who share none of that common outlook on detailed problems or responsibilities which evolves in the course of longtime daily collaboration.

In Friedman and Schwartz’s view, the consequence of a weak center is that the decision-making process becomes too cumbersome and slow, possibly resulting in suboptimal decisions. The risk of suboptimal decisions in the euro area is especially large if national central bank governors give priority to national objectives, as they did before EMU. If, for instance, inflation in Germany and France—thus most of the euro area—is in line with the objective of price stability while in the small countries inflation is increasing, will the governors of the central banks of the latter countries favor a reduction in interest rates? One could conjecture at this stage that the primary objective of the ECB, (i.e., price stability in the euro area) and the independence of the central banks within the Eurosystem should go against this risk. Still economic differences across countries may affect the voting behavior of national central bank governors within the ESCB. This is supported by the experiences of the Federal Open Market Committee (FOMC) of the US Federal Reserve where the voting is influenced by the economic situation in the various Federal Reserve Districts (Meade and Sheets 2002). Likewise Berger and De Haan (2002) have analyzed data on individual voting behavior in the decision-making body of the Bundesbank on discount rate changes in the period 1948 to 1961.3 They find that differences in economic performance translated into different views by the various state central banks on how monetary policy was to be conducted. The probability of a representative not to vote “yes” on a policy change is significantly larger, the larger are the differences between regional and average inflation. The same holds for growth differentials, though at a lower level of significance. The studies of Meade and Sheets (2002) for the United States and of Berger and De Haan (2002) for Germany therefore concur in their findings that economic differences across regions have affected voting behavior in the decision-making bodies of the central banks concerned. Can we be certain that the ECB will behave differently? Or should we worry that ECB decisions might lack the truly European perspective envisioned by the Maastricht treaty?4 And what will happen after the

132

Chapter 5

upcoming enlargement of the monetary union? Sidebar 11 summarizes a model developed by Berger (2002), suggesting that different economic developments may lead to suboptimal policies by the ECB. We illustrate the same point in a very simple way in table 5.5. In the table we show the GDP-weighted and the nonweighted inflation and output growth in the euro area for 2000 and 2001. Price stability in the euro area indicates that the ECB should try to keep weighted inflation below 2 percent in the medium run. If, however, national central bank governors focus on national inflation, the nonweighted inflation figure becomes more relevant. It follows from table 5.5 that there are substantial differences between the weighted and nonweighted inflation and growth figures. If national central bank governors take national interests into account, they will decide differently than when they focus on the euro area as a whole. Whether economic differences may lead to differences of opinion within the Eurosystem on the proper monetary policy stance depends, of course, crucially on the extent to which differences in business cycles, inflation, and monetary transmission in EMU exist and whether they will remain. We will examine these issues in the next sections. Say, for the moment, that inflation and business cycles are very similar, could there still be conflicts in the ECB Governing Council due to differences in preferences? Again, the experience of the Bundesbank can shed some light on this issue. Berger and Woitek (1999) show that dissent voting in the early Bundesbank Council was also correlated with political preferences. Following Vaubel (1997), they categorize Council members according to the political background of their nominating state or federal government.5 They report that left-wing Council members, meaning those members that were nominated by socialdemocratic governments, were significantly inclined to resist (or not to support) interest rate increases. Moreover Berger and Woitek (1999) find a systematic influence of the degree of conservatism in the Bundesbank Council on German monetary policy 1950 to 1994. As a rule more conservative Bundesbank Councils (i.e., Councils with a majority of members appointed by conservative governments) take a more “active” stance in their reaction to shocks. So the German experience suggests that not only divergence in regional economic development but also differences in political preferences seem to influence actual central bank behavior. Following Berger and Woitek, we list in table 5.6 the presidents of the various national central banks in the euro area in 2001, their

Centralization or Decentralization

133

11 A Model of ECB Decision Making Berger (2002) develops a simple Barro-Gordon type model of monetary policy making within an enlarged euro area that illustrates the potential danger of diverging economic developments within the union. Output in the “in” and “out” regions after euro area enlargement develops according to a Lucas supply function: yi = p - p ie + e i ,

i = in,out,

(1)

where y measures output, p (headline) inflation, pe (headline) inflation expectations, and e is a regional output shock with zero mean and known variance. Natural output is normalized to 0. The headline inflation is the sum of inflation in the tradable-goods sector, p˜ , assumed to be under direct and full control of the ECB, and a region-specific structural (exogenous) component, Dp: p i = p˜ + D p ,i , i = in,out. (2) The simplest possible representation of decision making within the ECB Governing Council is a bargaining approach that abstracts from possible strategic interaction among Council members. Along this line the ECB Governing Council sets a tradable-goods inflation in the euro area by weighing the preferred policies of the Board and the national central bank governors: p˜ ECB = bp˜ Board + (1 - b)p˜ NB , (3) with b Œ [0, 1] measuring the political weight or, to be precise, the overall vote share of Executive Board members in the Council and p˜ Board and p˜ NB representing the Board’s and the governor’s preferred tradablegoods inflation rate, respectively. Since the vote shares of the “in” and “out” region can differ, the preferred policy of the national central bank governors is equivalently described as p˜ NB = g p˜ in + (1 - g )p˜ out , (4) where g Œ [0, 1] measures the relative voting power of the “in” region. Thus the overall vote share or bargaining power, with which the interests of all parties involved enter into the ECB decision-making process, is as follows: for the “in” region (1 - b)g, for the “out” region (1 - b)(1 g ), and for the Board b. The “in” region’s central bank’s ideal monetary policy is assumed to proceed from a standard quadratic loss function that increases in the deviations of the region’s headline inflation and output from targets set to 0: Lin = p in2 + lyin2 .

(5)

The relative weight of the real target is l > 0. It is assumed that the “out” region’s central bank takes a more ambitious output target, y*out > 0, while

134

Chapter 5

the weight associated with output losses, l, as well as its inflation target is the same as in the “in” region: 2 Lout = p out + l ( yout - y *out )2 .

(6)

From equations (1), (2), (5), and (6), it is straightforward to derive the preferred policies of both regions. As Berger (2002) shows, the preferred policy of the “out” region differs from that of the “in” region in two respects. First, the more ambitious output target implies a higher rate of inflation. Second, to compensate for the presence of structural inflation, the “out” region’s central bank aims for lower tradable-goods inflation. The Executive Board can be assumed to choose a monetary policy that minimizes the loss function based on economically weighted euro area averages of the headline inflation and output, in line with the policy targets specified in the Maastricht treaty: LBoard = [ cp in + (1 - c )p out ]2 + l[ c yin + (1 - c ) yout ]2 ,

(7)

where c Œ [0, 1] is the economic weight (e.g., the share in aggregate euro area GDP of the “in” region). From equations (1) and (2), it is easy to show that this implies that the higher the inflation in the “out” region is and the higher the region’s impact on euro area inflation, the lower the Board will set tradable-goods inflation to compensate the impact of structural inflation in the “out” region on the overall euro area inflation rate. Berger (2002) shows that under this setup, and assuming rational expectations, equilibrium inflation in the euro area will be p˜ ECB = -b(1 - c )D p ,out + (1 - b)(1 - g )(ly *out - D p ,out ) (8) l {[bc + (1 - b)g ]e in + [b(1 - c ) + (1 - b )(1 - g )]e out }. 1+ l The first part on the right-hand side reflects the Board’s intention to compensate for the “out” region’s structural inflation by lowering tradablegoods inflation, weighted by b, which is the Board’s vote share in the ECB Council. The second element in (8) summarizes the conflicting policy objectives of the “out” region itself. On the one hand, the region’s central bank also aims at decreasing tradable-goods inflation to compensate for its higher structural inflation rate (-Dp,out); on the other, the presence of a more ambitious growth target gives rise to an inflationary bias (ly*out). The net effect of these competing forces on equilibrium inflation depends on the ambitiousness of the “out” region’s growth target, the importance of the output target in its loss function, and the size of its structural inflation problem. Finally, the third element on the righthand side of equation (8) shows that the ECB will shift tradable-goods inflation in the euro area to counter regional output shocks. The weight allocated to either region’s output shock is the sum of the weights attached to output stabilization in the relevant region by the respective central bank (i.e., (1 - b)(1 - g) for the “out” and (1 - b)g for the “in”

Centralization or Decentralization

135

region) and the Board (i.e., bc in case of the “in” and b(1 - c) in case of the “out” region). How does actual monetary policy in the euro area as described in equation (8) deviate from a benchmark policy that would be implemented if all Governing Council members were solely interested in targets specified by the Maastricht treaty? A reasonable benchmark policy rule along this line is one that stabilizes economically weighted euro area averages of output and inflation around zero. The idea is to keep, on average, prices stable and output at its natural level: p˜ * ∫ arg min L* = [cp in + (1 - c )p out ]2 + l[cy in + (1 - c ) y out ]2 (9) s.t. (1) and (2). In words, p˜ * is the rate of tradables inflation that minimizes the loss function L* subject to equations (1) and (2). If monetary policy were conducted in line with this rule, introducing rational expectations would yield equilibrium tradables inflation of l ( ce in + (1 - c )e out ). (10) 1+ l A comparison of the benchmark euro area policy with actual ECB policy as depicted by equation (8) reveals a number of crucial differences: p˜ * = -(1 - c )D p ,out -

Average tradables inflation as described by equation (8) will, in general, deviate from the benchmark (10). This is always true as long as ly*out π Dp,out. In particular, inflation is likely to exceed the inflation rate achieved under the benchmark policy if the inflationary bias introduced by the “out” region’s central bank is high relative to the region’s structural inflation problem, meaning ly*out > Dp,out. Moreover inflation in the euro area will be the higher, the higher is the political weight of the “out” region in the ECB Council. • Actual stabilization policy will, as a rule, not be optimal. In particular, it is straightforward to show that if the “out” region’s political weight in actual ECB decision making deviates from its economic weight and output shocks in the “in” and “out” region do not coincide, the ECB’s stabilization efforts will deviate from the “first-best” benchmark. •

Decreasing the mismatch between economic and political weights can bring actual stabilization policy closer to the benchmark and, if the “out” region was politically overrepresented before, reduce actual inflation.

136

Chapter 5

Table 5.5 Weighted and nonweighted inflation and output growth in the euro area, 2000 and 2001 Country

Inflation in 2000

Inflation in 2001

Output growth in 2000

Output growth in 2001

Austria Belgium Finland France Germany Greece Ireland Luxembourg Italy Netherlands Portugal Spain Nonweighted average GDP weighted average

2.0 2.7 3.0 1.8 2.1 (2.9) 5.3 3.8 2.6 2.3 2.8 3.5 2.9 2.3

2.3 2.4 2.7 1.8 2.4 3.7 4.0 2.4 2.3 5.1 4.4 2.8 3.0 2.5

3.0 4.0 5.6 3.8 3.0 (4.1) 11.5 7.5 2.9 3.3 3.5 4.1 4.7 3.5

1.0 1.0 0.7 1.8 0.6 4.1 5.9 3.5 1.8 1.3 1.7 2.8 2.2 1.5

Source: Own calculations based on data from De Nederlandsche Bank.

Table 5.6 Governors of national central banks and their political backgrounds, 2001 Country

Governor

Political background

Government responsible for appointment

Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain

K. Liebscher G. Quaden M. Vanhala J.-C. Trichet E. Welteke L. D. Papademos M. O’Connell A. Fazio Y. Mersch N. Wellink V. M. Ribeiro Constâncio V. Caruana

C S C C S S ? C C C S C

S+C L+S+ S+C+L C S+ S L S+C C+L S+L S C

Source: Berger and De Haan (2002). Note: S: socialist or social democratic; C: conservative or Christian democratic; L: liberal; + other parties than S, C, or L.

Centralization or Decentralization

137

political background, and the “political color” of the government that appointed the governor. Although the indicator is very crude, it suggests that there are differences in the degree of conservativeness appearing in the euro area.6 A debatable question is whether these differences in preferences are in decline, meaning whether there is convergence in preferences. 5.4 Diverging Business Cycles in the Euro Area? In the previous section we noted that differences in the business cycle in the various EMU countries could affect the voting behavior of national central bank governors in the Governing Council of the ECB.7 Business cycles can differ across countries for various reasons. First, countries experience different shocks. Second, they respond differently to common shocks. This may be caused by differences in the reaction of policy makers to a common shock or because of differences in the national or regional composition of output (Wynne and Koo 2000). Also differences in financial structure may lead to differences in the monetary policy transmission mechanism (see section 5.6). Various studies have pointed out that business cycles in the euro area diverged considerably in the past (e.g., see Christodoulakis et al. 1995). However, no matter how important divergent business cycles were in the past, in assessing dangers for suboptimal monetary policy making by the ECB, the crucial question is how likely it is that business cycles will diverge in the future. Two views have been put forward on this issue. According to a study by the European Commission (Emerson et al. 1992) further economic and monetary integration will lead to less divergence. For one thing, economic policies in the euro area will be more similar. As Lamfalussy (1997)—the first president of the European Monetary Institute, the predecessor of the ECB—puts it: With the emergence of a . . . “stability culture” in the conduct of monetary and fiscal policies, the risk of asymmetric policy reactions during the coming years would appear to me to be much smaller than any time since the end of the last war.

Krugman (1991), on the other hand, has argued that trade integration does lead to regional concentration of industrial activities. In Europe a similar concentration of industries can be expected to take place as in the United States (e.g., Silicon Valley), mainly because of economies of scale and scope. In such a concentration process, sector-specific shocks

138

Chapter 5

can become region-specific shocks, thereby increasing the likelihood of asymmetric shocks and diverging business cycles.8 In the ongoing debate over business cycle synchronization in the euro area, two related issues are being discussed. The first is whether business cycles in the euro area have become more similar. The second is what factors drive business cycle synchronization.9 On the first issue, the literature has not yet reached a consensus whether business cycles of the countries in the euro area have converged.10 In table 5.7 we illustrate this in summarizing the main aspects of some representative studies. The differences between the various studies are explained in part by the use of different data. Other reasons, however, include the use of different methods of identifying business cycles and assessing convergence. As pointed out by Massman and Mitchel (2003), competing methods for the computation of a business cycle have been suggested. One prominent view is that an economic time series can be decomposed into the sum of trend and cyclical components, although there remains disagreement over how the trend should be identified and estimated. An alternative view rejects the concept of trend-cycle decomposition and defines a business cycle in terms of the turning points in the original data series, hence not relying on the estimation and extraction of a trend series. There is also no consensus on how convergence between business cycles should be gauged. Suggestions include looking for increased bivariate correlation of cyclical components, for decreased cyclical disparity, or for evidence of an emerging common factor that drives individual countries’ business cycles. As to the second issue, the various factors considered to affect business cycle synchronization include trade relations (Frankel and Rose 1998), specialization (Imbs 2004a), monetary integration (Fatás 1997), financial relations (Imbs 2004a), and the presence of borders (Clark and van Wincoop 2001). A good illustration of the debate on the degree of business cycle synchronization in the euro area is the controversy between Artis and Zhang (1997, 1999), who conclude that European business cycles have become more synchronized, and Inklaar and De Haan (2001), who find that cycles are better correlated (against Germany) in the period 1971 to 1979 than in the period 1979 to 1987. They argue that this is inconsistent with Artis and Zhang’s (1999) view that increased monetary integration, specifically after the creation of the ERM in 1979, and business cycle synchronization are positively related. According to Artis and

Centralization or Decentralization

139

Table 5.7 Studies on business cycle synchronization in the euro area Study

Data used

Artis and Zhang (1997, 1999)

OECD monthly data of industrial production

Inklaar and De Haan (2001)

Ibid.

Wynne and Koo (2000)

Penn World Tables of GDP, annual data

Döpke (1999)

OECD data of Big 5 euro area countries

Agresti and Mojon (2001)

ECB euro area wide model (AWM) data of GDP and GDP components for 10 countries Harding and ECB AWM data Pagan (2001) of GDP for euro area, OECD data for United States

Massmann and Mitchel (2003)

OECD monthly data of industrial production, 1960:1–2000:8

Measure of cycle

Convergence measure

PAT, HP, linear trend; two subsamples (pre- and post-ERM) HP, two subsamples (pre- and post-ERM) Baxter-King

Lead and lag bivariate correlation with Germany and United States Ibid.

Pairwise correlations, using GMM

HP, linear, segmented trend

Rolling contemporaneous correlations based on 5-year moving average of each country with euro area Baxter-King Contemporaneous and lagged crosscorrelation between each country and the euro area

HardingPagan rule on level series and detrended (linear, HP, PAT) series Various methods

Conclusions Cycles have become more groupspecific after ERM, correlations not different acoss filters after ERM Mixed outcomes, no replication of results of Artis and Zhang (1999) Null of no correlation between EU founding members rejected, but lower correlation with more recent members Correlations among most countries and the euro area increases but that of Belgium falls

Each country highly correlated with euro area as whole, with lowest values for periphery Correlation and Relatively low regression methods correlations among on binary series member countries and euro area

Pairwise correlation coefficients using a method of moments estimator; the entire

Properties of the business cycle depend on how the business cycle is measured. Euro area has often alternated periods

140

Chapter 5

Table 5.7 (continued) Study

Artis (2003)

Darvas and Szapáry (2004)

Data used

IMF Quarterly GDP data for 22 countries, including ten countries in the euro area OECD’s Quarterly National Accounts GDP and components for 10 euro area countries; quarterly data between 1983 and 2002 grouped in four nonoverlapping 5-year periods

Measure of cycle

Convergence measure

of convergence and divergence in the last 40 years

Band-pass filtered series

distribution of all correlation coefficients is analyzed, using rolling windows Stylized facts analysis and cross-correlations

Cycle correlation with euro area, leads/lags, volatility, persistence of the cycle, and a measure of impulse response

Rather strong comovement with the euro area for most EMU members; more synchronization over time according to all the correlation measures calculated, particularly since 1993

HodrickPrescott and bandpass filter

Conclusions

Not much evidence of a European cycle

Source: Update of table provided by Massmann and Mitchel (2003).

Zhang, the ERM forced countries to follow similar monetary policies and has thereby reduced the possibility of following independent monetary policies. The latter are often regarded as a major source of asymmetric (demand) shocks. Not surprisingly, European monetary policy makers welcomed results suggesting that business cycles have become more similar in the euro area. For instance, Trichet (2001, pp. 5–6) states that . . . we can be reasonably confident in the increasing integration of European countries, and in the fact that economic developments are becoming more and more correlated in the area. This has been highlighted, in the academic field, by several empirical investigations: I would mention authors like Artis and Zhang (1999) who found evidence that business cycles are becoming more synchronous across Europe.

Centralization or Decentralization

141

Massman and Mitchel (2003) re-consider the evidence that sparked this controversy between Artis and Zhang and Inklaar and De Haan, using forty years of monthly industrial production data to examine the relationships among the business cycles of the 12 countries in the euro area, employing eight different measures for the business cycle. They compute pairwise correlation coefficients for the 12 countries’ business cycles using a method of moments estimator that also yields an associated measure of uncertainty. To examine the evolution of this estimate over time, Massman and Mitchel use a series of rolling windows, rather than windows of fixed width. Their measure of convergence of business cycles exhibits common features across the alternative measures of the business cycle. Interestingly Massman and Mitchel find that there have been periods of convergence and periods of divergence. The estimated mean correlation coefficient for the 12 European business cycles is on average positive and significant, but there has been considerable volatility. The correlation was trending upward until the mid 1970s and reaching peaks of around 0.8, for most measures of the business cycle. Then correlation, in general, falls to zero in the mid to late 1980s and is statistically insignificant, lending support to Inklaar and De Haan’s (2001) finding that correlations of euro area countries with Germany were higher in 1971 to 1979 than in 1979 to 1987. Correlation then rose in the late 1980s to values in the range 0.6 to 0.8, before slumping quite rapidly in the early 1990s. Since then correlation between the euro area countries has risen but remained volatile; during the mid 1990s correlation even appeared to have declined. The estimates for the most recent period indicate that the correlation for the 12 European cycles has risen from the trough in the early 1990s. Also Darvas and Szapáry (2004) find evidence in support of more business cycle synchronization in the euro area since the run-up period to EMU. These authors draw on GDP data and also on the extent of synchronization in the major expenditure and sectoral components of GDP. Their results suggest that Austria, Belgium, France, Germany, Italy, and the Netherlands show a high degree of synchronization by all measures used. This result applies not only for GDP but also for its components. The synchronization significantly increased between 1993 to 1997 and 1998 to 2002.11 Portugal, Finland, and Ireland show the lowest correlation with the euro area cycle, particularly for consumption and services. It should, however, be pointed out that when correlations are calculated with the euro area, an upward bias is created, since all countries are, by definition, included

142

Chapter 5

in the euro area aggregate. This bias may be substantial for the bigger countries. An alternative approach to assess business cycle similarities is to examine to what extent there is evidence of an emerging common factor that drives individual countries’ business cycles. Artis (2003, p. 2) is a recent example of this approach. In his study he comes to less optimistic conclusions than in his previous work with Zhang: “the European business cycle is a more elusive phenomenon than we might have expected; whilst some European countries seem ‘to stick together’, there are many which do not. In any case, the US and Japan are often to be found as closely associated with those European countries that do stick together as with others.” In line with the conclusions of Massman and Mitchel (2003), Artis finds that there is no coherent movement toward an exclusively “European” cycle. Artis et al. (2002) also find no systematic convergence or divergence tendencies in the classical business cycle (where turning points are identified on the basis of an absolute decline, or rise, in the value of GDP). The level of dispersion of the member countries’ growth rates peaked in the 1970s, declined until 1985, and creeped up in 1990 to 1995 and again in 1999 to 2000. Dispersion was relatively small during the periods 1985 to 1990 and 1995 to 1998. The volatility of growth in the last 15 years has been lower than in the period 1970 to 1985. When Artis et al. (2002) use the deviation cycle (whose turning points are defined as points where deviations occur in the rate of growth of GDP from an appropriately defined trend rate of growth), they find stronger evidence in favor of convergence, since there is a systematic tendency for cross-sectional dispersion to decrease over time. Artis et al. (1999) use univariate Markov switching autoregressions (MS-AR) for individual countries to detect changes in the mean growth rate of industrial production and a Markow switching vector autoregression model (MS-VAR) to identify a common cycle in Europe. They see evidence of a common unobserved component governing the business cycle dynamics in Europe, such as might suggest the existence of a common business cycle. Similarly Forni and Reichlin (2001) show that the business cycle of European regions can be decomposed into a European component, a national component, and a regional component, with the European component having a larger role than the national components. The share of the European regions GDP variance that is explained by the common European business cycle ranges between 40

Centralization or Decentralization

143

and 60 percent for most countries of the euro area (Portugal and Greece being the exception) while the share of the national components ranges between 20 and 35 percent. The rest of the variance is driven by the regions’ idiosyncratic components.12 Last, Kaufman (2003) uses time series on industrial production growth of individual countries to investigate whether there is a common growth cycle for the euro area countries and whether the synchronization has changed over time. She uses an autoregressive panel data framework and Bayesian simulation methods. Her results support the idea of increasing synchronization among the euro area countries due to the integration process. One of the factors that may affect business cycle synchronization is trade intensity. Theoretically trade intensity has an ambiguous effect on synchronization. Standard trade theory predicts that openness to trade leads to increased specialization in production and interindustry patterns of international trade. If business cycles are dominated by industry-specific shocks, higher trade integration, by bringing about more specialization, can lead to decreasing business cycle correlations. Furthermore the higher integration in both international financial markets and goods markets cushions any asymmetric shocks to contries because of the diversification of ownership and specialized production structures (Kalemli-Ozcan et al. 2003). On the other hand, if business is dominated by intra-industry trade, deeper trade links will not necessarily result in deeper specialization along industry lines. In that case industry-specific shocks can lead to more symmetric business cycles. Intensive trade relations between countries may also lead to the export or import of a business cycle caused by demand fluctuations, as changes in income in one country will normally also lead to a changed demand for foreign goods. Despite this theoretical ambiguity, starting with Frankel and Rose (1998), many studies have found support that trade intensity will, in general, enhance business cycle synchronization. Most studies examining this issue find a positive association in the trade between countries and their business cycle synchronization, regardless of the way in which the trade relationship is modeled. However, more recent studies tend to find somewhat lower effects than those reported by Frankel and Rose (1998). For instance, Gruben et al. (2002), using the same group of 21 countries as Frankel and Rose, confirm their general conclusion, that increased trade leads to increased business cycle correlation, but

144

Chapter 5

find the trade effect on business cycle correlation to be about half the Frankel and Rose point estimate (see section 6.3.1 for further references).13 Gruben et al. (2002) also find that increases in interindustry trade— which may indicate specialization—turn out not to have a significant effect on business cycle synchronicity. As pointed out by Imbs (2004a), specialization is likely to affect the international synchronization of business cycles. Two economies producing the same types of goods will be subjected to similar stochastic developments in case of sectorspecific shocks. Countries with similar production patterns will also react similarly to aggregate shocks. In contrast to Gruben et al. (2002), Imbs (2004a) finds that similarities in economic structure result in correlated business cycles. In other words, specialization reduces business cycle synchronization. However, the evidence reported by Otto et al. (2001) is not in line with this view. There is little agreement whether monetary integration will lead to more similar business cycles. An argument can be made in both directions. As suggested by Lamfalussy (1997), monetary integration may cause more similarity, since there will be less asymmetry in monetary policy. Also indirectly monetary integration may lead to more synchronization via the impact of exchange rate stability on trade relations. In a drastic departure from past empirical studies that failed to find a significant link between exchange rate stability and trade, Rose (2000) reports extremely large positive effects of common currencies on the volume of trade. The most dramatic, and widely cited, of his findings is that “two countries sharing the same currency trade three times as much as they would with different currencies” (Rose 2000, p. 7). Rose’s estimates of the trade-creating effects of adopting common currencies are obtained using a “gravity model” in which bilateral trade is a function of the relative economic size and distance between two trading partners. Rose also includes several variables that are intended to capture trading partners’ cultural and historical links and membership in regional trading blocs. The key feature of Rose’s gravity model is the inclusion of two monetary variables: a dummy variable to indicate whether trading partners use the same currency and a measure of the volatility of the bilateral exchange rate. Glick and Rose (2002) use a much larger data set and find that a common currency doubles trade.14 Other studies, however, arrive at considerably lower effects (see section 6.3.1 for a further discussion).

Centralization or Decentralization

145

Monetary integration may also lead to less business cycle synchronization. To the extent that exchange rates are considered as a shock-absorbing mechanism, fixed exchange rates may lead to less synchronization if the countries in the monetary union face asymmetric shocks. In face of an external shock, a fixed exchange rate regime requires the central bank to follow a policy so as to maintain the peg, forcing all the adjustment to take place in the real economy rather than in the exchange rate. According to Artis and Zhang (1997), business cycles in Europe were more similar after the start of the ERM than before, which they interpret as evidence that monetary integration will enhance business cycle synchronization. Other studies report less support for the view that exchange rate stability in Europe has led to more synchronization of business cycles. Frankel and Rose (1998), for instance, include an ERM dummy variable in their empirical model and find that its coefficient is not robust. Also Baxter and Stockman (1989) and Inklaar and De Haan (2001) report no effect of exchange rate stability on business cycle synchronization. Furthermore, possible evidence that since the run-up to EMU there is more business cycle synchronization may not reflect the effect of monetary integration. As pointed out by Darvas and Szapáry (2004), the business cycles of non–EMU European countries and the United States, and to some extent Japan and Russia, have also shown greater co-movement with the business cycle in the euro area. De Haan et al. (2002) have analyzed whether trade relations and exchange rate stability have affected the synchronization of business cycles in 18 OECD countries over the years 1961 to 1997.15 To calculate the correlation between business cycles they first detrended the series with a Hodrick-Prescott filter and calculate correlations of cyclical deviations for the period 1961 to 1997. These authors have estimated the following simple model: r ij,t = a 0,t + a 1 (log wtij,t ) + a 2 vij,t + m ij,t ,

(5.1)

where pij,t denotes the correlation of the business cycles of countries i and j during the period t. Dummies (a0,t) are included for the four subperiods (1961–73, 1973–79, 1979–87, and 1987–97). The definition of trade intensity (wtijt) is the same as in Frankel and Rose (1998): wtijt =

Xijt + Mijt , X i . t + X j . t + Mi . t + M j . t

(5.2)

146

Chapter 5

or the total trade (export X plus import M) between countries i and j divided by the sum of the trade of each country. The exchange rate volatility is measured as the standard deviation of the growth rate of the exchange rate. Their results, shown in table 5.8, suggest that both higher trade intensity and more exchange rate volatility are related to synchronization of business cycles. The fact that more exchange rate volatility leads to more business cycle synchronization may be interpreted as support for the view that exchange rates may function as an adjustment tool. This conclusion is confirmed by the negative sign of the “Bretton-Woods” dummy and the positive value of the “floating period” dummy. These results suggest that exchange rate fluctuations may reduce asymmetries between countries. Also financial integration has been argued to affect business cycle synchronization. As pointed out by Imbs (2004a), the impact of financial integration on synchronization is not unambiguous. On the one hand, a limited ability to borrow and lend internationally hampers the transfer of resources across countries and can increase GDP correlations. If, on the other hand, investors have imperfect information or face liquidity constraints, limiting capital flows can decrease GDP correlations, as investors herd (or withdraw) capital from many destinations simultaneously. Financial integration may also lead to specialization, and therefore less synchronicity. Using a variety of alternative measures of Table 5.8 Estimates of the relationship between business cycle correlation and trade intensity and exchange rate volatility

Constant Dummy 1960–1973 Dummy 1973–1979 Dummy 1979–1987 Trade intensity Exchange rate volatility

(1)

(2)

(3)

44.4 (8.45) -12.3 (-3.24) 17.2 (4.02) 0.8 (0.25) 2.51 (2.51)

44.6 (4.28) -10.0 (-2.48) 16.7 (3.95) 0.6 (0.18)

68.7 (5.33) -7.6 (-1.85) 16.9 (3.97) 0.9 (0.26) 3.03 (2.87) 4.50 (2.15)

2.52 (1.23)

Source: De Haan et al. (2002). Note: t-statistics in parentheses, n = 612; all parameters are multiplied by 100.

Centralization or Decentralization

147

financial integration, Imbs (2004a) finds evidence that economic regions with strong financial links are significantly more synchronized. Interestingly Imbs (2004a) obtains similar estimates across US states and across countries with substantially different monetary policies, suggesting that the synchronization of business cycles is not an artifact of an international convergence of monetary policy making. Imbs concludes that the positive direct effect of finance on synchronization dominates the negative indirect effect, working via higher specialization. Using some new indicators for financial integration (including bilateral asset holdings), Imbs (2004b) reports that apart from an influence on goods trade, financial liberalization has a direct positive effect on business cycle synchronization. Jansen and Stokman (2003) find that capital flows have played a role in synchronizing business cycles in recent years. For a sample of industrial countries in the period 1995 to 2001, they report that FDI linkages can explain the pattern of international business cycle linkages even better than foreign trade relations. Finally, Clark and van Wincoop (2001) suggest that the presence of borders matters when it comes to business cycle synchronization. They find for the four largest European countries that within-country correlations are substantially larger than cross-country correlations. These results continue to hold after controlling for exogenous factors such as distance and size. They find that the lower level of trade between European countries, and, to a lesser extent, the higher degree of sectoral specialization, can explain most of the observed border effect. Artis (2003) also examines which factors affect business cycle affiliation, departing from the idea that business cycles are the result of originating shocks feeding into a propagation mechanism. Asynchronous cycles may therefore arise from the interraction of different (nonsymmetric) propagation mechanisms with common shocks just as much as they arise from asymmetric originating shocks with similar kinds of propagation mechanisms at work. For instance, more or less flexible labor markets will make for less or more persistence in the response to a shock. On the basis of panel data estimation using averages over three periods (1970–79, 1980–92, and 1993–2001) Artis (2003) concludes that the variables “relative financial structure” (measured as the ratio of private credit to stock market value traded) and “relative share of oil imports” are significant with a negative sign. Relative labor market flexibility is not significant, while the role of trade is more ambiguous.

148

Chapter 5

On the basis of the foregoing analysis, we conclude that there is still no consensus on whether business cycles of the countries in the euro area have become more similar recently. This is due to the use of different data and methods of identifying business cycles and assessing convergence. There are various factors that may affect business cycle synchronization, such as trade relations, specialization, monetary integration, financial relations, and the presence of borders. There is considerable support for the notion that trade intensity enhances business cycle synchronization, but the effect of trade specialization on business cycle synchronization is less accepted. There is little agreement whether monetary integration and financial integration will lead to more similar business cycles. The presence of borders is likely to matter for business cycle synchronization. 5.5 Different Inflation Rates in the Euro Area Inflation differentials in the euro area have increased recently. As pointed out by Remsperger (2003), in the convergence and start-up phase of EMU, the average annual rate of inflation in Austria, France, and Germany was 0.8 percent. At that time the critical 2.3 percent mark (1.5 percent above the inflation in the three best-performing countries) was not being overshot by any country. Between 2000 and 2002 Austria, France, and Germany again had the lowest inflation in the euro area, although their average inflation rate was now 1.8 percent. Together with the margin allowed by the inflation criterion, this produced a critical value of 3.3 percent. The inflation in no fewer than three founding members of the monetary union—Ireland, the Netherlands, and Portugal—exceeded that level. Inflation differentials in the euro area since the beginning of 1999 have been quite marked. Notably Ireland, Spain, Greece, and Portugal have been persistently at the top of the inflation league table; also price increases in the Netherlands were generally higher than average inflation in the euro area. In contrast, German inflation has been below the euro area average (see table 5.9).16 Viewed over a longer time period, the degree of inflation dispersion in the euro area, measured in terms of the standard deviation, decreased over time, especially during the second half of the 1990s. The unweighted standard deviation declined from around 4 percentage points at the beginning of the 1990s to about 1 percentage point at the start of the monetary union (ECB 2003b).

Centralization or Decentralization

149

Table 5.9 Inflation differentials relative to the euro average, 1990–2002

Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain

1990–1993

1994–1998

1999–2002

1999

2000

2001

2002

-0.8 -1.1 0.3 -1.1 -0.6 12.9 -1.6 1.6 -0.6 -1.4 6.0 1.9

-0.4 -0.4 -0.9 -0.5 -0.6 5.4 0.3 1.5 -0.6 -0.3 1.2 1.3

-0.4 -0.1 0.2 -0.5 -0.6 1.1 2.1 0.3 0.3 1.3 1.2 1.2

-0.6 0.0 0.2 -0.6 -0.5 1.0 1.3 0.5 -0.1 0.9 1.0 1.1

-0.4 0.4 0.6 -0.5 -0.7 0.6 2.9 0.3 1.5 0.0 0.5 1.2

-0.2 0.0 0.2 -0.7 -0.5 1.2 1.5 0.2 -0.1 2.6 1.9 1.2

-0.5 -0.7 -0.2 -0.3 -0.9 1.7 2.5 0.4 -0.2 1.7 1.4 1.4

Source: ECB (2003b).

According to the ECB (2003b), the current degree of inflation dispersion in the euro area is close to that observed in the United States. However, as we pointed out in chapter 3, owner-occupied housing is included in the US CPI but not in the European HICP. As argued by Remsperger (2003), this is important because there is hardly any other component where regionally divergent price developments have such a major impact as they do in housing. The interregional standard deviation of the changes in rents is greater than that of services, which, in turn, is larger than that of industrial goods (excluding energy). House prices increased by 10 percent a year and more in Ireland, the Netherlands, and Spain while less in other countries. Thus, if owneroccupied housing were included in the HICP, the divergence of inflation rates in the euro would take a sharp rise, perhaps even exceed the dispersion of inflation rates in the United States. Also the inflation differentials in the euro area are more persistent than those in the United States. Inflation differentials are caused by many factors, including different consumption patterns, fiscal policies, differences in the dependence on trade, convergence of price levels of traded goods across euro area countries, different productivity developments, and diverging business cycles. We will discuss each factor in turn. Differences in inflation rates in the monetary union due to consumption patterns may be attributed to differences in the shares of

150

Chapter 5

individual consumer goods and services in total consumption across countries. The weights used in calculating HICP in the separate countries are only harmonized; they do not reflect the actual consumption baskets of these countries. So there are differences between the calculated inflation rates and the actual inflation rates. However, research by the ECB (2003b) indicates that differences in consumption patterns across the euro area countries do not have a major impact on inflation dispersion. Also the prices of the same goods may show different developments across countries as they may be administered. Preliminary estimates by the ECB (2003b) suggest nevertheless that the magnitude of the impact of administered prices on inflation dispersion in the euro area has been relatively small in the period 1999 to 2002. Countries in the euro area may experience different inflation rates due to differences in their fiscal policies. Although the scope of national fiscal policy is limited by the Stability and Growth Pact, it can put pressure on inflation. According to Hendrikx and Chapple (2002), for instance, VAT and energy tax increases in 2001 led to an increase of the inflation rate in the Netherlands of approximately 1 percentage point. These indirect tax effects should be temporary as they represent a oneoff change in the price level. In their empirical estimates of inflation differentials in the euro area, Honahan and Lane (2003) did not find much evidence that national fiscal policy differences contributed to the inflation dispersion. Likewise differences in dependence on trade may lead to inflation differentials. The nominal effective exchange rate of the euro had declined by about 20 percent in October 2000, before recovering in the period from 2001 to 2003. The impact of exchange rate shocks on consumer prices is determined by the share of imported goods in private consumption and the strength of the pass-through effect. As pointed out in section 3.6.1, the pass-through is not uniform across the countries in the euro area. Price developments may therefore differ in the short to medium term even with uniform shocks. One important factor influencing pass-through is the openness toward trading partners outside the euro area. In general, “extraopenness” should be reflected by a higher weight of extra–euro area goods in a country’s overall goods basket and, therefore, a stronger pass-through effect from exchange rate changes on domestic prices (ECB 2003b). A member country that depends on imports from a nonmember country will experience different inflationary pressures if the euro exchange rate depreciates as compared to a member country that

Centralization or Decentralization

151

conducts all its trade with other member countries (Hüfner and Schröder 2002). In addition to the direct impact of exchange rates on consumer prices, there may be a cyclical effect of exchange rate shocks via a change in the competitive position, which is also determined by an economy’s degree of openness to countries outside the euro area. The econometric estimates of the inflation differentials of Honahan and Lane (2003) suggest a significant role for effective exchange rate changes. Their point estimate implies that a relative deprecation of 3.5 percent is associated with an additional 1 percentage point of inflation. This is a large effect. The Irish nominal effective exchange rate depreciated 11 percent during 1998 to 2000, while the French exchange rate weakened by only 4 percent. According to these estimates, differences in the effective exchange rates of Ireland and France led to an inflation differential of 2 percentage points in this period. Most of the differences in the effective exchange rate across countries in the euro area reflect differences in their trade patterns. However, as pointed out by the ECB (2003b), euro area countries with similar degrees of “extra-openness” had very different inflation rates, indicating that other factors are also important for the divergence of inflation in the euro area. One such factor is also related to trade: differences in oil dependency. Oil prices tripled in 1999 and 2000. Since countries in the euro area differ in terms of oil dependency, oil price changes may also have led to inflation differentials. However, research by the ECB (2003b) suggests that the relationship between oil dependency and inflation differentials is rather weak. Countries in a monetary union share the same currency but need not have the same price level. If price levels differ initially across countries forming a monetary union, then price level convergence will generate temporary inflation differentials (Duarte 2003). Existing research for the United States suggests that price level differences can be persistent. For instance, using consumer price data for nineteen US cities from 1918 to 1995, Cecchetti et al. (2002) report that price level differences were large and persisted by as much as 1.55 percentage points in annual inflation rates measured over ten-year periods. Similarly Parsley and Wei (1996) employ commodity level price data for 48 US cities from 1975 to 1992 and find persistent deviations from the law of one price for both traded and nontraded goods. In the European Union, price levels differ across countries. According to the European Commission (2002b), indirect taxation, the structure of distribution networks, market power or competition, and

152

Chapter 5

inefficient services sectors are the main factors accounting for much of the remaining differences in the prices of tradable goods. At least some convergence of price levels of traded goods has already occurred in the euro area. Increased market integration and price transparency associated with the adoption of a common currency will further reduce the scope for deviations from the law of one price. Rogers (2001) has examined price level convergence, using a unique data set of prices of 168 goods and services across 26 European cities in 18 countries between 1990 and 1999. There was evidence of convergence for traded goods in the first half of the 1990s but none of convergence for nontradables over the decade. Table 5.10 is reproduced from this study. The empirical estimates by Rogers (2001) for inflation differences across Europe show that countries with low prices in 1999 experienced higher overall inflation in 2000. Rogers also obtained data for US cities, which provides a useful comparison, since the United States has long been a functioning monetary union. Rogers found little evidence of an active process toward price convergence across US cities. More recently Honahan and Lane (2003) find that a considerable part of the inflation differentials in the euro area can be explained by price level convergence. Their point estimate indicates that a country with a price level one-third below the European average will experience an additional 1 percentage point of inflation.17 Thus inflation of traded goods prices and nontraded goods prices, or both, can bring relatively high inflation to countries with low prices following economic integration. Since price level convergence works mostly through tradables, some of the current divergence of inflation Table 5.10 How much have prices in the euro area converged? Standard deviation of prices across locations Price index Euro Area Overall Tradables Nontradables United States Overall Tradables Nontradables Source: Rogers (2001).

1990

1995

1999

0.12 0.12 0.27

0.12 0.08 0.33

0.11 0.06 0.31

0.16 0.05 0.51

0.15 0.04 0.52

0.17 0.04 0.57

Centralization or Decentralization

153

in the euro area may be transitory. However, to the extent that price level convergence occurs through the relatively gradual process of convergence of productivity and living standards, the resulting crosscountry inflation differentials may be long-lived (see Rogers 2001). The Balassa-Samuelson effect (Balassa 1964; Samuelson 1964) suggests that prices of nontraded goods may converge as well. In a convergence of productivity levels, the initial low-productivity low-price level countries will experience faster productivity growth and price rises compared to the high-productivity high-price level countries. To show this mathematically, we take the simple model of a two-country world, both countries operating with a small open economy, two commodities, one scarce factor (labor), and constant input coefficient technology. We assume that the prices of nontradables are not equalized across countries. Another crucial assumption is that the growth rate of real wages in the traded goods sector is determined by the growth rate of labor productivity. So w T - p T = aT ,

(5.3)

where a, p, and w stand for changes in productivity, prices, and nominal wages, respectively, while the superscripts T and NT refer to the tradable and nontradable sector. Inflation in tradables is the same due to international trade arbitrage: p T = p T * +e ,

(5.4)

where the asterisk denotes the foreign country and e denotes the change in the exchange rate. It is assumed that wage increases in both sectors are the same so that a NT + p NT = a T + p T ,

(5.5)

Say that over time the productivity growth in the tradable goods sector of a country exceeds that in the nontradable goods sector. Clearly, higher productivity growth in the traded goods sector will entail higher wages in both sectors. Let inflation be defined as the weighted sum of the tradables’ and nontradables’ price inflation, with a and (1 - a) the traded goods and nontraded goods sector’s share in GDP: p = ap T + (1 - a )p NT

(5.6)

Now assume that a ª a*. The inflation differential between the home and foreign country can then be written as

154

Chapter 5

p - p * = e + a [(a T - a NT ) - (a T * - a NT *)].

(5.7)

We see in (5.7) that the difference between the inflation rates equals the rate of depreciation of the nominal exchange rate plus the (common) share of nontraded goods in the consumption basket, multiplied by the excess of the productivity growth differential between the traded and nontraded goods sectors in comparison to this productivity differential abroad. The high inflation that results from rapid productivity growth in the traded good sector is nothing more than an equilibrating mechanism. Without this equilibrating mechanism the country with high productivity growth would gain competitiveness and accumulate current account surpluses. The Balassa-Samuelson effect can explain inflation differentials in a monetary union if the countries in this union show large difference in income per capita. The euro area includes many economies in which GDP per head is considerably below the EU average. In these so-called catching-up countries, productivity, especially in manufacturing, is likely to grow faster than in the other EMU countries. The resulting sector price increases influence relative prices and push inflation to levels above those recorded in more advanced economies. According to De Grauwe and Skudelny (2000), the prices of nontraded goods grew faster than the prices of traded goods in all EMU countries because labor productivity growth in the tradable goods sector exceeded labor productivity growth in the nontradable goods sector. Figure 5.2 shows the average productivity growth over the period 1971 to 1995 as reported by these authors. 5 T NT

4 3 2 1 0 A

BL

DK

FIN

FR

GE

IT

NL

PO

SP

SW

UK

Figure 5.2 Average productivity growth in European countries (%), 1971 to 1995. (Source: De Grauwe and Skudelny 2000)

Centralization or Decentralization

155

Co-integration tests for 11 EU countries for the period 1975 to 1995 by Alberola and Tyrväinen (1998) suggest that sustained inflation differentials of 2 percentage points exist between the more and less advanced euro countries due to the Balassa-Samuelson effect. However, De Grauwe and Skudelny (2000) find in their panel estimates for the period 1970 to 1995 for 12 EU countries that the average bilateral inflation differential caused by the Balassa-Samuelson effect generally does not exceeded 1 percentage point. These authors also estimated the maximum bilateral inflation differentials due to the Balassa-Samuelson effect. The calculations showed that the inflation differential can go up to 8 percent in absolute value. Sinn and Reutter (2001) also estimated the contribution of the Balassa-Samuelson effect to inflation differentials, but they found for most countries a smaller impact. To construct a complete sample for all members of the currency union, they had to restrict the tradable goods sector to agriculture and manufacturing and treat the remainder as the nontradable goods sector (construction, services, electricity, gas and water supply). Inflation rates were higher in countries where the intersectoral growth difference was higher (see equation 5.7). To avoid deflation in any one of the countries, the aggregate inflation rate for all countries taken together must rise as countries become more diverse in their productivity differentials. The following equation serves as the basis for the empirical estimates of Sinn and Reutter (2001): n

p min = Â b ia iNT (aiT - aiNT ) - min [a iNT (aiT - aiNT )]. i =1

i =1,...,n

(5.8)

Here p min is the minimum aggregate inflation rate compatible with the requirement that no single country will face a deflation, bi is country i’s share in aggregate value added, a iNT is the share in value added produced by country i’s sector of nontraded goods and (aTi - a iNT) is the change in the marginal product of labor in the traded goods sector minus the change in the marginal product of labor in the nontraded goods sectors of country i. Equation (5.8) can be explained as follows: Imagine a fully developed industrial country, such as Germany, whose manufacturing sector enjoys no particular productivity gains relative to the nontraded goods sector (no change in relative prices) and the relative prices of nontradable goods is rising in all other countries. The price of traded goods cannot fall without generating a deflation in this industrially mature

156

Chapter 5

country. The aggregate price level must rise so as to accommodate for the required increases in the relative prices of nontraded goods in the other countries that result from the productivity increases in the traded good sectors. This is covered by the first term on the right-hand side of the equation. To understand the second term on the right-hand side of the equation, consider that even the most industrially mature country can experience some extra productivity growth in the traded goods sector. The price of traded goods can fall in this country without effecting a deflation because of compensation from the increase in the price of nontraded goods. As a consequence the no-deflation constraint is relaxed. Because the price of traded goods is falling, the relative price of nontraded goods can increase in all countries without implying as much inflation as before (Sinn and Reutter 2001). As pointed out by the ECB (2003b), it is difficult to isolate the BalassaSamuelson effect from other historical influences on inflation, in particular, differences in monetary and exchange rate policies across countries. There is actually a large spectrum of estimates of BalassaSamuelson effects and contradicting results for individual countries. In table 5.11 we present a summary of studies on the importance of the Balassa-Samuelson effect in the euro area (recalculated for euro area inflation of 2 percent), including the studies that are discussed here. Although these studies are not all directly comparable because of differences in methodology and sample periods and because of a wide variety of results, some broad patterns are discernible. For instance, Germany and France are generally below the average, while Greece and Ireland are above the average. This result is broadly consistent with the fact that catching-up countries can expect to experience a real appreciation. Nevertheless, there is the slim chance, as pointed out by the ECB (2003b), historically, that their catching-up will not in every case lead to high inflation or an appreciating nominal exchange rate, as is the case of Ireland. Indeed, as Honahan and Lane (2003) convincingly argue, little, if any, of the Irish inflation deviation is a reflection of the Balassa-Samuelson effect. Ireland’s boom has been apparently largely due to employment growth, and not exceptional productivity gains. Furthermore some of the estimates of the Balassa-Samuelson effect are not in line with actual inflation after the start of the monetary union. For instance, Belgium and Finland are sometimes found to have high Balassa-Samuelson effects, but this is not confirmed by actual inflation differences. Conversely, the Netherlands has had a higher inflation differential than predicted by the Balassa-Samuelson model (ECB 2003b).

Sample

Alberola and Tyrväinen 1975–1995

HICP Proxy IMFa

IMF (1999) 1960–1996

Canzoneri et al. 1973–1997

De Grauwe and Skudelny

Sinn and Reuterb 1987–1995

Average HICP inflation 1995–2002c

Belgium Germany Greece Spain France Ireland Italy Netherlands Austria Portugal Finland Euro area Standard deviation

3.1 1.3 — 3.1 1.7 — 2.4 2.3 1.8 — — 2.0 0.6

2.0 1.9 2.7 2.3 1.9 3.4 1.9 2.3 2.5 2.7 2.3 2.0 0.4

3.8 1.5 2.8 — 2.8 3.0 2.7 1.6 — 4.3 2.9 2.0 0.9

2.6 1.0 — 2.4 2.4 — 2.8 — 1.8 — 2.6 2.0 0.6

2.1 1.7 — 2.0 1.6 — 2.4 2.0 2.5 2.1 1.4 2.0 0.4

1.8 1.0 5.3 2.5 2.3 3.4 2.5 2.4 2.4 1.8 3.7 2.0 1.1

1.7 1.2 3.8 3.0 1.5 3.1 2.8 2.5 1.5 3.0 1.6 1.9 0.9

157

Source: ECB (2003). Note: Euro area inflation normalized to 2 percent. a. The IMF (2002) calculates an HICP proxy, which assumes that the historical trend differential between price developments of industrial goods and services between 1995 and 2001 remains the same. This measure is immune to some of the criticism of the other BS studies as it relies directly on observed inflation rather than on productivity differentials. However this analysis also entails an important caveat in that it is based on a short period, which does not comprise a complete business cycle, and therefore may be biased. b. Sinn and Reutter (2001) assume that historical productivity differentials will be reflected in equally large inflation differentials between sectors. Most other studies do not find a unitary relationship, implying that the dispersion found by Sinn and Reutter is likely to be upward biased. c. Greece since 1997.

Centralization or Decentralization

Table 5.11 “Equilibrium” inflation rates implied by Balassa-Samuelson effect according to selected studies

158

Chapter 5

The Balassa-Samuelson hypothesis is based on the assumption that PPP holds in the long run in the traded goods sector. Furthermore relative prices of the nontraded and traded sectors are assumed to mainly depend on the different evolution of productivity in the two sectors. Ortega (2003) has investigated whether such long-run hypotheses can be accepted across the main European economies (Germany, France, Italy, and Spain). The sample period covers 1970 to 2001 for Italy and Germany, 1978 to 1999 for France, and 1986 to 2001 for Spain. She finds that deviations from PPP are persistent for the sample periods under consideration, although somewhat less than is typically found in the literature. Ortega splits up the real exchange rate into that of the traded sector and the differential across countries of the relative price of the nontraded sector. The latter is further broken into its determinants, namely the cross-country differentials of relative markups, relative labor costs, and relative labor productivities. Ortega finds that the driving forces of the relative price differentials differ across countries and periods. They persistently diverge from the predictions of the Balassa-Samuelson hypothesis. Business cycle differences may also contribute to the inflation differentials in the euro area. As we pointed out in the previous section, business cycles diverge among the countries in the euro area. Countries with output above trend tend to have upward pressure on inflation, while countries with output below trend will experience downward pressure on inflation. Figure 5.3 plots the average inflation rate after the euro was adopted in each euro area country against its average output gap in the same period. The plot suggests a positive relationship between the average output gap and average inflation. Business cycles can be out of sync for various reasons. Remsperger (2003) argues that one reason could have been the nominal convergence process in the run-up to EMU. The elimination of the residual foreign exchange risk since the beginning of 1999 and the dwindling of the risk premia brought about a largely uniform interest rate level on capital markets. In some countries this could have generated a substantial economic boost. It was exactly in those countries that already had relatively high inflation that the fall in real interest rates pushed up prices.18 An important factor was the rapid rise in property prices encouraged by the convergence of interest rates (Remsperger 2003). Indeed, Honahan and Lane (2003) report a fairly strong negative cross-sectional correlation between real interest rate declines in the run-up to EMU and commercial property inflation in 1995 to 2001 (the correlation is

Centralization or Decentralization

159

4.5 4 Average inflation

3.5 3 2.5 2 1.5 1 0.5 0

-2

-1

0

1

2

3

4

5

Average output gaps Figure 5.3 Output gap and inflation, 1999 to 2002. (Source: OECD, Economic Outlook)

-0.67). This source of inflation differentials is only temporary. The expansionary effects of real interest rate changes over time are offset by the equilibrating effect of changes in national competitiveness triggered by an increase in inflation differentials. Inflation differentials among economies in a monetary union lead to changes in their real exchange rates. Owing to a real “appreciation,” countries with higher-than-average inflation rates suffer a loss in price competitiveness, while countries with relatively low inflation rates gain in price competitiveness. The consequence is that export demand in the countries with higher inflation rates tends to decline, which has a dampening effect on prices in these countries. Conversely, demand tends to increase in countries with lower inflation rates. The coolant effect of real appreciation through a loss of competitiveness is likely only to operate at a gradual pace. This persistence mechanism is reinforced if the wagesetting process is not perfectly flexible, such that current rather than prospective inflation influences wage determination (Honahan and Lane 2003). Arnold and Kool (2002) have examined the regional inflation dispersion in the United States. They find that the pro-cyclical impact of a lower real interest rate dominates the countercyclical impact

160

Chapter 5

of the real exchange rate in the short term. So, when regional inflation differentials occur, initially they tend to increase. In conclusion, like in other monetary unions, inflation differentials exist and are likely to remain in place in the European Monetary Union. From our review of the literature we find that despite the popularity of the Balassa-Samuelson argument, the evidence in support of it is not very compelling. Other factors, like different reactions to the development of the external value of the euro, price level convergence, and diverging business cycles seem to be more important. Regional inflation dispersion appears to be an adjustment mechanism through which regional economic imbalances are corrected. Inflation differentials can, however, become problematic in two instances. First, if the composition of the group of countries with above average inflation remains constant, the political support for monetary union and the ECB’s policies will likely erode in these countries. This can put pressure at national central bankers. Second, if an inflation differential in a particular country is transitory, it can be a potentially dangerous trigger of persistence mechanisms that continue to operate after the original shock has disappeared. Overshooting can occur through price–wage dynamics, especially if current inflation feeds into future wage growth (Honahan and Lane 2003). There may be differences in the way expectations are formed across the euro area countries. According to Benigno and Lopez-Salido (2002) the expectation formation in the wage- and price-setting process in Germany has been forwardlooking, whereas in other countries it has been geared more to the past. This can have a significant impact on price dynamics and give rise to inflation differentials as pointed out by Remsperger (2003). For example, an oil price shock that affects wages in one country because its expectation formation is oriented to the past does not necessarily affect wages in other countries, so temporarily this situation creates inflation differentials. 5.6 Differences in Monetary Policy Transmission and Financial Structure It may be difficult for the ECB to keep inflation in the medium term below 2 percent if the impacts of monetary policy decisions differ across countries in the euro area. Many critics of EMU have considered the differences in monetary transmission across countries in the euro area as an (additional) argument against a common currency.

Centralization or Decentralization

161

In this section we examine two aspects of monetary transmission. First, we review in section 5.6.1 recent research on the so-called passthrough of monetary policy decisions (the extent to which changes in monetary policy interest rates are reflected in market interest rates). Among other things, the impact of monetary policy on the real economy depends on how changes in policy rates are transmitted to market interest rates. Two elements are crucial for the transmission of monetary policy decisions: the degree to which changes in the policy rate affect the cost of borrowing and the speed of adjustment of market rates to changes in policy interest rates. We analyze to what extent the passthrough of policy interest rates differs across countries in the euro area and whether there is convergence of the pass-through. Second, we examine recent studies in which asymmetries in monetary policy transmission are related to differences in financial structure. The best-known study in this area is by Cecchetti (1999, p. 22), who argues: Most economists believe that the monetary transmission mechanism will vary systematically across countries with differences in the size, concentration, and health of the banking system, and with differences in the availability of primary capital market financing. The countries of the EU differ quite dramatically in all of these dimensions that would seem to matter, leading to the prediction that the impact of interest rates on output and prices will not be consistent across countries. While the estimates of the impact of interest rate changes on output and inflation tend to be quite imprecise, they do differ, and in the way that is predicted by the state of the countries’ financial systems.

In contrast, in summarizing a large project on differences in monetary transmission across euro area countries, the ECB concludes in its Monthly Bulletin of October 2002 that the empirical evidence does not suggest that there are systematic differences among countries in policy transmission that are robust across different studies and methodologies. In section 5.6.2 we analyze this issue in some detail. 5.6.1 Pass-through of Monetary Policy Decisions Various studies have investigated the pass-through empirically in a multicountry setting. Cottarelli and Kourelis (1994) report important differences in pass-through in EMU countries. Table 5.12 shows their estimates of the short-term (initial) effect for the six biggest EU countries. Also the results of similar studies by Borio and Fritz (1995), BIS (1994), and Mojon (2000) are presented in this table. Mojon (2000) examines the pass-through in these six EMU countries for the period 1979 to 1998 for a whole range of deposit and credit rates (but which cannot

162

Chapter 5

Table 5.12 Short-run and long-run effects of policy rate increase of 100 basis points on lending rate (basis points)

Cottarelli and Kourelis (1994) Borio and Fritz (1995) BIS (1994) Mojon (2000)

Belgium

Germany

Spain

France

Italy

Netherlands

21 87 61 127 85 96

37 100 11 105 18 68

36 94 0 117 — 65

— — 43 74 43 86

12 83 26 122 14 50

52 82 108 108 125 99

Note: Top row shows the short-run effect, and next row the long-run effect. The BIS and Mojon studies only report short-run multipliers. The short-run effect refers to impact multipliers for Cottarelli and Kourelis (1994), and BIS (1994), to one-month multipliers for Borio and Fritz (1995), and to three-month multipliers for Mojon (2000). For the latter we report the multipliers that refer to short-term loan rates.

be fully compared across countries) and confirms the conclusion of heterogeneity of previous studies. Hofmann (2002) analyzes the passthrough in France, Germany, Italy, and Spain over the period 1984 to 1998, using the Johansen co-integration analysis. He concludes that innovation in the money market rate is fully passed through to shortterm and long-term business loan rates over time. In the short run the response of lending rates is, however, sluggish. Although the EMU countries now share the same currency, their financial systems show considerable differences. As pointed out by Mojon (2000), national segmentation in the retail banking industry will remain significant despite the EMU because retail banking involves heavy investments in advertising, in a network of office branches, and in customer services. Also differences in regulation can cause retail banking markets to remain segmented along national lines. Then there are differences in the balance sheet structure of households and firms as they only gradually adjust to the new monetary regime. As a consequence the pass-through from policy-controlled interest rates to bank interest rates will tend to remain country specific. This potential source of asymmetry in monetary transmission is particularly relevant in the euro area where bank rates are a key determinant of the cost of capital and the yield on savings. Toolsema et al. (2002) examine this issue in some detail using rolling regression techniques in an error correction framework. These authors estimate their model first for the period 1980 to 2000, and then apply rolling regressions. The idea behind the latter approach is to take a

Centralization or Decentralization

163

fixed number of observations and to redo the regressions, every time adding one observation at the end of the sample while dropping one at the beginning. The results indicate whether the monetary policy transmission has been stable over time in the country under consideration and whether or not convergence has occurred. They find that the long-term equilibrium multipliers in all countries have moved toward similar levels. In other words, this evidence suggests that there is convergence of long-term multipliers. This process is visible in windows that start at the end of the 1980s. Over all, most recent evidence on pass-through of monetary policy suggests that there has been a tendency toward convergence, although it is unclear to what extent differences as reported in the earlier literature have fully disappeared. In this regard Mojon (2000) and Toolsema et al. (2002) conclude that some differences still exist, whereas Hoffman (2002) finds that monetary policy changes are fully transmitted to market rates in all countries in his sample. 5.6.2 Monetary Transmission and Financial Structure There is plenty of evidence that monetary policy has a diverging effect on many countries in the euro area (e.g., see Rawaswamy and Sløk 1998). A number of studies claim that asymmetries in monetary policy transmission in the euro area result from differences in financial structure (e.g., see Dornbusch et al. 1998; Cecchetti, 1999). Cecchetti (1999) bases his view on the lending view of monetary policy transmission, according to which monetary policy actions change the reserves available to the banking system, thereby affecting the willingness of banks to lend and, ultimately, the supply of loans. Countries in which firms are more bank dependent and banking systems are less healthy will be more sensitive to ECB interest rate changes. Cecchetti (1999) relates the estimates of monetary policy impact on output and inflation of Ehrman (1998) to indicators for financial structure. He finds that countries with many small banks, less healthy banking system, and poorer direct capital market access display a greater sensitivity to monetary policy changes than do countries with big, healthy banks and deep, welldeveloped capital markets. Table 5.13 is reproduced from Cecchetti (1999). The final column in table 5.13 shows the predicted effects of monetary policy on output and inflation, with higher values indicating a stronger lending channel and therefore a larger impact. The indicator is an average of the first three columns. The first column summarizes

164

Chapter 5

Table 5.13 Summary of factors affecting the strength of the monetary policy transmission mechanism

Country

Importance of small banks (1)

Bank health (2)

Availability of alternative finance (3)

Predicted effectiveness of monetary policy (4)

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Netherlands Portugal Spain Sweden United Kingdom

3 1 2 1 2 3 2 1 2 1 1 2 1 1

2 2 2 3 3 2 2 1 3 2 3 2 3 1

3 1 1 2 2 2 3 3 3 2 3 2 3 1

2.67 1.33 1.67 2.00 2.33 1.67 2.33 1.67 2.67 1.67 2.33 2.00 1.67 1.00

Source: Cecchetti (1999).

Cecchetti’s assessment of the importance of small banks, which is based on more detailed information on the number of banks, the number of banks per million population, and the concentration ratio for the top five banks. The second column summarizes indicators for the health of the banking system (return on assets, loan loss provisions, net interest margin, and operating costs). The scores as reported in the third column are based on the number of publicly listed firms, the extent of secondary equity and debt markets, and the ratio of bank loans to all forms of finance. Table 5.14 connects the predicted effect to the estimated effects in the model of Ehrmann (1998) as reported by Cecchetti (1999). It is remarkable that—in contrast to what Cecchetti (1999) suggests—the linkage is not very strong. Indeed, if we plot Cecchetti’s indicator for financial structure and the estimated impact on production growth, a positive, but insignificant effect shows up (see figure 5.4). So the evidence is that differences in financial structure do not cause much difference in monetary transmission. This conclusion is broadly in line with the outcomes of a Eurosystem research project, which is summarized in sidebar 12.

Table 5.14 Predicted and estimated effects of monetary policy Maximum impact on Country

Predicted effectiveness of monetary policy

Output

Inflation

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Netherlands Portugal Spain Sweden United Kingdom

2.67 1.33 1.67 2.00 2.33 1.67 2.33 1.67 2.67 1.67 2.33 2.00 1.67 1.00

— -0.72 -0.48 — -1.30 -1.21 — -0.76 -0.64 — -0.39 -0.46 -0.56 -0.53

— -0.05 -0.34 — -0.21 -0.48 — -0.25 -0.25 — -0.28 -0.23 -0.11 -0.37

Note: Second column reproduces column 4 of table 5.13; the third and fourth columns show the maximum impact of a 100 basis point increase in interest rates as reported by Cecchetti (1999) on output and inflation.

y = 0.1976x + 0.3295

1.4

R 2 = 0.109

Maximal output effect

1.2

1

0.8

0.6

0.4

0.2

0 0

0.5

1

1.5

2

Financial structure indicator Figure 5.4 Financial structure and impact of monetary policy

2.5

3

12 Differences in Monetary Policy Transmission: The ECB View In its Monthly Bulletin of October 2002 the ECB reports on a large-scale research project on monetary transmission in the euro area.19 As pointed out by the ECB, there are two approaches to determining the importance of the different channels in affecting the evolution of prices and output in the euro area. On the one hand, structural econometric models can be used to try to disentangle some of the channels and identify their relative quantitative importance at the macroeconomic level. A drawback is that the analysis is dependent on the model, and the result can be driven partly by modeling choices. On the other hand, disaggregated data from balance sheets of nonfinancial firms and banks can be used to analyze specific key links in the transmission mechanism, such as the role of financial factors and the supply of bank credit. This approach is promising because of the importance of bank lending as a source of finance in the euro area. With this type of research, however, it is not always easy to deduce the macroeconomic importance of evidence in favor of financial factors playing a role. Both approaches have been used in this project. The main results of the research can be summarized as follows: Investment seems to be an important driving force behind output changes in the wake of a monetary policy shock. The results confirm that business investment is sensitive both to changes in the user cost of capital and, to a more limited extent, to liquidity or cash-flow effects. Financial and credit constraints seem to play a role in explaining the response to monetary policy in some countries and for specific groups of firms or banks but are not of central importance for the euro area as a whole. The empirical evidence does not suggest that there are systematic differences among countries in policy transmission that are robust across different studies and methodologies. However, there is evidence for diverging monetary policy measures affecting economic sectors and also evidence that the effects of monetary policy on output is stronger in periods when the balance sheets of households and firms are weak, such as during a downturn. A number of studies have used large-scale macroeconomic models at the disposal of the European Central Bank and the national central banks of the Eurosystem to compare the effects of monetary policy. Table 5.15 gives the results of a common monetary policy experiment with these econometric models, using a 100 basis point rise in the policy interest rate over two years. As the results show, this measure is found to lead to a maximum aggregate drop in output in the national models of about 0.4 percent after two years. A similar effect is found for inflation, but in this case it occurs two years later. There are some notable variations in the results across models with respect to both the magnitude and timing of the effects. The impacts on output and prices were found to be relatively modest in Belgium, France, the Netherlands, and Luxembourg and relatively strong in Italy, Spain, Portugal, and Greece.

Centralization or Decentralization

167

Table 5.15 Effects of 100 basis point rise in the policy interest according to models of national central banks Year

2001

2002

2003

2004

2005

Consumption deflator Belgium Germany Greece Spain France Ireland Italy Luxembourg Netherlands Austria Portugal Finland Aggregate Euro areawide model

-0.10 -0.05 -0.16 -0.04 -0.07 -0.09 -0.15 -0.02 -0.12 -0.10 -0.07 -0.53 -0.09 -0.15

-0.18 -0.19 -0.24 -0.25 -0.10 -0.15 -0.33 -0.05 -0.20 -0.14 -0.22 -0.50 -0.21 -0.30

-0.21 -0.38 -0.29 -0.46 -0.12 -0.15 -0.47 -0.06 -0.22 -0.14 -0.26 -0.17 -0.31 -0.38

-0.17 -0.56 -0.35 -0.68 -0.14 -0.17 -0.50 -0.09 -0.30 -0.13 -0.27 -0.02 -0.40 -0.49

2001

2002

2003

2004

2005

-0.10 -0.09 -0.69 -0.62 -0.25 -0.43 -0.55 -0.27 -0.25 -0.49 -0.81 -0.15 -0.31 -0.71

-0.05 0.15 -0.72 -0.56 -0.16 -0.38 -0.21 -0.23 -0.22 -0.36 -0.74 -0.22 -0.14 -0.63

-0.03 0.26 -0.74 -0.39 -0.08 -0.32 0.05 -0.15 -0.16 -0.32 -0.61 -0.25 -0.02 -0.57

Real GDP -0.12 -0.56 -0.38 -0.86 -0.16 -0.22 -0.37 -0.13 -0.38 -0.12 -0.35 -0.08 -0.40 -0.66

-0.15 -0.28 -0.41 -0.12 -0.15 -0.25 -0.26 -0.17 -0.20 -0.25 -0.12 -0.34 -0.22 -0.34

-0.20 -0.33 -0.78 -0.43 -0.28 -0.48 -0.60 -0.25 -0.27 -0.47 -0.56 -0.24 -0.38 -0.71

Source: Van Els et al. (2001).

5.7 Conclusions A consequence of the decentralized Eurosystem is that many of its central banks have political weight that exceeds their economic weight. While not necessarily a problem, this structure can introduce an unwelcome bias into the ECB’s decision making if country representatives put some weight on national economic developments and these developments deviate notably from the behavior of euro area aggregates. Although members of the ECB Governing Council do not act as national representatives, but as fully independent persons, it is certainly possible that national economic welfare plays at least some role in the (voting) behavior of regional representatives in the ECB Council. Indeed, if national background should not play any role in the ECB decision-making process, why then have the European governments not delegated monetary policy fully to the Executive Board in the first place? The hypothesis of regional influences in a federal central bank system is supported by recent studies on US and German monetary policy.

168

Chapter 5

There is quite some evidence that within the euro area, countries diverge in terms of their business cycles, inflation, and—to a lesser extent—monetary transmission. Our findings cannot give a definite answer to the question of whether business cycles will become more or less synchronized under EMU. It is likely that EMU will further intensify trade relations among the EMU countries, and this will lead to more synchronization according to most studies. However, there is also evidence indicating that monetary integration can lead to less synchronization as the stabilizing influence of exchange rate fluctuations is removed. The evidence discussed in this chapter further suggests that inflation differentials within the euro area can be persistent. As far as the differences in monetary transmission are concerned, it appears that differences in the estimated impact of monetary policy on output and prices across countries do not tend to be robust across different methodologies, data, and models.20 Although differences in transmission can be detected in individual studies, they are often not statistically significant and moreover inconsistent across studies. Also the connection between transmission differentials and differences in financial structure can be questioned. When it comes to the pass-through of monetary policy measures to money market rates, there is evidence suggesting at least some convergence. What are the policy implications of the analysis? Our analysis seems to lend some support to the view of euroskeptics as it appears that the optimism on synchronization of business cycles—as, for instance, expressed by Trichet (2001)—is not fully warranted: there is only mixed evidence that further integration will lead to more synchronization of business cycles and inflation differentials will not disappear. The upcoming enlargement of the monetary union will probably only increase economic divergence. In order to limit the risks of suboptimal policies due to diverging economic developments, we advocate a more centralized decision-making process within the ECB. In chapter 7 we will return to the issue of ECB reform. However, next, in chapter 6, we discuss the future enlargment of the Monetary Union.

6

New Member Countries

6.1 Introduction In this chapter we discuss the implications of the enlargement of the European Union (EU) for EMU. The current members of the monetary union will be joined by a number of new entrants that have substantially lower incomes per capita. On May 1, 2004, ten countries joined the European Union, and Bulgaria and Romania will become members in 2007. The new EU countries are members of the EMU with a socalled derogation. After a two-year waiting period, their convergence will be evaluated based on the Maastricht criteria. At the earliest, the new EU member states may therefore join the euro area in 2006 after a positive assessment. Several other countries, notably Slovenia and Estonia, are aiming to join the euro area very soon after EU accession, and the incumbent members of the monetary union are not likely to be able to do much to keep the aspirants out. In section 6.2 we assess the convergence achieved so far. Apart from fulfilling the Maastricht criteria, the new member countries should have an independent central bank before they can enter the monetary union. Section 6.2 therefore starts with a review of the independence of the central banks in the acceding countries. In the remainder of the section we try to find an answer to the question of whether a quick entry into the monetary union is in the best interest of these countries and, if so, which exchange rate will be optimal in the intermediate period. What are the implications of the entrance of new member countries? Will the enlargement of EMU threaten the viability of the monetary union? In other words, is enlargement of the monetary union in the interest of its current members? Section 6.3 deals with these questions.

170

Chapter 6

Similar considerations as those dealt with in chapter 5 (i.e., business cycle synchronization, inflation differentials, and financial-structurerelated differences in monetary transmission) will be discussed. Section 6.4 offers our conclusions. 6.2 Central Bank Independence and Convergence 6.2.1 Central Bank Independence Negotiations about EU membership started with the Czech Republic, Cyprus, Estonia, Hungary, Poland, and Slovenia in March 1998, while in February 2000 negotiations were opened with Bulgaria, Malta, Latvia, Lithuania, Romania, and Slovakia. To qualify for membership the applicants must have largely adapted their laws to comply with EU legislation and have the ability to implement that legislation. This implies that, before entry, the applicant countries should also have implemented those aspects of the Economic and Monetary Union acquis communautaire as defined by Title VII of the EC Treaty. This legislation refers to the prohibition of direct public sector financing by the central bank, the prohibition of privileged access of the public sector to financial institutions, and independence of the national central bank. In this section we discuss the recent literature on central bank independence that relates to the new member countries. Most of this literature focuses on the transition countries; much less is known about Malta and Cyprus. As we saw in chapter 4, it is often argued that central bank independence (CBI) is crucial for the credibility of monetary policy. A large body of empirical research on industrialized countries has provided evidence that a high degree of CBI is positively correlated with lower average inflation rates.1 Many countries in transition delegated monetary policy to a (legally) independent central bank. However, legal independence is not enough, as stressed by Jean-Claude Trichet (2002), at the time governor of the Banque de France, at the Munich Economic Summit, June 8, 2002: The effective implementation of the acquis communautaire is not only a legal prerequisite for accession to the EU. It also implies the effective transformation of accession countries’ economic framework, which should facilitate their integration into the EU and, later, the euro area. In this context, it should be ensured that there is no discrepancy between the central banks’ formal status in the legislation and the implementation of that legislation. We consider that comprehensive concept as an essential contribution to the clarity and the credibility of

New Member Countries

171

the single monetary policy. It is of utmost importance that all present and future Member States respect this economic and institutional ground rule of the European framework.

The main purpose of this section is therefore to analyze the degree of legal CBI achieved by the new member countries and to analyze to what extent it has been implemented. In the literature, various definitions of CBI and proxies for CBI can be found. The bottom line of indicators for CBI is that they all try to measure how much influence government has on monetary policy making. Most indicators are based on central bank laws. Still the criteria taken into account, the grading and weighting methods, and the interpretation of laws differ widely across various studies. Consequently it is no surprise that research on CBI comes up with sometimes strikingly different results. Most of the research on CBI refers to industrial countries. However, the literature on legal and, to a lesser extent, actual CBI in countries in transition has increased substantially in recent years. Simultaneously various indexes of legal independence have been built.2 Relevant contributions are Cukierman et al. (2002), Hochreiter and Kowalski (2000), Lybek (1999), Loungani and Sheets (1997), Radzyner and Riesinger (1997), Malizewski (2000), and Dvorsky (2000). Table 6.1 presents various indexes of legal CBI. Loungani and Sheets (1997) derive two indexes of CBI: the first covers the goal of economic and political independence (CBI-DF), while the second assesses similarity between the analyzed law and the statute of the Bundesbank (SIB). Lybek (1999) has built an indicator for 15 former Soviet republics consisting of 21 criteria, including most of the elements found in earlier CBI indexes. Malizewski (2000) has constructed a slightly modified version of the index of Grilli et al. (1991) for transition countries.3 Similarly Cukierman et al. (2002) have calculated the values of the independence index of Cukierman (1992) for the transition countries. Also Dvorsky (2000) has calculated CBI for the CEEC-5 applying the Cukierman index. The diverging results in table 6.1 illustrate that the degree of CBI largely depends on the design of the index and the criteria included. It is remarkable that at least on paper most central banks in the new member countries are independent. On average, aggregate legal independence is substantially higher than in developed economies during the 1980s (Cukierman et al. 2002). After the most recent central bank

172

Chapter 6

Table 6.1 Legal central bank independence in (potential) new EU member states LounganiSheets Country

Dvorsky

CBI-DF

SIB

Lybek

Maliszewski

Cukierman et al.

Bulgaria Cyprus Czech Republic Estonia Hungary Latvia Lithuania Malta Poland Romania Slovakia Slovenia

NA NA 0.70 NA 0.75 NA NA NA 0.90 NA 0.69 0.60

0.88 NA 0.88 1.00 0.31 NA 0.13 NA 0.50 0.50 NA NA

1.00 NA 1.00 0.67 0.72 NA 0.33 NA 0.61 0.56 NA NA

NA NA NA 19 NA 15 18 NA NA NA NA NA

15 NA 13 13 10 12 15 NA 14 7 11 11

0.55 NA 0.73 0.78 0.67 0.49 0.78 NA 0.89 0.34 0.62 0.63

Note: The index of Cukierman et al. (2002) is the most recent LVAW index after CB reform and removal of the ruble.

reform (enacted in 1997), the central bank in Poland has, for instance, a score of 0.89 on the Cukierman index compared to 0.69 for the Bundesbank during the 1980s. It has been widely acknowledged that, especially in developing countries, legal indicators of CBI are often incomplete and noisy indicators of actual CBI. Central bank legislation may not always be implemented as initially intended for several reasons.4 So there may be substantial deviations between the law and actual practice (see also sidebar 13). Furthermore the selection of criteria, the grading and the weighting imply arbitrariness to a certain extent and are, therefore, subject to discussion. In response to these shortcomings of legal CBI indexes, Cukierman (1992) and Cukierman et al. (1992) provide a different approach to measuring CBI. Their proxy is based not on central bank laws but on the actual average term in office of the central bank governor. This indicator is based on the assumption that a higher turnover rate of the central bank governor (TOR) reflects a lower degree of CBI. The TOR is calculated as follows:

New Member Countries

173

13 CBI and the Rule of Law Eijffinger and Stadhouders (2003) present an extension of the empirical research on the relationship between CBI and inflation. According to these authors, legal arrangements are a prerequisite of CBI, but the translation of these legal arrangements into actual practice is of much greater importance. Eijffinger and Stadhouders argue that the translation of legal to actual CBI mainly depends on the rule of the law in a country. The authors therefore introduce institutional quality indicators (IQIs) as proxies for the rule of the law. These indicators are treated as a missing link between legal and actual CBI. The idea behind this approach is that if there is an interaction between legal CBI and IQIs, a measure for actual CBI can be constructed using the IQIs as correction factor for legal CBI. The outcome is a measure for effective CBI. Eijffinger and Stadhouders construct proxies for actual CBI for both developed and developing countries and test whether the institutional environment is relevant to achieve price stability. They find that the rule of law and the institutional framework do matter in keeping inflation low, especially in countries in transition. Legal independence is a necessary, but not sufficient, condition for actual CBI; it needs to be accompanied by much liberalization. The peculiarities and characteristics of transition economies—narrow capital markets and limited access to foreign financing, large price shocks, and macroeconomic imbalances— undermine over ambitious institutional reform plans the same as they do the rule of law. The effectiveness of CBI entails a high degree of freedom to build up the central bank’s reputation. Eijffinger and Stadhouders investigate effective CBI in transition economies by using legal transition indicators (LTIs). These indicators are provided by the European Bank for Reconstruction and Development (EBRD) and are measures for the quality of the commercial legislation in terms of two criteria: extensiveness and effectiveness of commercial law. The overall legal transition indicator is significantly and positively correlated with inflation during early phases of transition. This result suggests that during these phases the ambitious institutional reforms are inefficient because macroeconomic imbalances and huge price shocks raise the temptation to bend the law. When liberalization has reached a sufficiently high level, the legal transition indicator becomes significantly and negatively related to inflation. The measure indicative of effective CBI comes from multiplying the index of legal CBI (based on the LVAW indicator of Cukierman et al. 2002) by the legal transition indicator.

174

Chapter 6

Table 6.2 Legal transition (LTI) and effective CBI

Bulgaria Czech Republic Estonia Hungary Latvia Poland Romania Slovak Republic Slovenia

LTI

LVAW

LVAW*LTI

0.57 0.86 0.86 0.86 0.57 0.86 0.57 0.57 0.57

0.55 0.73 0.78 0.67 0.49 0.89 0.34 0.62 0.63

0.31 0.63 0.67 0.58 0.23 0.77 0.19 0.35 0.36

Source: Eijffinger and Stadhouders (2003).

TOR =

Number of central bank governors (6.1) . Length of actual term of office (years or fractions of years)

The idea behind this approach is that central bank governors who are not as willing to follow government instructions will more often be dismissed than “cooperative” ones. Of course, it is possible that a low TOR score is the result of a central bank governor being more compliant to government, and who therefore remains longer in office than his renegade colleagues. Another reason to cautiously regard this indicator is that a governor who resigns could have done so for voluntarily, personal reasons. The studies in which the TOR is used as a proxy for CBI generally conclude that there is a significant relationship between the TOR and the inflationary picture of developing countries. However, Sturm and De Haan (2002) have extended the data set covering almost twice as many countries as Cukierman (1992) and included information from the 1990s. Their findings are in sharp contrast to most of the previous work, leading to the conclusion that in developing countries the turnover rate of the TOR is often not related to inflation. Likewise Lybek (1999) failed to detect any connection between the number of governors to the average annual inflation and average annual growth in the countries of the former Soviet Union for the period from 1992 to September 1998. However, the period taken into account was probably too short to obtain reliable results. Lybek also provided some anecdo-

New Member Countries

175

tal evidence that some CB governors resigned before their term officially expired due to increasing political pressure. Apparently some governors stayed in office because they pursued an accommodating monetary policy. It is interesting to observe the degree of effective CBI in the transition countries using the TOR as a proxy. The TOR can offer valuable insight on the gap between legal and actual CBI (Cukierman et al. 2002). Nevertheless, the short history of central banks in transition countries should make one careful in drawing conclusions. Any small shift in the base period can lead to big differences in the TOR. Radzyner and Riesinger (1997) have calculated the TOR for the Czech Republic, Hungary, Poland, Slovenia, and the Slovak Republic. They conclude that although all central banks covered enjoy a comparatively high degree of legal CBI, the actual term of office was significantly shorter in some countries. The highest TOR is recorded in Poland (see table 6.3). Dvorsky (2000) has updated the study of Radzyner and Riesinger (1997).5 Except for Slovakia, the countries Table 6.3 Central bank independence in (potential) new EU member states: TOR Country

Radzyner-Riesinger

Dvorsky

Czech Republic

0.23 (12/92–3/97)

0.13 (12/92–8/00)

Estonia Hungary

0.38 (12/91–3/97)

0.23 (12/91–8/00)

Slovakia Slovenia

0.49 (2/89–3/97) 0.23 (11/92–3/97) 0.17 (6/91–3/97)

0.35 (2/89–8/00) 0.26 (11/92–8/00) 0.11 (6/91–8/00)

a. Source: Hochreiter and Kowalski (2000).

5 6

0.16 (1/92–9/98) 0.62 (1/92–9/98)

Lithuania

Legal term in officea 6

0.32 (1/92–9/98)

Latvia

Poland

Lybek

6 5 6 6 6

176

Chapter 6

monitored came up with a lower TOR compared to 1997, mainly resulting from a longer base period and almost no personal changes at the top of the central banks. The highest TOR is again registered for Poland (0.35), which is quite interesting, since Poland has by far the highest scores in terms of legal CBI. According to Dvorsky, the high TOR is exclusively due to the frequent changes of governors from 1989 to 1992 and does not adequately reflect the positive track record after 1992. Sidebar 14 provides some further discussion of the Polish case.

14 CBI in Practice: Poland According to Radzyner and Riesinger (1997) and Dvorsky (2000), Poland had a very high TOR. In terms of legal independence, the Polish central bank is in the top flight of central banks and displays the highest scores in applying various measures of legal CBI. This apparent paradox is, however, quite easy to explain. The high score for the TOR in the reference period (1989–2000) is exclusively due to the frequent changes of central bank governors in the very early stage of transition from 1989 to 1992. The frequent changes from 1989 to 1992, however, turned out to be evidence of the high political dependence of the National Bank of Poland’s (NBP) at the time. Both Pakula and Baka resigned after the appointment of a new government. Wojtowicz, the third governor was tripped by a financial scandal and served the shortest term in office. When Gronkiewicz-Waltz was governor, a number of important central bank law amendments were approved. In 1998 a new legal and institutional framework for Poland’s central bank and the whole banking system was adopted. The key issues taken into account were as follows (see NBP 1998): The need to enhance CBI (with particular reference to monetary policy). • The desire to foster the collective nature of the NBP management process. • The European Treaty, which commits Poland to coordinate its domestic legislation with EU law. •

Gronkiewicz-Waltz shortened her second term of office to take up her new position as vice president of the European Bank for Reconstruction and Development (EBRD) in January 2001.

New Member Countries

177

Table 6.4 Central bank governors in Poland in chronological order Zdzislaw Pakula Wladyslaw Baka Grzegorz Wojtowicz Hanna Gronkiewicz-Waltz Leszek Balcerowicz

July 1988 September 1989 January 1991 March 1992 since January 2001

September 1989 January 1991 August 1991 December 2000

6.2.2 The Maastricht Criteria EU membership does not imply immediate membership in EMU. However, the new member countries will not get a similar position as the United Kingdom and Denmark: they have an obligation to join EMU. Before they can enter EMU, the new members have to fulfill the convergence criteria as stipulated by the Maastricht Treaty. These include a two-year participation in the Exchange Rate Mechanism Mark II (ERM II), without a devaluation of the parity rate against the euro during this period. ERM II is solely open to EU member states, and participation in this mechanism can thus begin only after accession to the European Union (see sidebar 15).6 Whether and when the new member countries satisfy the Maastricht criteria will be, to a significant extent, at their discretion. Sweden has thus far evaded the obligation to join EMU by not satisfying the exchange rate criterion (Buiter and Grafe, 2002). The Maastricht Treaty contains four convergence criteria: 1. Price stability. The average inflation rate (measured on the basis of the consumer price index) must not exceed by more than 1.5 percentage points that of, at most, the three best-performing member countries. 2. Sustainable fiscal position. There can be no excessive deficit. Excessive deficit exists if the budget deficit is higher than 3 percent of GDP—unless either the ratio has declined substantially and continuously and has reached a level that comes close to 3 percent, or the excess over the 3 percent reference value is only exceptional and temporary and the deficit remains close to 3 percent;

•

the ratio of gross government debt to GDP exceeds 60 percent— unless the ratio is sufficiently diminishing and approaching the reference value at a satisfactory pace.

•

178

Chapter 6

15 ERM II At its meeting in Amsterdam in June 1997, the European Council decided to replace the Exchange Rate Mechanism (ERM) by a new ERM. The Exchange Rate Mechanism Mark II (ERM II) offers an opportunity for EU members that participate in the monetary union (the “ins”) as well as members that do not (the “outs”) to stabilize their exchange rates. By the Maastricht Treaty, each member state that is not yet allowed to participate in the euro area must treat its exchange rate policy as “a matter of common interest.” In principle, this should also apply to the countries with an opt-out clause, namely Denmark and the United Kingdom. Nevertheless, membership of ERM II is voluntary for all “outs.” The operating procedures for ERM II were laid down in an agreement between the ECB and the national central banks of EU countries outside the euro area. ERM II is designed as an asymmetrical, euro-centered exchange rate system. The main feature of ERM II is the wide fluctuation of ±15 percent between the euro and the currency of the country participating in the mechanism. The bilateral central rate and the upper and lower intervention rates (expressed as a percentage of the central rate) are quoted using the euro as the base currency. Analogous to the “old” mechanism is another important feature of ERM II. This is an automatic and unlimited intervention at the margins by both the ECB and the participating national central bank using the Very Short-term Financing Facility. So the “new” mechanism involves a reciprocal commitment by the ECB and respective central banks. It must be emphasized that these credit lines are of a very temporary nature because the debt has to be repaid within 75 days after the end of the month in which the intervention took place. However, the ECB can suspend automatic intervention if this conflicts with its primary objective of maintaining price stability. At the time of writing, the currencies in ERM II are the Danish krone, the Estonion kroow, the Lithnanian litas, and the Slovenian tolar. Before Greece joined EMU, the Greek drachma also participated in ERM II. ERM II is compatible with a fairly broad range of exchange rate arrangements. The Ecofin Council has only excluded three regimes: any regime without a mutually agreed central rate to the euro, crawling pegs, and pegs to currencies other than the euro. Entering the EU with a currency board where the currency is pegged to the euro is compatible with ERM II, but the exchange rate target must be clearly viable, for instance, because the system has been in operation for a substantial period. A major challenge for entering ERM II will be finding a central rate against the euro. To avoid the risk of misalignment, the central parity should be as close as possible to the equilibrium exchange rate. There is, however, a wide variety of methods and approaches to calculate equilibrium exchange rates (see Égert 2003 for an excellent survey).

New Member Countries

179

Wyplosz (2003b) argues that ERM I could survive for a decade thanks to the widespread presence (in the countries with a weak currency) of capital controls. When these controls were eliminated, ERM I became unstable. According to Eichengreen and Wyplosz (1993), some of the speculative attacks on currencies participating in ERM I were not related to “fundamentals.” The new members are committed to repeal all capital controls upon EU accession. According to Wyplosz, this means that ERM II will become structurally unstable, possibly even with the wide ±15 percent margins.

3. Exchange rate stability. The currency must respect the “normal” fluctuation margins of the Exchange Rate Mechanism Mark II (ERM II), without severe tensions for at least two years (especially no devaluation on the initiative of the member country concerned). 4. Low interest rate. The average long-term interest rate must not exceed by more than 2 percentage points the interest rates in, at most, the three best-performing countries in terms of price stability. Although these criteria have been criticized for their lack of theoretical foundation (e.g., see Eijffinger and De Haan 2000), the member countries of the European Union have made it clear that the new member countries must stick to this part of what is called the acquis cummunautaire. The Maastricht convergence criteria cannot be changed for new members, and the criteria must be applied in the same manner as in the convergence examinations so far. This is to “ensure equal treatment between future Member States and the current participants in the euro area” (Ecofin 2000). One of the most critical issues seems to be the fulfilment of the budget deficit criterion, as various new member countries at the time of writing record major fiscal imbalances. The Czech Republic, Hungary, Poland, and Slovakia had budget deficits ranging from more than 4 percent to 8.5 percent of GDP in 2002 (Backé and Wójcik 2003). Another important issue is the choice of the proper exchange rate regime for the new member countries in the period between entering the European Union and becoming a (full) member of the Monetary Union. The exchange rate regime is a key determinant of a country’s macroeconomic stability, which affects the investment climate. Apart from the perspective of future EMU membership, the choice of exchange rate regime is therefore of great relevance for the new member

180

Chapter 6

countries. Table 6.5 shows the exchange rate regimes of the (potential) new EU members in 2003. An important consideration in choosing an exchange rate regime is that the new members have to liberalize international capital flows as part of the acquis communautaire, making them more vulnerable to speculative attacks. As Wyplosz (2003b) argues: “The repeated worldwide experience is that, almost unavoidably, disappointments of one kind or another (economic, political, international) eventually arise and trigger speculative reversals of capital flows. EU accession will reinforce this phenomenon, while the mandatory repeal of capital controls will magnify the size of capital movements in either direction. . . . Capital inflows and outflows stand to rock the exchange rate and to trigger unmanageable speculative attacks.” As follows from table 6.5, a number of countries currently have a currency board. A currency board can be considered the most credible form of a fixed exchange rate regime. It ensures that the own currency is convertible against a fixed exchange rate with some other currency(ies) that is codified, be it in a law or otherwise. The anchor currency is generally chosen for its expected stability and international acceptability. There is, as a rule, no independent monetary policy as the monetary base is backed by foreign reserves. A currency board is a strong, “double-barreled” commitment device (Buiter and Grafe 2002). Through the currency peg it represents a comTable 6.5 Exchange rate regimes of (potential) new EU member states, 2003 Country Bulgaria Cyprus Czech Republic Estoniaa Hungary Latvia Lithuaniaa Malta Poland Romania Slovakia Sloveniaa

Exchange rate regime Fixed peg to euro (currency board) Fixed peg to euro with de facto band ±1–2% Free float (inflation targeting) Fixed peg to euro (currency board) Crawling peg to euro with band ±15% (implicit inflation targeting) Fixed peg to SDR (quasi-currency board) Fixed peg to euro (currency board) Peg to weighted basket of euro, USD, GBP (±0.25% band) Free float (inflation targeting) Managed float Managed float Managed float

Source: EEAG (2004). a. Participates in ERM II since June 2004.

New Member Countries

181

mitment to price stability. Through the “no domestic credit expansion” constraint, it represents a commitment to budgetary restraint. The value of these commitments depends either on the currency board arrangement being perceived as credible and permanent or on the belief that, if it is abandoned, it will be replaced by something representing a comparable commitment to price stability and budgetary responsibility as a credible currency board, like the EMU. At the other extreme, a country may choose a floating exchange rate regime with an independent central bank with some kind of an inflation-targeting strategy. Berger et al. (2001b) show that a currency board becomes, ceteris parifus, more attractive under the following conditions: The imported foreign monetary policy is in the hands of an independent and “conservative” (i.e., inflation-averse) foreign central bank.

•

The home country’s central bank is relatively dependent and outputoriented compared to the foreign central bank.

•

The correlation between the home and foreign country’s output shocks is high.

•

Compared to a full-fledged central bank, a currency board is a cheap way of managing monetary policy. As pointed out by Buiter and Grafe (2002), all that is needed is a sufficient number of modestly skilled bank clerks who exchange, at a fixed rate, domestic currency for the foreign currency in terms of which the peg is defined. As a currency board implies that the central bank cannot (fully) act as lender of last resort, no country should consider a currency board unless it can afford to do without a lender of last resort. As this safety net for the financial sector is missing, a prerequisite for a currency board is a reasonably healthy financial system. Likewise no country should consider a currency board unless it has a sound fiscal framework that will not require discretionary access to central bank financing by the general government. A currency board runs the risk of a real misalignment. If a country’s inflation remains higher than that of the pegging country, the currency can become overvalued (Pautola and Backé 1998). While fixing the exchange rate is a fast way to disinflate an economy starting with a higher inflation rate, pegging the exchange rate will not necessarily reduce the inflation rate instantaneously to that of the pegging country. There are several reasons why inflation will not fall right away (Roubini 1999). First, purchasing power parity does not hold exactly in

182

Chapter 6

the short run, since domestic and foreign goods are not perfectly substitutable and the mix of goods and services in the countries concerned may differ. Second, nontradable goods prices do not feel the same competitive pressures as tradable goods prices, so inflation in the nontraded sector may fall only slowly. Third, there is significant inertia in nominal wage growth. Often wage contracts are backward looking and the adjustment of wages occurs slowly. Finally, differing productivity growth rates may be reflected in differences in price increases (BalassaSamuelson effect; see section 6.3.2). If domestic inflation does not converge to the level of the pegging country, a real appreciation will occur over time. As Roubini (1999) points out, such a real exchange rate appreciation may cause a loss of competitiveness and a structural worsening of the trade balance, which makes the current account deficit less sustainable. As we pointed out above, according to many observers, speculative attacks are a huge risk in the run-up to membership of the monetary union. To what extent are currency boards less vunerable than other exchange rate systems to specultative attacks? Although it is sometimes claimed that speculative attacks cannot occur under currency boards, recent experience has shown otherwise. Still, existing evidence suggests that fixed exchange rate regimes are more likely to survive than pegs with more flexibility (Frankel, 1999). Buiter and Grafe (2002) argue that from an economic point of view, a currency board with the euro can make sense for the new member countries. A country with a currency board would have a natural “strong exit” in the form of EMU membership. Furthermore they have restructured their banking sectors, so that the lack of a lender of last resort may not be problematic. Monetary independence through a floating exchange rate permits flexibility (the valuable ability to respond to shocks), but the downside of this flexibility is the risk of opportunistic policies. According to Buiter and Grafe (2002), the benefits of monetary independence in most new member countries should not be overstated. They argue that monetary policy in these countries is particularly unlikely to be very effective in stabilizing output because credit, deposit, and debt markets are still rather underdeveloped. Consequently changes in the cost and availability of domestic credit are unlikely to have a large immediate effect on output, either through the interest rate or through the credit channel. In monetary and financial systems undergoing rapid transformations the monetary transmission mechanisms are both poorly understood and

New Member Countries

183

quite unstable, which increases the likelihood that central banks in the region may miss the announced inflation targets. Finally, as we have shown above, the strength of the political commitment to central bank independence remains questionable in some new member states. Even if no further challenges to central bank independence occur, it will take time for markets to become assured of the independence of the central banks. It follows from the preceding analysis that a currency board with a peg to the euro may be the proper exchange rate regime for the new member countries on their road to full EMU membership. Apart from the (related) risk of misalignment, there may, however, be a serious problem. Together, the exchange rate and the inflation criterion restrict the scope for changes in the real exchange rate of the new member countries vis-à-vis the euro. Due to the Balassa-Samuelson effect, these countries may experience higher inflation than the euro area in case of a nominal fixed exchange rate. This has led Szapary (2000) and Buiter and Grafe (2002) to argue that the inflation criterion of the Maastricht Treaty should be relaxed or reinterpreted.7 To examine whether this conclusion is justified, in section 6.3.2 we will discuss the literature on the Balassa-Samuelson effect in the transition countries. However, first we must consider the optimal currency area (OCA) literature, as it is relevant to our discussion on the entrance of the new member countries to the monetary union. 6.2.3 Optimal Currency Area Considerations It is well known from the OCA literature that the Maastricht convergence criteria do not help a country decide whether—from an economic perspective—it should join a currency union (e.g., see Eijffinger and De Haan 2000). OCA theory suggests that the answer should depend on the balance of benefits and costs, whether it is beneficial for a country to join a monetary union. In essence, by OCA theory, countries exposed to symmetric shocks or possessing mechanisms for the absorbtion of asymmetric shocks should find it optimal to use a common currency. The benefits of monetary union include lower transaction costs, reduction of exchange rate volatility and uncertainty, more price transparency, and a better functioning internal market. The costs of a monetary union mainly derive from the loss of an instrument of economic policy, meaning the exchange rate. In a monetary union the participating countries no longer can change the price of their own currency

184

Chapter 6

vis-à-vis other currencies. How serious this loss is depends on three factors: the need for exchange rate adjustments, the effectiveness of exchange rate adjustments, and the availability of other instruments. The need for exchange rate adjustments depends on the importance and character of economic shocks. Country-specific shocks only strike one country, in contrast to general shocks, which strike all countries. Shocks that strike a single country are, by definition, asymmetric. However, general shocks can also have an asymmetric impact, depending on the structural characteristics of the economies being struck. For instance, an oil price hike affects oil-importing countries differently than oil-exporting countries. Also the nature of the shock matters. As pointed out by Borghijs and Kuijs (2004), flexible exchange rates can generate rapid adjustment in international relative prices even where domestic prices adjust slowly. This makes them potentially useful absorbers of real shocks, which require an adjustment in relative prices in order to “switch expenditure.” For instance, a sudden drop in demand will, under flexible exchange rates, cause a depreciation, that “crowds in” extra demand. However, exchange rate adjustment in response to monetary and financial shocks leads to undesired changes in relative prices. For instance, a negative financial shock putting upward pressure on interest rates will cause the exchange rate to appreciate, so it acts to amplify rather than dampen the negative impact on output. The more important asymmetric real shocks are, the higher are the costs of relinquishing the exchange rate instrument. Whether an exchange rate adjustment will help to adjust a country to an asymmetric shock depends on the pass-through to import prices (Borghijs and Kuijs 2004). If changes in the exchange rate do not generate an adjustment in international relative prices because passthrough to import prices is very small, the exchange rate is of little use as a shock absorber even in the case of asymmetric real shocks. According to Obstfeld (2002), empirical evidence to date suggests that exchange rate changes affect relative prices. Finally, alternative adjustment mechanisms—like labor mobility, price and wage flexibility, and fiscal policy—can, in principle, solve the adjustment problem after the occurrence of an asymmetric shock. The stronger these alternative adjustment mechanisms are, the less costly it will be to give up monetary sovereignity. The following characteristics have been argued on the basis of the OCA theory to favor retention of the national currency, and the associated scope for nominal exchange rate flexibility (Buiter 1999b):

New Member Countries

185

1. High nominal rigidity in domestic prices and/or costs. If wages and prices are flexible, adjustment after an asymmetric shock will be smooth and quick, so there is no need to use the exchange rate instrument to counter shocks. 2. Relatively low openness to trade in real goods and services. The more open an economy is, the more likely it is that domestic nominal costs, including wages and prices, will be strongly linked to the exchange rate. 3. High incidence of asymmetric (nation-specific) shocks rather than symmetric or common shocks and/or dissimilarities in national economic structures or transmission mechanisms that cause even symmetric shocks to have asymmetric consequences. As explained above, low asymmetry can call for a common currency. 4. Less diversified structure of production and demand. If a country has a well-diversified structure of production and demand, it is less likely to experience asymmetric shocks. 5. Low degree of real factor mobility (especially labor mobility) across national boundaries. As pointed out above, factor mobility is an alternative adjustment mechanism to asymmetric shocks. So a high factor mobility makes it less costly to give up the exchange rate instrument. 6. Absence of significant international (and supranational) fiscal tax–transfer mechanisms. A fiscal transfer system may act as an alternative adjustment mechanism. The country suffering an asymmetric shock will receive support in one way or another. For instance, in the US system states receive more funds from the federal budget when they are experiencing low rates of growth. So such a system makes it advantageous to enter a monetary union even if asymmetric shocks occur frequently. These criteria are sometimes used to argue that the new member countries should not aim for a quick entry into the monetary union. What is the evidence to date? Little is know about wage flexibility in the new member countries. The only empirical multiple-country studies that we are aware of are Blanchflower (2001) and Iara and Traistaru (2003).8 These studies estimate so-called wage curve models that relate wage levels to local unemployment rates. The stronger the responsiveness of wages to unemployment, the more flexible wages are. Blanchflower (2001)

186

Chapter 6

estimates wage curves for fifteen transition countries, including nine new EU members, using both individual micro data and aggregate regional data sets. He finds unemployment elasticities of pay ranging from -0.021 for the Czech Republic to -0.462 for Latvia (in regressions without fixed effects, and -0.096 (GDR) to -0.520 (Latvia) in regressions with fixed effects.9 Iara and Traistaru (2003) investigate the responsiveness of regional average earnings to local labour market conditions in Bulgaria, Hungary, Poland, and Romania. They find empirical support for the wage curve in Bulgaria, Hungary, and Poland. At -0.12, unemployment elasticity of pay in Bulgaria is highest and close to that found in advanced economies (-0.10). Hungarian and Polish elasticities, -0.05 and -0.04 respectively, are only half of that. Table 6.6, based on Van de Coevering (2003), presents some other OCA indicators. The first indicator refers to convergence in production structures. Krugman (1993) developed an output divergence index to measure the degree of specialization in any given country compared to another country or group of countries. The index is the sum of the absolute differences in share between the given country and the benchmark in a number of economic sectors. The production structure of each country is compared to the euro area as a whole for six different sectors. It follows that considerable differences exist among the production structures of the countries considered. All new member countries display a considerably higher index than the average of the euro area countries. The next columns of table 6.6 refer to openness (to the world and the euro area), measured as half the ratio of exports plus imports over GDP. Because annual trade data fluctuate considerably, three-year averages were used in calculating openness. In addition the share of intraindustry trade in total trade—measured by the so-called Grubel-Lloyd index10—is shown. If the index is 0, all exports and imports are in different types of goods. In that case countries are fully specialized and all trade is interindustry. A value of 100 means that imports and exports are of the same size in all types of goods so that all trade is intraindustry. All new member countries have strong trade relationships with the euro area. Hungary, Estonia, Slovenia, the Czech and Slovak Republics, trade even more with the euro area than any of the current euro area members, with the exception of Belgium/Luxembourg. Only Latvia, Lithuania, Poland, and Romania are less open than the average of euro area countries.

New Member Countries

187

Table 6.6 OCA Indicators: Euro area countries and new member countries

Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Average euro area Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovak Republic Slovenia

Output divergence index

Openness 1999–2001

Grubel-Lloyd index with EU15

1995

2001

World

Euro area

1993

2000

11 6 15 10 9 27 NA 11 31 8 15 19 15a NA 33 27 16 40 41 40 60 34 25

11 8 18 7 9 33 NA 9 33 9 11 18 15a NA 38 33 21 40 43 46 57b 34 24

33 68c 29 22 25 16 60 19 NA 52 26 20 35 41 58 67 60 36 41 25 30 64 50

20 42 10 12 11 8 18 9 NA 28 18 11 17 19 35 28 37 13 13 14 17 33 31

77d 81 56d 86 77 27 60 65 NA 74 47 69 65 46 58 17 57 13 15 45 30 42 61

82 85 57 88 77 26 56 73 67 72 61 87 68 39 74 57 75 20 37 58 41 68 71

Source: Van de Coevering (2003). a. Excluding Ireland. b. Data for 2000. c. Plus Luxembourg. d. Data for 1996.

188

Chapter 6

Both the euro area and the CEE countries show considerable heterogeneity with respect to the level of intra-industry trade in total trade. During the 1990s most new member countries showed substantial increase in their intra-industry trade. Still there are important differences among these countries. Hungary, the Czech and Slovak Republics, and Slovenia, even surpassed the euro area average in intraindustry trade, while some other countries showed only moderate growth of intra-industry trade. Jan Fidrmuc (2001) assesses the efficacy of interregional migration as a channel of regional adjustment in the Czech Republic, Hungary, Poland, Slovakia, and Lithuania. The analysis relates gross and net migration flows scaled by the region’s population to regional unemployment rates and average wages. The main finding is that the efficacy of labor mobility as a channel of regional adjustment to idiosyncratic shocks has been rather limited in the transition economies. Regions with high wages tend to display both high immigration and emigration. The response of net migration flows to regional economic conditions is economically small, and migration flows have been declining since the onset of transition. In our view, the OCA criteria do not unambiguously suggest that it may be better for the new member countries to wait before they enter the monetary union. The data show that considerable structural divergence within the euro area exist. So, if the current euro area is considered optimal for all countries involved, there is no reason for most of the new member countries to postpone entry. Furthermore OCA theory has recently been severely criticized, mostly on three grounds (Backé and Wójcik 2003). First, the OCA criteria are endogenous. As argued by Frankel and Rose (1998), joining a currency union will eliminate exchange rate uncertainty and reduce currency transaction costs, which will stimulate bilateral trade. This in turn will foster business cycle synchronization and reduce the exposure to asymmetric real shocks, which will validate (ex post) the adoption of the common currency (see also section 5.4). Rose (2000) finds that the trade effects of using a common currency are huge: countries with a common currency trade over three times as much as countries using different currencies. Similar results are reported by Rose and van Wincoop (2001). Frankel and Rose (2002) also find large potential benefits from the use of a common currency in trade and also that the additional trade has substantial positive effects on growth. Other studies,

New Member Countries

189

for instance, those by Melitz (2001) and Persson (2001) arrive at considerably lower benefits.11 Some studies explicitly focus on the trade-enhancing effects of the introduction of the euro. Micco et al. (2003) find that the effect of EMU on bilateral trade between member countries ranges between 4 and 10 percent, when compared to trade between all other pairs of countries, and between 8 and 16 percent, when compared to trade among nonEMU countries. Bun and Klaassen (2002), using a dynamic panel model for annual bilateral exports, find that the euro has significantly increased trade, with an effect of 4 percent in the first year and cumulating to around 40 percent in the long-run. Table 6.7 summarizes these and other studies. Clearly, the literature suggests that EU membership can lead to business cycles that bring new members in sync with those of the euro area. Fidrmuc (2001) computed the potential correlation to the business cycle in Germany for five new member countries (the Czech Republic, Hungary, Poland, Slovenia, and the Slovak Republic) using Frankel and Rose’s (1998) relation between the degree of trade intensity and similarity of business cycles. His predicted correlations for the Czech Republic, Pland, and Hungary average around 0.35, while the correlations for Slovenia and the Slovak Republic average around 0.24, since their trade is less oriented toward Germany.12 The issue of business cycle synchronization of the new member countries and the current euro area countries will be discussed in more detail in section 6.3.1. Table 6.7 Studies on the trade-enhancing effect of the euro Study

g

Standard error of g

Micco et al. (2003) Bun and Klaassen (2002) De Souza (2002) Flam and Nordström (2003) Barr et al. (2003) De Nardis and Vicarelli (2003)

0.089 0.33 0.17 0.139 0.25 0.061

0.025 0.10 0.24 0.02 0.033 0.027

Source: Rose (2004). Note: Reported are the estimated impact (g) of a common currency on bilateral trade intensity and the standard error. The reported figures are the most preferred or representative estimates according to Rose (2004).

190

Chapter 6

Second, risk-sharing arguments support the view that under full financial market integration, countries exposed to asymmetric shocks can profit from monetary unification. A common currency is seen to facilitate portfolio diversification, which allows countries to adjust more smoothly and at lower costs to asymmetric real shocks (Mélitz and Zumer 1999). For a country experiencing an adverse country-specific shock, its foreign assets holdings can smooth the income effect of the adverse shock. Third, the exchange rate can be seen as a source of shocks rather than a shock absorber, in particular for small open economies (e.g., see Artis and Ehrmann 2000). As Buiter (1999b) put it, a “fatal flaw in the OCA literature is its failure to allow properly for the international mobility of financial capital. This has led to an overemphasis on the stabilising, buffer stock potential of a market-determined nominal exchange rate, and a failure to recognise its destabilising potential. I view exchange rate flexibility as a source of shocks and instability as well as (or even rather than) a mechanism for responding effectively to fundamental shocks originating elsewhere.” Similarly Borghijs and Kuijs (2004) find for the Czech Republic, Hungary, Poland, the Slovak Republic, and Slovenia that on the basis of a Clarida and Gali-type structural VAR, in these countries the exchange rate served more as an unhelpful propagator of monetary and financial shocks than as a useful absorber of real shocks. Finally, many new member countries already use the euro (or the dollar) for transaction purposes even though they have not officially adopted the euro. Table 6.8 shows some estimates of unofficial euroization (dollarization). In conclusion, we believe that a cost–benefit assessment of the choice between a quick entry and waiting definitely favors the “earliestpossible” strategy than that of waiting (see also Thygesen 2002). Budgetary consolidation is much easier to accomplish with the help of the lower interest rate resulting from a clear perspective of entering the monetary union. Furthermore the risks of exchange rate instability become lower as the waiting period becomes shorter. For new member countries with a currency board that have already shown for a number of years that they can live with a unilaterally fixed exchange rate, it is in any case difficult to advocate waiting. There appears to be no compelling reason for persuading these member countries to give up their currency boards during the two-year waiting period as defined in the Treaty.

Feige (2003)

Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovakia Slovenia

De Nicolo et al. (2003)

FCC/TCC in 2001

FCD/M2

Years

FCD/M2

Years

FCD/total dependence

Years

0.41** 0.21* 0.59** 0.06* 0.79** 0.11** 0.27** 0.55** 0.28* 0.54*

0.392 0.111 0.172 0.163 0.3 0.329 0.156 0.428 0.156 0.353

1991–2001 1993–2001 1993–2001 1990–2001 1993–2001 1993–2001 1990–2001 1990–2001 1997–2001 1997–2001

0.324 0.09 0.132 0.128 0.286 0.291 0.199 0.234 0.126 —

1990–1999 1993–2000 1992–2000 1990–99(excluding 1996) 1993–95 1993–2000 1990–99 1990–2000 1993–2000 —

0.417 0.102 0.184 0.063 0.0185 0.423 0.178 0.283 0.141 —

1993–99 1993–2000 1995–2000 1997–99 1996–97 1993–2000 1994–99 1990–2000 1993–2000 —

New Member Countries

Table 6.8 Euroiazation and dollarization in new member countries

Note: FCC = foreign currency in circulation; TCC = total currency in circulation; FCD = foreign currency deposits. * = dollar and holdings of European legacy currencies (AST, DM, and SF); ** = dollar holdings only.

191

192

Chapter 6

So, from the perspective of the new member countries, a quick entry in the euro area would be a sensible strategy. But what about the consequences of an enlargement of the monetary union for countries that are currently in the euro area? This issue is the subject of the next section. 6.3 Implications of Enlargement of the Monetary Union 6.3.1 Business Cycle Synchronization Eichengreen and Ghironi (2001) use a model from the empirical growth literature, estimated on data for the 1980s and 1990s, to forecast growth rates in an enlarged EMU in the subsequent period. Their results are reproduced in table 6.9. The admission of new members is expected to increase the dispersion of growth rates within EMU considerably. However, Eichengreen and Ghironi (2001) conclude that growth-rate variability will not be greatly aggravated by enlargement to include the members of the 1998 Accession Group—the Czech Republic, Estonia, Hungary, Poland, and Slovenia. The expansion of the monetary union to include the 1998 Accession Group should therefore be relatively easily accommodated, while expansion to include the 2000 Accession Group (Bulgaria, Latvia, Lithuania, Romania, and Slovakia) will pose a much more serious challenge for EMU. Eichengreen and Ghironi (2001) also argue that if institutions are quickly upgraded to EU levels, the dispersion of growth rates will fall, even in the short run, reducing the strains on monetary policy. In reaching this conclusion, they use a measure of institutional quality based on indicators for voice and accountability (a measure of political and civil freedom), political stability, government effectiveness, adequacy of the regulatory framework, rule of law, and corruption control. Table 6.9 Standard deviations of growth rates in an enlarged EMU

Current EMU members in 1999 Current EMU members in 2006 All EU members in 2006 Plus 1998 accession group Plus 2000 accession group

No institutional reform

Institutional convergence

1.80 0.79 1.39 1.41 2.10

1.29 1.18

Source: Eichengreen and Ghironi (2001).

New Member Countries

193

Other authors reach less optimistic conclusions. For example, Berger et al. (2004) argue that the correlation between the cyclical components of industrial production in the various (potential) new member states and the cyclical part of industrial production in the euro area is quite low.13 Table 6.10 is reproduced from their study. Industrial production is decomposed into a trend and a cyclical component, using a HodrickPrescott filter. It follows that except for Slovenia and, to a lesser extent, Cyprus, the new member countries have business cycles that are hardly synchronized with the business cycle in the euro area. The same holds true for some other euro area countries, notably Greece and Portugal. Furthermore, as pointed out by Fidrmuc (2001), at the beginning of the 1990s the transition countries were in a transitional recession. Fidrmuc (2001) has calculated the correlation of business cycles in five accession countries and of Germany for the period 1993 to 1999 and finds that the business cycle synchronization of most of these countries was very similar to that of current euro area countries. The Autumn 2003 Report of the European Forecasting Network (EFN 2003) contains an analysis of business cycle similarities of the new member countries and the euro area. Table 6.11 is reproduced from this report. It shows some cyclical characteristics of the deviation cycles Table 6.10 Business cycle correlation (with EU12) for the period, 1990–2001 Old EU members Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Denmark Sweden United Kingdom Source: Berger et al. (2004).

New EU members 0.49 0.36 0.36 0.76 0.75 0.18 0.26 0.62 0.38 0.33 0.06 0.71 0.52 0.36 0.31

Bulgaria Cyprus Czech Republic Estonia Hungary Latvia Lithuania Malta Poland Romania Slovakia Slovenia

NA 0.32 0.11 0.11 0.20 0.17 -0.17 NA 0.17 -0.04 0.12 0.65

194

Chapter 6

Table 6.11 Business cycle properties in new member countries (deviation cycles based on industrial production)

Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Average Euro area

Proportion of time in expansion

Average duration recession

Output loss (%)

Steepness

0.42 0.44 0.52 0.50 0.54 0.43 0.43 0.53 0.45 0.43

31.3 27.0 29.0 20.0 19.0 17.3 17.0 19.0 22.4 22.7

3.36 12.65 11.03 8.00 11.98 3.88 5.11 4.63 7.58 4.52

0.11 0.47 0.38 0.40 0.63 0.22 0.30 0.24 0.34 0.20

Correlation with euro area 0.16 0.40 0.91 -0.02 -0.04 0.67 0.32 0.65

Source: EFN (2003).

extracted from the monthly indexes of industrial production from 1993: 1 onward. The deviation cycle (where the turning points are characterized by changes relative to trend) is determined by applying a band-pass filter based on two low-pass Hodrick-Prescott filters. Subsequently dating rules (which incorporate minimum phase and cycle duration restrictions) are applied along the lines of Artis et al. (2002). It follows from table 6.11 that the average proportion of time spent in expansion hovers around the theoretical benchmark of 0.5. The average length of a recession in the new member countries is very similar to the euro area average, although there is quite some variation among the new members. The strongest difference is the output loss; the amplitude of a business cycle in the new member countries is generally greater than in the euro area. Also the steepness of recessions is greater in these countries. The final column of table 6.11 shows the business cycle correlation with the euro area. It follows that Hungary, Poland, and Slovenia have high correlation coefficients, while the business cycles in Latvia and Lithuania seem to be very much out of sync with the euro area cycle. For some countries, notably Hungary and Poland, the correlations are somewhat higher than those reported in table 6.10, which probably reflects differences in time periods considered. Süpel (2003) also finds that Hungary and Poland have the

New Member Countries

195

strongest correlation with the euro area, both in terms of cycles in GDP and industrial production. EFN (2003) further shows the pattern of the business cycle synchronization over time. It reveals that Hungary, Poland, and Slovenia see their business cycle getting closer to the euro area cycle, while the Czech Republic and Slovakia move away from the euro area; the Baltic countries share similar tendencies, but they have been in the past less correlated with the euro area. Darvas and Szapáry (2004) also analyze the synchronization of business cycles between the euro area and the new member countries. In contrast to previous studies, these authors also focus on the major expenditure and sectoral components of GDP, using several measures of synchronization. Like EFN (2003), they find that Hungary, Poland, and Slovenia have achieved a high degree of synchronization with the euro area. This conclusion holds for GDP, industrial production, and exports, but not for consumption and services. The other new member countries have achieved less or no synchronization. Hungary, Poland, and Slovenia show strong improvement in cyclical correlation from 1993 to 1997 to 1998 to 2002. However, the other five countries considered show almost no tendency to move toward greater synchronization during this period. In all the new member countries there is minimal or even negative correlation with the euro area cycle of private consumption and also of services. Most papers discussed so far use industrial production for constructing business cycles, given the availability of high-frequency data for this variable. However, as pointed out by Süpel (2003), industry data can overstate the co-movements of the overall economies, a finding also supported by Darvas and Szapáry (2004). Süpel has therefore constructed a broad cycle factor, using three separate indicators for each country: annual growth of industrial production, annual growth of retail sales volumes, and annual growth of construction output. The joined cyclical component has been estimated by using a state space model. He finds that the average correlation of the cyclical component of this broad indicator in the new member countries with that of the euro area is close to zero. This should act as a clear warning: conclusions concerning synchronization of business cycles that are based on industrial production can be too optimistic. Business cycles can differ across nations or regions within a nation for various reasons. First, nations and regions can experience different shocks. Second, they may respond differently to common shocks. This may be due to differences in the reaction of policy makers to a common

196

Chapter 6

shock or to differences in the national or regional composition of output. Also differences in financial and economic structure can lead to differences in the monetary policy transmission mechanism. Figure 6.1 displays the correlation of demand shocks (y-axis) and supply shocks (x-axis) in quarterly real GDP for individual euro area and new member countries with demand and supply shocks in the euro area aggregate computed by Fidrmuc and Korhonen (2003). This approach follows that suggested in the seminal paper by Bayoumi and Eichengreen (1993). The sample period is 1991–92 to 2000 for most countries. Shocks are identified using two-variable VARs for output and prices and the Blanchard and Quah (1989) assumptions. Despite considerable variance within groups, the results indicate that, on average, demand and supply shocks are more closely correlated in today’s euro area. Most euro area countries are located in the upperright part of the figure, while most candidate countries find themselves in the lower-left, indicating nonsignificant or even negative coefficients of correlation.14 Notable exceptions from the rule are two of the more advanced countries among the candidate countries, Estonia and Hungary, which rank among the euro area countries. The correlation of shocks in Greece, which entered the euro area late and is thus early in its real convergence process, and in Ireland, which was among the

Demand shocks correlation

0.8 0.6

ITA

0.4 POL 0.2 0.0

ROM GRC

IRL BGR

-0.2

SVK CZE

HUN ESP ESTAUS PRT LUX FIN NLD BEL

FRA GER

SVN

-0.4 LTU -0.6 -0.2

0.0

LVA 0.2

0.4

0.6

0.8

Supply shocks correlation Figure 6.1 Demand and supply shocks in the euro area and in accession countries. (Source: Fidrmuc and Korhonen 2003)

New Member Countries

197

fastest growing countries in the European in the sample period, is as low as the correlation of the average new member country. Fidrmuc and Korhonen (2004) have updated their earlier study to examine the correlation of new member countries’ supply and demand shocks with the euro area (the sample period is extended to 2002). They find that the slowdown of the economy increased the heterogeneity within the European as well as within new member countries. This is particularly true for demand shocks. Among the new member countries, Poland appears to have the highest correlation of supply shocks (above 0.6). The Czech Republic, Hungary, and Latvia all have correlations of approximately 0.2. This is roughly comparable to the situations of Denmark and Greece. For Estonia, the correlation of both shocks is small, and for Lithuania and Slovakia correlation of supply shocks is negative. Similar studies were done by Horvath (2000), Frenkel and Nickel (2002), and Süpel (2003). Whereas Horvath finds that idiosyncratic shocks are prominent in the accesion countries, Frenkel and Nickel conclude that “several individual CEECs exhibit shocks and shock adjustment processes that are fairly similar to some euro area countries.” Süpel’s main findings are that supply and demand shocks in Hungary, Poland, and Slovenia are positively correlated with shocks in the euro area, while Lithuania’s correlation in both demand and supply shocks is negative. The Czech Republic and Slovakia are found to be positively correlated with the euro area in terms of their supply shocks but not in terms of demand shocks. Table 6.12 compares the results of Fidrmuc and Korhonen with those of Horvath (2000) and Frenkel and Nickel (2002) in somewhat more detail.15 It follows that there is wide variety in results across these three studies. Take, for example, the findings for Poland. Frenkel and Nickel report supply shocks in Poland to be negatively related with supply shocks in Germany, while Horvath finds close to zero correlation. Fidrmuc and Korhonen report supply shocks in Poland to be positively related to similar shocks in the euro area, but Frenkel and Nickel find a positive correlation. Also the correlations of the demand shocks in Poland and Germany vary considerably. Finally, Korhonen (2003) uses a somewhat different methodology. He examines monthly indicators of industrial production in the euro area and nine new member countries. VARs are estimated for each country. If a euro area shock is quickly reflected in production, this is considered as evidence of symmetry of the business cycles. Korhonen finds that some of the advanced new members (especially Hungary) exhibit

198

Chapter 6

Table 6.12 Correlation of exogenous shocks between euro area, Germany and accession countries (vector autoregressive estimations) Supply shocks Country

Germany a

EMU b

c -0.03 (-0.03) 0.04 (0.18) 0.25 (0.02) 0.46 (0.20) 0.30 (0.16) -0.11 (-0.31) 0.08 (0.67) 0.02 (0.22) 0.05 (-0.09) 0.15 (0.19)

Bulgaria

0.462

—

Czech Republic Estonia

0.538

-0.05

0.036

0.08

Hungary

0.263

0.28

Latvia

0.022

-0.07

Lithuania

—

Poland

-0.494

Romania

—

Slovak Republic Slovenia

Demand shocks

-0.16 0.00 —

0.384

-0.04

0.434

0.02

Germany a

a

EMU b

c

a

—

0.03 (-0.14) -0.15 (-0.06) 0.12 (-0.00) 0.25 (0.20) -0.49 (0.02) -0.49 (-0.35) 0.28 (0.05) 0.03 (-0.06) -0.05 (0.05) -0.18 (-0.21)

-0.224

0.280

0.250

0.052

0.321

0.10

0.339

0.343

0.05

0.726

-0.197

-0.40

0.333

0.260

0.11

—

—

0.33

-0.690

-0.200

0.14

—

—

—

0.182

-0.097

0.04

0.658

0.049

0.03

-0.213 -0.241 0.122 -0.428 — 0.217 — -0.433 -0.147

Source: update of Horvath (2003). a. Frenkel and Nickel (2002); quarterly data for the period 1993:I to 2001:IV. b. Horvath (2000); quarterly data for the period 1993:I to 2000:III, for Hungary 1995:I to 2000:III. c. Fidrmuc and Korhonen (2003), quarterly data for 1991 to 2000. In parentheses the updates of Fidrmuc and Korhonen (2004) are shown; the latter refer to the period 1991 to 2002.

a high correlation with the euro area business cycle. Moreover correlation seems to be at least as high as in some smaller countries in the euro area like Portugal and Greece. Table 6.13 summarizes all studies discussed so far. The main message stemming from the analysis is that most new member countries are subject to different macroeconomic shocks—and thus a different business cycle—than the current euro area. Although according to some estimates the more advanced new members have business cycle

Period (longest)

Comparison

Method

Conclusion

CZ, ES, HU, LA, LT, PL, SL, SI CZ, HU, PO, SL, SI

1993:I– 2000:III

FR, GE, IT, UK

Correlation of shocks determined by SVAR

1991/3:01– 1999:12

GE

Frenkel and Nickel (2002)

BU, CZ, ES, HU, LA, PL, SL, SI

1993:I– 2001:IV

EMU as a whole and individual countries

Correlation of detrended industrial production and GDP Correlation of shocks determined by SVAR

Fidrmuc and Korhonen (2003)

BU, CZ, ES, HU, LA, LT, PL, RO, SL, SI

1991:I– 2000:IV

Euro area and Euro area countries

Correlation of shocks determined by SVAR

Berger et al. (2004)

CZ, CY, ES, HU, LA, LI, PO, SL, SI CZ, ES, HU, LA, LT, PL, SL, SI

1990:01– 2001:12

Euro area

1993:1– 2002:12

AU, GE, IT, Euro area

Correlation of detrended industrial production (HP) Stylized facts analysis; correlation of detrended industrial production and GDP

Low correlation with GE; HU and SI have the highest supply shocks correlation, while LT has the highest demand shock correlation. High correlation notably for HU, SL (only for GDP), SI; for CZ correlation of GDP is almost zero Wide variety of results, correlation of supply shocks somewhat higher than those of demand shocks. Country with highest supply (demand) shock correlation with EMU is HU (PO); most countries have negative demand shock correlation. Wide variety of results, correlation of supply shocks somewhat higher than those of demand shocks. CR, LA, LT, SL, SI have negative demand shock correlation. Country with highest supply (demand) shock correlation is HU (PO). Low correlations except for CY and SI.

Study

Countries

Horvath (2000) Fidrmuc (2001)

EFN (2004)

New Member Countries

Table 6.13 Summarizing recent literature on business cycle synchronization: New member countries

High correlation of industrial production with euro area, notably for HU, PO, SI; negative for LA, LT

199

200

Table 6.13 (continued)

Comparison

Method

Conclusion

BU, CZ, ES, HU, LA, LT, PL, SL, SI

1996:01– 2002:6

Euro area, GE, Big 5

Correlation of detrended industrial production and GDP (HP); correlation of shocks determined by SVAR

Fidrmuc and Korhonen (2004)

BU, CZ, ES, HU, LA, LT, PL, RO, SL, SI

1991:I– 2002:II

Euro area

Correlation of shocks determined by SVAR

Korhonen (2003)

CZ, ES, HU, LA, LT, PL, RO, SL, SI CZ, ES, HU, LA, LT, PL, SL, SI

1992:3– 2000:12

Euro area

1993:I– 2002:IV

Euro area

Correlation of VAR impulse functions, industrial production Several measures of synchronization, including correlation of detrended GDP and components

High correlations with euro area for GDP for HU, PO, SI; negative for SL, CZ, LT. For industrial output HU, PO and ES have highest correlation with euro area; LT and RO have negative correlation. Demand shocks negatively correlated with euro area in CZ, SL, LA, LT, ES. Highest supply shock correlation in HU, PO, SI. Low-demand shock correlations (negative in BU, CZ, LT, SI, RO) higher supply shock correlations, notably for PO; negative for BU, LT, SL; lower correlations than in Fidrmuc and Korhonen (2003a) Correlations vary widely, but they are comparable with those of Greece, Ireland, and Portugal. HU, PO, SI have achieved a high degree of synchronization with the euro area for GDP, industrial production and exports, but not for consumption and services. The other accession countries have achieved less or no synchronization. All accession countries have minimal or negative correlation with the euro area cycle of private consumption and services.

Countries

Süpel (2003)

Darvas and Szapáry (2004)

Note: AU = Austria, BU = Bulgaria, CZ = Czech Republic, CY = Cyprus, ES = Estonia, FR = France, GE = Germany, HU = Hungary, IT = Italy, LA = Latvia, LT = Lithuania, PL = Poland, RO = Romania, SL = Slovakia, SI = Slovenia, UK = United Kingdom, HP = Hodrick-Prescott filter, SVAR = structural vector autoregressive model, VAR = vector autoregressive mode.

Chapter 6

Period (longest)

Study

New Member Countries

201

correlations comparable to those of some current members in the euro area, it is clear that enlargement of the monetary union will imply more cyclical diversity among the countries in the union at least in the short and medium term. The fact that business cycle synchronicity in some of the new member countries seems to be similar to that of some small countries in the euro area only reinforces this conclusion. In the absence of reform of the present ECB framework, this could have an adverse impact on monetary policy making in the euro area, as was pointed out in chapter 5. 6.3.2 Inflation It is often argued that due to the Balassa-Samuelson effect, transition countries experience an appreciation of their real exchange rates. As a consequence of economic restructuring, many transition countries experience rapid productivity growth in their industrial sectors. As productivity growth in the traded goods sector exceeds that in the nontraded goods sector, nontraded goods prices should increase due to the wage equalization process between both sectors. When productivity growth in the transition countries exceeds productivity growth in the countries in the euro area, the transition countries should have a higher inflation rate. According to Eurostat (2001), average productivity in manufacturing in transition countries was only about 40 percent of the EU average in 1998. Therefore we can expect further high productivity growth. This restructuring will, however, take some time. During restructuring these countries will probably experience higher inflation than the current EMU countries. This raises two questions. First, how big are these inflation differentials between current and potential future EMU members? Second, what are the policy implications? There is clearly no consensus in the literature on the magnitude of the Balassa-Samuelson effect in the transition countries. Table 6.14 provides a summary of various recent studies. Estimates vary widely. Whereas Rogers (2001), for instance, estimates that the BalassaSamuelson effect is likely to imply two additional percentage points of annual inflation in the new member economies, Égert (2002a, b) finds little evidence of a higher inflation rate due to the Balassa-Samuelson effect in the Czech Republic and Slovakia. The extremely high inflation differentials implied by sectoral productivity developments and labor shares for Hungary and Poland as reported by Backé et al. (2002) attract attention. The figures of these authors reflect mainly the massive gains in productivity in the tradable-goods sector that were achieved during

202

Chapter 6

the 1990s in these two countries. They argue, however, that past figures are probably not a good guide for the future as convergence implies that productivity increases will decelerate as higher productivity levels are reached. Some of the diverging outcomes in the literature are the result of differences in method. An important factor is that not all studies summarized in table 6.14 are restricted to estimates of the Balassa-Samuelson effect. The cited studies have pointed out various other channels that can give rise to inflation differentials. Some authors take these into account. For instance, Halpern and Wyplosz (2001), who estimate the Balassa-Samuelson effect for a panel of nine transition countries (including Russia), include demand factors. The same is true for Coricelli and Jazbec (2001), who in addition add a variable that captures structural misalignments. A very different estimation procedure has been followed by Pelkmans et al. (2000). These authors base their estimation on relative price levels in new member countries compared to existing EMU member countries rather than on productivity growth differentials. The authors proceed in four steps. First, they regress the deviation of inflation rates of euro area countries from the euro area average on the relative consumer price levels of these countries. Next, they regress the relative consumer price levels of 29 OECD countries on the GDP-based comparative price levels of these countries (i.e., on ratios of GDP measured in PPP and at current exchange rates). The coefficients of the independent variables in both equations are negative and highly significant. In a third step, Pelkmans et al. (2000) calculate the relative consumer price levels of the 10 Central and Eastern European countries, based on their comparative price levels and the coefficient estimated for the OECD countries in the second equation. Finally, these authors use the coefficient estimated in the first equation for the euro area countries to compute the new member countries’ inflation differentials from the average euro area, which are implied by their relative consumer price levels. Their results show on average an inflation differential of 3.8 percentage points. As to the policy implications, the evidence reviewed suggests that at least according to some estimates of the Balassa-Samuelson effect, new member countries with a fixed exchange rate regime may have problems in meeting the inflation criterion of the Maastricht Treaty. Countries with a somewhat more flexible exchange rate regime are unlikely to have problems to meet the Maastricht criteria for BalassaSamuelson reasons. The Balassa-Samuelson effect is not likely to

New Member Countries

203

Table 6.14 Estimates of the inflation differentials (%) in the transition countries Study

Countries

Jakab and Kovacs (1999) Pelkmans et al. (2000) Rother (2000)

Hungary

Sinn and Reuter (2001) Halpern and Wyplosz (2001) Corizelli and Jazbec (2001)

De Broeck and Sløk (2001) Égert (2002a)

CEE 10 Slovenia Czech Republic Hungary Poland Slovenia Estonia Panel of 9 transition countries (including Russia) Panel of 19 transition countries

Vis-à-vis (if relevant)

1.9 29 OECD countries

Germany

Based on model for service-to-consumer goods price ratio Based on model for relative price of tradable goods

Panel of transition countries Czech Republic

Germany

Poland Slovakia Slovenia Panel of Czech Republic, Hungary, Poland, Slovakia, and Slovenia

3.8 2.6 during 1993–98 2.88 6.86 4.16 3.38 4.06 2.9–3.1 for the period 1991–99 1 in the medium term (1990–98)

On average 1.5

Hungary

Égert (2002b)

Size

Germany

0.648a 0.303 for 1991–2000 2.589 1.295 for 1991–2000 3.245 1.901 for 1991–2000 -0.154 -0.075 for 1993–2000 1.321 0.661 for 1993–2000 With share of nontradables as in GDP it ranges from 0.094 to 1.903 depending on time period and data. Estimates for 1996–2001 period range from 1.707 to 1.903. With share of nontradables as in CPI the latter range from 0.810 to 1.059.

204

Chapter 6

Table 6.14 (continued) Study

Countries

Backé et al. (2002)

Czech Republic, Hungary, Poland, Slovenia

Vis-à-vis (if relevant) Main trading partnersb

Size 0.35 1995–2000 3.84 1995–2000 9.76 1995–2000 3.88 1995–2000

a. First figure shows results using GDP deflator; second figure shows results with CPI. b. Under the assumption that there are no productivity-inflation differentials between tradable and nontradable goods in the main trading partners, which seems unrealistic.

exhaust the 15 percent bands of ERM II in two years (Buiter and Grafe 2002). If the lower estimates of the Balassa-Samuelson effect are closer to the truth, new member countries with fixed exchange rate regimes will meet the inflation criterion. In fact the countries with a currency board have the lowest inflation rates among the new member countries. Some observers have argued that the convergence criteria should be modified (e.g., see Coricelli and Jazbec 2001). One could, for instance, compare the inflation rates of the new member countries with those in the least developed EMU countries or allow for a higher than the 1.5 percentage point differential. These suggestions have met little support from the current monetary union members. Admitting countries with relatively higher inflation rates could increase the HICP inflation in the euro area. However, this argument should not be overstressed as the weight of inflation in the new member countries in the total euro area inflation rate is quite low (about 6 percent). For instance, a 3 percent difference in inflation rates between the 1998 Accession Group and the rest of the euro area would only imply a 0.1 percent increase in the euro area’s GDP-weighted inflation (Égert 2002a). More important, however, is that the increase in the dispersion of inflation rates in the euro area may increase the risks implied by the decentralized setup of the ECB (see chapter 5). Since the catch-up process of the new member countries will continue after they have joined EMU, the enlargement of the monetary union may in time display more inflation divergence within the union. If national considerations play a role in the behavior of national central bank governors in the Governing Council of the ECB, it is more likely that the focus on

New Member Countries

205

euro area-wide developments will be undermined. From this perspective, the future enlargement of the monetary union only underscores the need for reform of the ECB. 6.3.3 Monetary Transmission and Financial Structure As pointed out before, differences in financial structure may lead to diverging transmission of monetary policy measures. Although the evidence for the current euro area is rather mixed (see section 5.6.2), many authors have argued that the important differences between the current and future euro area countries in terms of their financial systems may lead to differences in monetary transmission between both groups. As far as the new member countries are concerned, there is only scant evidence on differences in monetary transmission among countries. Hardly any attempts have been made to compare the transmission mechanisms of these countries with one another.16 The evidence generally relates to just one or a few countries, while the link with financial structure is often absent in these studies. Economic research on monetary transmission in transition countries is hampered by three major constraints. First, a lack of usable data caps the number of econometric tools that can be used. Second, the quality of the available data is often low. It is well known that many, if not most, macroeconomic time series are subject to measurement errors of unknown importance, and this seems to apply, in particular, to transition countries. Third, a highly dynamic economic environment makes it difficult to separate the effects that stem from a specific development. Conditions for research will slowly improve over time as time series data accumulate and as the reform process associated with postsocialist transition and the preparations for EU accession come to an end. There are three main approaches used to investigate monetary transmission in transition countries (Ganev et al. 2002). The first research method includes rather unsophisticated comparison techniques by which general inferences are made on what might have happened under different circumstances. A second, more formal approach involves the construction of (small) structural macroeconomic models. These models are used to investigate the development of certain economic indicators under specific monetary conditions in a highly stylized environment.17 The third research method is based on vector autoregressions (VARs). This type of econometric modeling generally enables the researcher to limit the number of strong theoretical constraints used to investigate the effects of monetary policy.

206

Chapter 6

Studies in which the VAR methodology has been used include Durjasz (2001), Gottschalk and Moore (2001), and Christoffersen et al. (2001) for Poland, and Kuijs (2002) for the Slovak Republic. Maliszewski (2002) compares monetary transmission in Poland and the Czech Republic, while Ganev et al. (2002) present VARs for ten transition countries. Unfortunately, the various studies cannot be directly compared because they refer to different sample periods and also employ rather diverse modeling strategies. Still, despite the wide variety of models, some general conclusions can be drawn from these studies (Elbourne et al. 2003). Not surprisingly, many studies find evidence that the monetary transmission mechanism is unstable over time. Furthermore the reported impact of monetary policy measures is often counterintuitive. Durjasz (2001), for instance, reports that the Polish experience suggests that only after the implementation of direct inflation targeting (which was introduced in Poland in late 1998) reasonable transmission patterns emerged. Another conclusion that is found in many papers is that the exchange rate mechanism is more powerful in various transition countries than the interest rate channel. In countries that have relatively flexible exchange rate regimes, the exchange rate channel is the dominant way in which monetary contractions affect inflation. Gottschalk and Moore (2001), for instance, find that prices respond usually within a year to a shock in the nominal exchange rate. The price level does decline in response to an interest rate hike, but this effect is not clearly significant. Also the evidence for the Slovak Republic reported by Kuijs (2002) suggests a strong exchange rate channel. Although there is broad agreement about the importance of the exchange rate channel, there is less agreement over whether this dominance remains in the more recent period. According to Hamecz (2001), the dominance of the exchange rate channel under the tightly managed exchange rate regime became less clear after the Hungarian central bank switched to inflation targeting in mid-2001. However, Maliszewski (2002) concludes that the exchange rate channel is still dominant in Poland and the Czech Republic. One of the most comprehensive VAR studies is by Ganev et al. (2002), who report impulse responses of industrial output, inflation, and the exchange rate to a one standard error interest rate shock involving ten transition countries. Their study only covers the period January 1995 to December 2000, however. The positive short-term interest-rate shock brought about very different reactions of industrial output in different

New Member Countries

207

countries. It dampened output in the short run in Slovakia, Hungary, and Slovenia while it raised it in Lithuania, Estonia, the Czech Republic, and Poland. Latvia, Bulgaria, and Romania had a mixed pattern. The impact seemed to ease off after 12 months in most countries. Likewise the core inflation response to interest rate shocks varied across countries. In some, namely Lithuania, Hungary, and Slovenia, the response was consistent with theory in that higher interest rates dampened inflation. In Bulgaria after initial boost, inflation subsided. In Slovakia and the Czech Republic an interest rate shock raised inflation persistently and led to higher inflation over the three years following the shock. In Romania there was an instability problem—the core inflation is still on the rise after three years. Also Elbourne et al. (2003) estimated VARs for ten transition countries. Apart from inflation and output growth, these authors use the deposit rate and the lending rate as interest rates in the models. For those countries that had either a currency board or a fully fixed exchange rate for a large proportion of the sample period, they modeled the exchange rate as an exogenous variable. For the others, that had either floating exchange rates or crawling pegs, the exchange rate is endogenous. All variables are defined in the form of their deviation from their stochastic trend. German output growth, the German call money rate, and commodity prices are included as exogenous variables, as are dummy variables to take into account the effects of the differing exchange rate regimes and financial crises. Although the results of this study differ from those of Ganev et al. (2002), Elbourne et al. (2003) also conclude that there appear large differences in monetary transmission among the countries considered. Table 6.15 is reproduced from this study. It shows the effects on inflation and output growth of a 100 basis points shock to the domestic deposit rate. The table shows the peak impact of inflation and output growth after the monetary shock, the time (in months) it takes before this maximum is reached, as well as the cumulative impact of the shock over a three-year period. The financial structure of the transition countries differs from that of the current monetary union members. As banks play an important role in the transmission of monetary policy, table 6.16 shows a number of key indicators for the banking system of the transition countries. Bank assets in relation to GDP are almost three times higher in the euro area than in the transition countries. Of course, this does not imply that the credit and bank lending channels are less important in the new member countries, as capital markets are still in their infancy. As Buiter and Taci

208

Chapter 6

Table 6.15 Summary of impulse responses of VAR model for 10 transition countries Inflation

Output growth

Country

Peak impact

Time (months)

Sum 36 months

Peak impact

Time (months)

Sum 36 months

Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovak Republic Slovenia

-5.30 -0.70 -0.18 -0.28 -0.46 -0.84 -0.18 -0.20 -0.28 -0.08

3 7 10 1 5 7 2 5 4 2

-16.38 -3.09 -1.78 1.43 -4.71 -8.51 -1.61 -1.89 -1.30 0.01

-0.12 -2.25 -0.58 -2.07 -0.87 -0.71 -0.64 -0.13 -0.59 -0.11

5 5 5 3 2 1 6 1 2 2

-0.32 -6.47 -2.20 -9.92 -4.49 -1.30 -5.93 -0.25 3.26 -0.89

Source: Elbourne et al. (2003).

(2002) argue, the financial systems in the transition countries have developed more as bank-based systems than as market-based systems. Given the insufficient scope and effectiveness of legal contract enforcement and with inappropriately or imprecisely defined property rights, these countries had no alternative but to develop a relationship-based financial system, with banks as the main financiers. Comparative research on the relationships between the financial structure and monetary policy transmission in the transition countries is scarce. Elbourne et al. (2003) use a similar approach to that suggested by Cecchetti (1999). They gathered various indicators for the financial structure in the transition countries and put them into three categories: indicators for the importance of small banks in a country’s financial system, indicators for the health of the banking system, and indicators for the importance of external finance. In contrast to Cecchetti (1999), they do not combine the indicators into one single financial sector indicator as this is a rather subjective and ad hoc procedure. Instead, Elbourne et al. (2003) use rank correlation coefficients of the estimated impact of monetary policy decisions as reported in table 6.14 and the various financial structure indicators. They find no clear evidence of a relation between financial structure and the impact of monetary policy shocks as most of the correlations are not significant.

Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovak Republic Slovenia

Assets share of five largest banks (%)

Loans share of five largest banks (%)

Deposits share of five largest banks (%)

Domestic credit of banks (% GDP)

Nonperforming loans (as % of loans)

Average capital ratio

Net interest margin

Average rate of return on assets

60.5 66.1 98.8 53.3 62.3 88.5 48.6 70.1 63.4 62.5

NA 67.5 99.5 52.9 71.9 85 48.4 65.4 63.8 49.1

NA 74.5 99.5 61.5 66.2 93 49.1 NA 69.1 53.2

25.6 56 38.1 35.2 21.7 16.8 36.5 8.9 61.4 44.6

10.9 19.3 1.5 3.1 5 10.8 15.9 3.8 26.2 8.5

0.1 4.6 12.6 8.8 8.4 10 8.2 NA 6.4 8.1

4.1 2.1 0 3.7 4 3.6 4 7.4 1.8 4.2

4.1 0.5 1.1 1.3 2 0.4 1.1 2.3 0.5 1.1

New Member Countries

Table 6.16 Some indicators of the banking sector in transition countries (2000)

Source: Buiter and Taci (2002).

209

210

Chapter 6

6.4 Conclusions EMU is being enlarged over time. Some of the new member countries will join EMU some two years after their EU entrance. Central bank independence is a prerequisite for EU membership. The new member countries must grant their central banks in legal terms a very independent position. However, the analysis of actual independence, proxied by the turnover rate of central bank governors, sometimes suggests deviations from the law. All new EU members have an obligation to join the EMU, but a new member has to fulfill the convergence criteria stipulated by the Maastricht Treaty before it can enter. This is a way of ensuring equal treatment of member countries in the euro area. The adherence to the convergence criteria can be problematic. One critical issue is the need to fulfill the budget deficit criterion. Another is achieving exchange rate stability, by which the currency must respect the “normal” fluctuation margin of the ERM II. All new member countries have to adhere to the inflation criterion. Recent estimates of the Balassa-Samuelson effect do suggest that the inflation criterion can probably be fulfilled. Quick entry is a sensible strategy from the perspective of the new members, since there are benefits involved. One is a lower interest rate after entering the EMU, which makes budgetary consolidation easier to accomplish. The shorter the period accession countries must wait, the lower are the risks of exchange rate instability. For the accession countries with currency boards that have demonstrated over the years that they can live with a unilaterally fixed exchange rate, the wait is especially difficult. As the monetary union enlarges, reform in the decision-making process of the ECB will become more critical. The disparate business cycles, inflations, and monetary transmissions could increase within a larger monetary union. Most new members have business cycles that differ from that in the euro area. Over the long run these idiosyncrasies will probably be lessened as a consequence of full convergence, but the idiosyncrasies will certainly not disappear in the short and medium term. This impact will be felt in monetary policy making in the euro area, but a reform of the ECB framework could avoid this. Although interpretations of the so-called Balassa-Samuelson effect differ substantially, the new EMU members can be expected to have higher inflation levels than the older members. While this does not seem particularly serious in the impact on euro area wide inflation, the

New Member Countries

211

increased divergence of inflation within the euro area may cause the decision making of the ECB to become more focused on individual cases than on inflation in the euro area as a whole. A lot of authors have already argued that the big differences in financial systems between current and prospective euro area countries can lead to differences in monetary transmissions between both groups. There is agreement over the importance in the past of the exchange rate mechanism compared to the interest rate channel, but there is less agreement over whether this dominance will continue. In the financial structures of the transition countries, banks play an important role in the transmission of monetary policy because capital markets are still in their infancy. Although there is mixed evidence that the impact of monetary policy shocks depends on financial structure, enlargement of the monetary union may imply more structural diversity in monetary transmission within the union. So ECB policy decisions may have different impacts on the various countries comprising the union. Our findings in this chapter therefore support our argument in chapter 5, that a reform of the ECB in which economic and political weights are brought more in line with each other is needed. In chapter 7 we discuss some ways to reform the ECB.

7

Options for Reform

7.1 Introduction The Governing Council of the ECB currently has 18 members—6 Executive Board members and 12 national central bank governors, one for each of the 12 euro area countries. In principle, all members have equal weight in the decision-making process. However, eighteen members have already proved too many from the point of view of effective discussion, deliberation, and decision making. Enlarging an unreformed ECB to include up to some 15 additional national central bank governors could turn the Governing Council into an unwieldy and unmanageable group of 33 members. The central bank’s tradition of consensus-based policy making—said to have an important place in today’s ECB decision-making process—could further amplify the ECB’s numbers problem and raise decision-making costs. Baldwin et al. (2001) argue that the practice of the Executive Board to initiate Council decisions will become seriously impaired as the number of euro area member countries increases and the Board’s relative power decreases. Likewise the increase of euro area member countries without reforms could broader the wedge between the economic and political weights of countries in the euro area (see figure 7.1). Since nearly all accession countries are small in economic terms relative to current euro area members, enlargement within the given institutional setup may lead to significant overrepresentation of the area’s smaller member countries in the Council. For instance, in a monetary union with 27 members the current ECB statute implies that the representatives of its smallest 17 member states, representing only about 10 percent of the area’s aggregated GDP, can determine monetary policy for the entire euro

214

Chapter 7

MLT EST LVA CYP LTU BGR SVN LUX SVK RO HUN CZE IRL PRT GRC FIN DNK POL AUS BEL SWE NLD ESP ITA FRA GBR GER

Political weight (PW)

Economic weight (EW)

0

5

1 Council seat

10

15

20

25

Share (%)

Figure 7.1 Economic and political weights of central banks after enlargement of the euro area. (Source: Berger et al. 2002)

area (Berger 2002). Without a reform of the ECB, nearly 80 percent of the countries will have larger political than economic weight. If the “one person, one vote” principle is strictly applied, all newcomers but the United Kingdom will be allocated a political weight that surpasses their economic weight, and in most cases by a substantive margin (Berger et al. 2004). Overrepresentation can introduce an unwelcome bias into the ECB’s decision making if country representatives try to put some weight on national economic developments and these developments deviate significantly from the behavior of euro area aggregates. As we have argued in chapters 5 and 6, there is reason to believe that such asymmetries can affect ECB policy making. In this chapter we will therefore evaluate various ECB reform options from three perspectives: (1) decision-making costs, (2) the gap between economic and political weights, and (3) political feasibility. Section 7.2 discusses various reform options, while section 7.3 zooms in on the proposal put forward by the ECB. Section 7.4 concludes.

Options for Reform

215

7.2 Options for Reform Among the ways of reform voiced for the ECB (see Berger 2002), are centralization (the Executive Board will become responsible for policy decisions), vote-weighting (the vote of a national central banker depends on the size of the economy), representation (one central banker represents various central banks), extending regional central banks across national borders, and rotation (the governors of national central banks have rotating voting rights). Baldwin et al. (2001) argue for more centralization. The pragmatic version of centralization would put a policy decisions into the hands of the ECB Executive Board. This, of course, would limit the role of the Council to that of an informational forum in which the area’s regional central banks would be apprised of policy decisions and implementation issues would be discussed. A larger role for the Executive Board in ECB decision making can go a long way toward limiting decision-making costs and preventing any diverging economic developments within a larger euro area to have an undue impact on monetary policy in the euro area. The EU Treaty specifies that the Board is appointed by “the governments of the member states at the level of Heads of State or Government, on a recommendation from the Council, after it has consulted the European Parliament and the Governing Council of the ECB” (EU 1997, Article 112.2b). This highly centralized political process at the European level should support a euro area-wide perspective of the nominees selected for the Board. However, the political feasibility of centralization seems limited. The principle “one person, one vote” is an important feature of the current ECB framework also in day-to-day monetary policy making. As the ECB Statute clearly states, “Each member of the Governing Council, shall have one vote” (EU 1992, Article 10.2), and this includes the national central bank governors. An equal right to participate in ECB policy decision making was thus an integral part of the Maastricht Treaty that established the currency union. Clearly, a reform of the ECB that fails to safeguard the established voting rights of current member countries’ central banks will not be politically acceptable. Some member governments have already experienced opposition in letting go the “last” bit of influence on ECB policy making after having exchanged monetary sovereignty for a seat at the ECB Council in 1999. A version of this scheme that only slightly departs from full centralization and meets some of these objections was put forward by Gros

216

Chapter 7

(2003), who argues in favor of redefining the division of labor between the Executive Board and the Governing Council. The function of the Governing Council would be exclusively in the direction of monetary policy, whereby the Council would decide on proposals from the Executive Board, constitute a platform for the exchange of views on the euro area economy, and monitor the work of the Executive Board. These are tasks that can be performed efficiently by a large body composed of representatives of all member countries, and the Governing Council provides the appropriate legitimacy for such a controlling institution. Gros sees the primacy of the Governing Council as not affected, in that all powers would continue as mandated by the EU Treaty. This plan does, however, reduce the right of the Governing Council to control every single act of the Executive Board. Thus the Executive Board would come to enjoy a certain degree of discretion, as is justified by the fact that it represents not only an aggregation of individual state interests but a “general European monetary interest.” According to Gros, this division of labor is based on one key difference between NCB presidents and members of the Board: their respective information bases. Board members concentrate on area-wide aggregates in their daily work and are likely to be in closer contact with global financial markets than the NCB presidents. The latter perform a wide variety of functions at the national level: they supervise the national banking system, they are influential participants in national debates about almost all economic policy issues, and so on. By contrast, the members of the Board can concentrate almost exclusively on issues related to the formulation of the common monetary policy stance. Although we are quite sympathetic toward this reform proposal, we doubt that it can sufficiently address the potential gap between political and economic weight. This is because in the Gros (2003) proposal all national central bank governors keep one vote in the Governing Council and the Council still decides on the direction of monetary policy. In a scheme proposed by Bofinger (2003), a similar setup is favored whereby the Executive Board takes decisions on interest rates and the Governing Council is allowed veto power over interest rate decisions. This scheme only brings us back basically to the current situation. Furthermore it seems that the political feasibility of reform to give the Executive Board more power is limited. Berger et al. (2003) suggest a type of reform in which economic size and political power are matched as closely as possible. When countries

Options for Reform

217

have as much voting power as GDP share, deviations from a purely “European” perspective by Council members would not have an undue influence on monetary policy in the euro area. There are basically four options here: vote-weighting, representation, extending regional central bank jurisdictions across national borders, and rotation.1 For instance, under vote-weighting, the votes of non-Board members of the ECB Council would be weighted in terms of a member countries’ share in euro area GDP. By definition, a reform along this line could better align the political and economic weights of the national Council members. Vote-weighting has a precedence in the qualified voting schemes of the EU Council, which the Treaty of Nice updated in the first EU enlargement. Yet another voting scheme that takes into account differences in economic size is the idea of a required “double majority” of votes and population. Under such a system there is still an equal voting right for all Board members. Every decision requires a majority of the votes. In addition, however, it is also required that the votes in favor represent a majority of the population of the euro area. An alternative would be to require that these votes represent a majority of the euro area’s GDP. A problem with any of these vote-based reform scenarios is clearly that they do not necessarily address the problem of decision-making costs. Decision-making costs in the narrow sense of voting on interest rate changes, say, do not need to be made so problematic. The Council’s decision-making process involves more than a simple aggregation of votes, it also includes a more or less extensive discussion of the views of all members. In this regard weighting votes does not necessarily solve the ECB’s “large number problem.” In fact, just like the argument made in opposition to the centralization solution, it should be noted that a weighted-voting scheme can interfere with the “one person, one vote” principle embedded in the ECB Statute, although somewhat less than may apply to the “double majority” system. An alternative reform scenario is representation that combines some of the characteristics of the centralization and the weighing approaches. The principal idea is to create groups of euro member countries with joint representation and joint voting rights in the ECB Council, integrating the concept of a strong regional anchor with the necessity of restricting the size of the ECB’s main decision-making body after the enlargement.2 The representation scenario requires a number of

218

Chapter 7

specific institutional decisions, in particular, on group selection. The selection principle is based on the idea of common economic regions (taking into account similarities in business cycles or economic structure), economic size, or both. Related issues are the number of groups, the overall Council size, and the delegation of voting power from group members to their representative in the ECB Governing Council. The alternative institutional designs range from a restricted or “imperative” mandate (votes in the Council are predetermined at the group level) to an unrestricted mandate (group members delegate their full voting rights to their representatives). However, since the latter arrangement can, in principle, deprive individual group members of their right to participate in the decision making, there is a potential conflict with the idea of national representation and the “one person, one vote” principle. This makes a solution entailing some form of explicit involvement of national central banks at the group level before a Council decision (i.e., a restricted mandate for the group representatives in the Council) a likely part of any representation scenario. Such a restriction is likely to encompass contributions to Council discussions as well as formal voting. In this sense it can alleviate the decision-making costs problem at the level of the ECB Governing Council. However, the costs will substantially increase at the level of the group. If the mandate of group representatives in the Council is restricted, in the sense that their actions require the explicit consent of group members, the overall time and effort needed for a Council decision will be of a similar magnitude, if not higher, as in the previously discussed scenario. A variant of the representation idea is the extension of central bank jurisdictions across national borders. For instance, the regional central banks in the US Federal Reserve System extend their reach over the borders of some states. Likewise some of the (post-1992) Landeszentralbanken in the Bundesbank Council represent more than one German state. An application of this principle to the ECB after enlargement could help reduce the number of decision makers in the Council. If the design of central bank areas were aimed at establishing regional banks with approximately similar economic weight, significant mismatches could be avoided between voting power and economic size. However, as with the previous scenarios, there can be problems of political feasibility where a reform works to abolish the existing voting rights of current euro area member states. Furthermore implicit in such a reform is that one of the basic principles of the current ECB setup, “representation” of countries, would be abandoned.

Options for Reform

219

An alternative reform scenario that, in principle, might address both the mismatch between political and economic weights and the decision-making problem associated with the enlargement of euro area membership (while avoiding some of the political constraints discussed above) could be (asymmetric) rotation. The basic idea is that national central bank governors would take turns sitting at the Council, with the frequency of their participation scaled to match the relative economic weight of their respective country. Rotation would thus work to weight the votes of national central bank governors in an implicit fashion. Arguably rotation poses less conflict with the “one person, one vote” principle than centralization, weighted-voting, or the representation scenario. Although not all governors can participate in every Council meeting, those who participate will be casting a full vote. Rotation could also serve to limit the overall size of the ECB Council by allowing only a fraction of central bank governors to participate in meetings. The ability to address the potential problems posed by enlargement while avoiding part of the political feasibility problems associated with some of the other types of reform makes the rotation scheme a good structural solution for ECB. Berger et al. (2004) recognize that in basing representation on GDP size, more than 50 percent of the euro area’s GDP is guaranteed to always have a voice in the Council. Again, the principle of equal voting rights is lost. The easiest way to ensure that always more than half of the euro area’s GDP is represented is to give the “Big Five” a permanent seat in the Governing Council. If this turns out to be politically unacceptable, one could consider a system in which the Big Five also rotate, but substantially less than the other countries. Berger et al. (2004) present the example of four seats for the Big Five, who rotate, and five seats for the others, who also rotate. 7.3 The ECB Proposal In its meeting of December 19, 2002, the Governing Council of the ECB adopted a proposal for ECB reform after enlargement of the monetary union.3 As in the analysis of Berger et al. (2004), the ECB proposal put a limit on the number of central bank governors exercising a voting right. However, the ECB put this maximum at 15, instead of 9.4 Consequently the Governing Council will consist of 21 members, which may be too large from a decision-making cost perspective. No modern

220

Chapter 7

central bank has a decision-making body this size. Moreover all members of the Governing Council (with and without voting rights) will continue to sit at the table and have the right to participate in the discussion. As Bofinger (2003) puts it, “in spite of its complexity the Recommendation clearly fails to meet the main target of the ECB’s institutional reform. The danger that the Council would be paralysed by too many participants is still there.” The ECB proposes that if the euro area increases to more than 15 countries, there will be two groups with rotating voting rights. The first group will consist of the governors of the member states that occupy the highest positions in the country rankings on the basis of a so-called composite indicator of “representativeness.” They share four voting rights. The second group will consist of all other governors, and they will share 11 voting rights. The principal component of the “representativeness” indicator will be the member state’s GDP. The second component will be the total assets of the aggregated balance sheet of monetary financial institutions (TABS-MFI) within the territory of the member state concerned. The relative weights of the two components are 5/6 for GDP and 1/6 for TABS-MFI. Once there are 22 euro area members, there will be three groups with rotation. The allocation of central banks to the groups will be based on a ranking according to the composite indicator. The rotation scheme as proposed by the ECB is as follows: The first group, which will have four votes, will be composed of the five central bank governors from the euro area member states that occupy the highest positions (the Big Five). The second group, with eight voting rights, will consist of half of all national central bank governors selected from the subsequent positions in the ranking. The third group will be composed of the remaining governors and will share three voting rights. Thus, when there are 27 members, the intertemporal voting power of a national governor will be 80 percent in the first group, 57 percent in the second, and 38 percent in the third (Bofinger 2003). The ECB proposal met considerable criticism from academic observers. Gros (2003) argued, for instance, that “the solution proposed by the ECB is worse than the status quo. It is inefficient, opaque, internally inconsistent and arbitrary.” Apart from critique on the size of the Governing Council that we share, Gros had the following objections to the ECB proposal: First, he saw it as compromising the principle of

Options for Reform

221

equality of member states, and thus as potentially undermining the idea that members of Governing Council should ignore the particular interests of their home country and act only in the interest of the entire euro area.5 As we have argued before, the best way to ensure that national interests will not unduly influence ECB policy making is to bring political power and economic weight of national central bank governors as closely in line as possible. Even if national central bank governors take the economic needs of their home countries into account, this will not necessarily lead to decisions that are much out of line with the ECB’s mandate on price stability in the euro area as a whole. Some preliminary calculations under the assumption of a 27 member EMU by Berger et al. (2004) suggest that the average share of GDP represented is high under the reform as proposed by the ECB: 73 percent. Reform along these lines can be expected to overcome the most important institutional design failure of the ECB. Second, Gros saw the proposal as lacking clarity. Mainly, it was not clear to him how the first group of 5 countries will share four votes. Will they rotate every meeting, every month, every year? By what order? What happens to new members of the euro area? We agree that the ECB proposal has to be better specified, but is seems that these practical matters do not give good reason to object to the proposal as a whole. Third, Gros argued that the ECB proposal is not transparent because it is too complicated. He saw as arbitrary the weight given to the indicator of the size of financial markets (one-sixth) and as designed to ensure a better position for one country, Luxembourg. This way, according to Gros, Luxembourg is granted larger weight than Finland, a country with about ten times the population and six times the GDP of Luxembourg. The third group with the lowest voting power would consist exclusively of the new members. We agree that the criteria for determining the voting groups is rather arbitrary. As we pointed out before, we prefer grouping on the basis of economic size (GDP) only. Furthermore there should be a clear rule on how often the grouping can be reconsidered to take into account that the relative size of countries may change over time. If the new member countries grow faster than the current euro area countries, they should get a higher voting share. Finally, Bofinger (2003) argued that the ECB reform proposal would undermine the independence of the ECB:

222

Chapter 7

[F]rom the experience of the Bundesbank one can see that it is of decisive importance for the political independence of a federatively structured central bank system that all regional/national representatives are treated equally in all respects. This insight is expressed by the principle of “ad personam participation.” It implies that a NCB governor is not regarded as an agent of its specific region/country but as an expert on monetary policy which takes part in the discussions and decisions in personal and independent capacity. The German experience shows that such an institutional set-up leads to a “Thomas Becket” effect according to which governors from very different political and regional backgrounds soon feel a responsibility for the currency area as a whole. . . . Under the new rules a NCB governor would permanently experience that his national origin plays a role in the decision-making process. This could make him prone to decide from a national perspective which threatens the independence of the ECB’s decision-making process from national considerations.

However, as we showed in chapter 5, there is evidence that voting in the Bundesbank decision-making body is influenced by regional economic considerations (Berger and De Haan 2002). The best way to mitigate any political influence that is at odds with the aim of price stability in the euro area appears to be to align political and economic weights as closely as possible. If politicians put pressure on their national central banks to improve the economic situations of their countries, and the central bankers act upon this, there should not be distorted decision making. 7.4 Conclusions Clearly, the ECB should be reformed in view of the enlargement of the euro area. Under the current setup the Governing Council would become excessively large. From the decision-making cost perspective, a larger role for the Executive Board would be an attractive option for reform. If the Board also were to become responsible for monetary policy making, the risks of a distorted monetary policy would be minimized, provided that the Board has a truly euro area-wide focus. If, however, the Governing Council remains the decision-making body, it should be reformed to bring the political power and economic weights of national central bank governors in line as closely as possible. From this perspective, the rotation system proposed by the ECB is in the right direction. In this system the likelihood that a national central bank governor will have voting right depends on the size of the economy of his home country. However, potential problems remain concerning the

Options for Reform

223

Council size (15 national central bank governors with voting rights and 6 Executive Board members) and the participation of all central bank governors in the discussions of the Council. In our view, a smaller Governing Council, with about 15 members, in which only national central bank governors with a voting right participate could lead to more efficient decision making than the scheme proposed by the ECB.

8

Conclusions

In this book we have focused on two sets of issues: the transparency and credibility of the ECB, and the decentralized setup of the central bank in the euro area. We showed that the ECB has a relatively good overall score on the Eijffinger-Geraats disclosure indicator, which is due to its good scores for political and economic disclosure. The ECB also has a high score according to most other indicators for central bank disclosure. However, an important finding of many surveys, including our own, is that the ECB is widely considered to be nontransparent by financial market participants. The ECB is also considered to be less credible than the Fed or the (previous) Bundesbank. Our indicator for accountability suggests that the ECB does not rank high but does go further than required by law in providing information. The ECB has been criticized by many acedemics, notably for its monetary policy strategy. Although we agree with some of the criticism, we believe that the ECB has pursued a successful policy so far. Inflation in the euro area has remained low, although it is often slightly above rather than below the 2 percent mark. The projected inflation rate has been relatively stable. Most ECB policy decisions have been in line with market expectations, although there have been occasional surprises. Despite this favorable assessment we conclude that the ECB can improve the transparency of its monetary policy in two ways. First, it should drop altogether the first pillar and concentrate on a more explicit inflation-targeting strategy. The main problem with the role of money in the ECB strategy is that it is widely misunderstood by many observers and therefore creates the impression that the ECB lacks transparency. We think that the ECB should abandon the monetary analysis as a dominant navigation system for its monetary policy. The second improvement would be to announce a midpoint inflation target of 2 percent with a range of 1.5 to 2.5. A clear definition of

226

Chapter 8

price stability can enhance the transparency and credibility of the ECB. In our view, the arguments put forward for raising the inflation objective substantially are not compelling. From our reading of the literature, we see that the consequences of the zero bound are negligible for target inflation rates as low as 2 percent. Likewise the upward measurement bias in inflation seems to be less of a problem in the euro area than in the United States. Also the risk of deflation in an individual country in the euro area seems to be quite small. The evidence on the Balassa-Samuelson effect is so mixed that we do not consider this to be a convincing argument for a higher ECB inflation objective. Although we prefer an inflation-targeting strategy, we feel that a change toward inflation targeting will not automatically increase the transparency of the ECB. For example, the Fed and the Bundesbank are not inflation targeters but are perceived to be more transparent than the ECB; the Bank of England, on the other hand, has an inflation-targeting strategy but performs only slightly better than the ECB in terms of transparency. In other words, inflation targeting is neither a necessary nor a sufficient condition for transparency. The second main issue of the book is the decentralized structure of the central bank of the euro area. This setup is not necessarily problematic, but it has the potential to introduce an unwelcome bias into the ECB’s decision making if country representatives weigh in their own national economic developments and these developments deviate notably from the behavior of euro area aggregates. Already there is evidence that within the euro area, countries diverge in terms of their business cycles, inflation, and—to a lesser extent—monetary transmission. Quick entry into the monetary union remains the sensible strategy from the perspective of the new EU members, since it is beneficial for them. Membership euro area means lower interest rates, allowing the required budgetary consolidation to be easier to accomplish. The shorter the wait period, the lower are the risks of exchange rate instability. For countries with currency boards and with well-established unilaterally fixed exchange rates, the advocated waiting is especially difficult. The upcoming enlargement of the monetary union supports our view that reform of the ECB is necessary as it will increase economic diversity within the union. Business cycles of the new member countries will not yet be synchronized and inflation rates and monetary transmission will differ. Although an argument can be made that further economic and monetary integration will reduce this divergence

Conclusions

227

over time, there is evidence showing that integration does not always lead to less divergence. We therefore conclude that a reform of the ECB decision-making structure is needed to reduce the risk of suboptimal policy making. For the Governing Council, we favor a reform that brings the political power and economic weights of national central bank governors in line. From this perspective the proposed rotation system goes in the right direction, since every national central bank governor will have a voting right depending on the size of its home country economy. To avoid the potential problem of the size of Council and to forgo national central bank governors without voting rights to participate in the discussion, we recommend a smaller Governing Council—in which only national central bank governors with voting rights participate so that more efficient decision making can take place in the ECB.

Notes

Chapter 1 1. At the time of writing, apart from the new EU member countries, also Denmark, the United Kingdom, and Sweden do not (yet) participate in EMU. Still, it is expected that at some time these countries will also use the euro. 2. The central banks of the EU member states that do not participate in the euro area are members of the ESCB with a special status. They are allowed to conduct their respective monetary policies, but they do not take part in the decision making for the single monetary policy for the euro area and the implementation of such decisions. 3. Since not all EU member states participate in the monetary union, the ESCB also comprises central banks from countries that still have their own currency. The Eurosystem, a subset of the ESCB (excluding central banks of countries that do not participate in the monetary union), is responsible for carrying out the common monetary policy. See chapter 2 for further details. 4. For instance, the UK Prime Minister John Major, wrote in The Economist, “to recite the mantra of full economic and monetary union . . . will have all the quaintness of a rain dance and about the same potency.” 5. The ECB also has a number of other tasks. Article 105.2 of the Maastricht Treaty states, for instance, that the ESCB “promote the smooth operation of payment systems.” TARGET (the Trans-European Automated Real-time Gross settlement Express Transfer system) plays an important role in this respect. TARGET consists of 15 national real-time gross settlement (RTGS) systems and the ECB payment mechanism (EPM), which are all interlinked. More than 40,000 banks and branches can be addressed via TARGET. A daily average of more than 253,000 transactions (domestic and cross-border) were processed in 2002, with an average daily value of more than €1,500 billion. TARGET’s market share of large-value euro payment systems traffic is significant, with the system processing nearly 85 percent in terms of value and 59 percent in terms of volume. 6. See http://www.imf.org/external/np/mae/mft/. 7. The Economist, “Euro Towers or Fawlty Towers?” 31 October 1998. 8. Negotiations with Turkey and Croatia about EU membership will start in 2005. 9. See ECB Press Release at http://www.ecb.int.

230

Notes to Pages 9–39

Chapter 2 1. The first part of this chapter draws heavily on Eijffinger and De Haan (2000). 2. Moreover, and “without prejudice to the objective of price stability,” the ECB is obligated to “support the general economic policies in the Community with a view to contributing to the achievement of the objectives of the Community.” This includes a “high level of employment” and “sustainable and non-inflationary growth.” 3. In the HICP market transactions do not include interest rates, owner-occupied housing, nor business expenditures. Included are consumption expenditures of foreigners in the reference country but excluded are the consumption expenditures of residents abroad (see Camba-Mendez et al. 2002 for further details). 4. Energy prices constitute 8.2 percent of the HICP in 2003; see ECB (2004). 5. Standard means that the tender is conducted in accordance with a pre-announced schedule and is completed within a period of 24 hours from the announcement of the tender to the communication of the results; see ECB (2004) for further details. 6. Introductory statement by Duisenberg before the ECB Press Conference, December 5, 2002. 7. Similarly De Grauwe (2003) states: “the ECB is downgrading the importance of the money stock (M3) in its monetary policy strategy, and rightly so. It just did not make sense anymore to pretend that the money stock is the most important variable to watch. This variable is so much polluted by noise that it rarely gave the right warning signal of future inflation.” 8. In 2002 the Panel members were Christian de Boissieu, Guillermo de la Dehesa, Sylvester C. W. Eijffinger, Jean-Paul Fitoussi, Giampaolo Galli, Daniel Gros, Gustav A. Horn, Lars E. O. Svensson, Niels Thygesen, Norbert Walter, and Charles Wyplosz. The briefing papers and the full text of the hearing with the transcript of the questions and answers are downloadable from the European Parliament Web site: www.europarl.eu.int/comparl/econ/emu/default_en.htm.

Chapter 3 1. Blinder et al. (2001), for instance, argue that throughout 2001 the ECB has repeatedly refused to endorse market anticipations that it would cut its interest rates. By the time the ECB reduced rates by 25 basis points in May, the markets did no longer expect any change. 2. In chapter 4 we will deal more extensively with the issue of transparency. 3. Still improvements to the HICP could be made. For instance, owner-occupied housing costs could be included, see also Camba-Mendez (2003). 4. See Bernanke et al. (1999) for a good discussion of the experience of inflationtargeting countries. 5. It is often argued that the introduction of euro coins and banknotes raised the inflation rate substantially. However, available official data do not show a significant impact of the euro cash change-over on prices so far, although the estimated range has been

Notes to Pages 39–45

231

revised slightly up by Eurostat from 0 to 0.2 percent based on data for the first half of 2002; the previous estimate was 0 to 0.16 percent. Still, unusually high price increases did occur in some services (cafés and restaurants, health-related services, repairs, etc.) and some regularly purchased low-priced goods. This might explain the public perception that the introduction of euro banknotes and coins had a large inflationary impact (European Commission 2002a). 6. “For all these conceptual uncertainties and measurement problems, a specific numerical inflation target would represent an unhelpful and false precision. Rather price stability is best thought as an environment in which inflation is so low and stable over time that it does not materially enter into the decisions of households and firms.” Remarks by Chairman A. Greenspan on “Transparency in monetary policy” before the Federal Reserve Bank of St. Louis, Economic Policy Conference, October 11, 2001. 7. For example, according to calculations of the Deutsche Bundesbank in 1998, the consumer price index tends to overstate the effective inflation rate by around 0.75 a percentage point. See Deutsche Bundesbank, Monthly Bulletin, May 1998, p. 55. 8. As pointed out by Kumar et al. (2003), the costs of deflation depend on the deflationary source as well as on extent and duration. Deflation may not entail significant costs and be accompanied by growth in output under certain circumstances. Among these are temporary price declines due to overexpansion in aggregate supply caused by, for instance, a productivity spurt or some other positive external shock. In those cases deflation is the manifestation of an adjustment to a new equilibrium in the context of rising incomes. 9. Financial supervision and regulation in the euro area is even more decentralized. In some countries banking supervision is carried out by the central bank, whereas in others this task is performed by another institution, sometimes in close cooperation with the central bank (see Oosterloo and De Haan 2003). The current setup has led to a discussion about the proper role of the ECB in this regard (see Kremers et al. 2003 for an excellent discussion). The ECB is not entrusted with any direct responsibility related to prudential supervision of credit institutions and the stability of the financial system. Also fiscal policy-making in the euro area is decentralized. Still national fiscal policies are restricted by the Stability and Growth Pact (SGP). There is an intense debate about the question of whether rules like these are needed in a monetary union and, if so, whether the rules of the SGP are the proper ones (see Eijffinger and De Haan 2000, Buti et al. 2002, and De Haan et al. 2004 for a further discussion). 10. “Problems of price measurement,” remarks at the Annual Meeting of the American Economic Association and the American Finance Association, Chicago, Illinois, January 3, 1998. 11. This part heavily draws on Ullersma (2002); see also Yates (2002). 12. Price level targeting may be a better way to anchor expectations as undershooting of the target in the current period leads to inflationary expectations in the next period (Smets 2000). 13. In Akerlof et al. (1996) the model is based on nominal wage rigidity, while in Akerlof et al. (2000) the mechanism is near rationality in the use of inflationary expectations in price and wage setting. 14. Another view in which nominal wage rigidity also plays a role is that a temporary downturn can have persistent effects on unemployment due to hysteresis effects

232

Notes to Pages 45–83

(Blanchard and Summers 1986). This can be explained in a number of ways. The unemployed, because they lose human capital, may become less attractive to potential employers. Also, in an insider–outsider wage-setting framework, unions only take the interests of the currently employed into account, so wages are set at levels that are too high. Issing (2001b) argues for going to the root cause of the problem by structural reform. 15. Dickens (2001) points out some major weaknesses in the analysis of Wyplosz, for instance, that the long-run Phillips curve is not necessarily stable. 16. According to Jaeger (2003), some ECB communication suggests that the second pillar covers the higher business cycle frequencies of inflation (in the short run), and that the first pillar provides a “low-frequency crosscheck” for the second pillar. By low frequency Jaeger means cycles taking more than eight years to complete. However, certain ECB communications suggest that the first and second pillars both relate to the business cycle frequencies. 17. This episode provides a good illustration of the discrepancy between what the ECB does, and why, and how it is (wrongly) being reported. Gros et al. (2000, p. 2) write: “It is also interesting to note that the first pillar (monetary growth) did not seem to play a role in the ECB’s April 1999 decision. In fact, at the time of the cut, the money supply (M3) was expanding at about 5 percent, substantially above the target (sic) of 4.5 percent.” 18. Koen et al. (2001) point out that the evolution of effective exchange rates was far from uniform in the euro area. The depreciation in Belgium, for instance, was far less pronounced than in Ireland. This is because Belgium trades widely with other euro area countries, whereas Ireland trades mostly with the United Kingdom and the United States. So the depreciation was more of a problem for Ireland than for Belgium, because it was larger and because Ireland was ahead in the cycle (see also section 5.5). 19. As pointed out by Wyplosz (2003a), exchange rate changes can also indirectly affect inflation. A depreciation automatically increases the price of imported intermediate goods. However, as long as there is some pass-through, a depreciation can improve trade competitiveness, since with more exports and fewer imports, there is an improvement of the current account. As world demand for domestic production increases, the level of activity is stimulated and this exerts an upward influence on prices and wages. 20. Figure 3.5 shows the case where CES utility is assumed. See Feenstra et al. (1993) for further details. 21. This argument fits in with a branch of literature in which the microstructure of financial markets is used to describe the effects of sterilized foreign exchange market interventions. Examples include Bhattacharya and Weller (1997) and Vitale (1999). See Frenkel et al. (2001) for further details. 22. Other recent studies using another approach include Ito (2002), Kim (2003), King (2003), Saacke (2002), Vitale (2003), and Taylor (2004). 23. This view goes back to Alchian and Klein (1973).

Chapter 4 1. Many authors, however, use the term transparency when they are in fact referring to disclosure.

Notes to Pages 84–112

233

2. In contrast, Thornton (2002) argues that disclosure is neither necessary nor sufficient for accountability. In his view, central banks are accountable to the public via elected public officials. Consequently central bankers are accountable not only to these officials for achieving the objectives, which they or the elected officials set, but also for the way these objectives are being pursued. Secrecy to enhance the efficiency of central banks does not violate the principle of democratic accountability. Thornton argues that if disclosure diminishes policy effectiveness, accountability is not a justification for disclosure. 3. Of course, some of the information is directly transferred to the public, such as via the Internet. 4. The remainder of this section heavily draws on Eijffinger and Geraats (2002). 5. See Geraats (2002) for an excellent survey of the literature on central bank transparency. 6. Blinder et al. (2001) do not agree that publishing the model will enhance transparency. They argue (2001, p. 3) that “most central bank watchers will care only about the bank’s basic view of how the economy works and how it thinks monetary policy affects output and prices. Well-chosen words supplemented by a few key numbers may suffice. In fact, they may convey the relevant information better than masses of equations—which are often far from transparent.” 7. The Siklos index also includes the item “special recognition of the role of financial system stability” (on which the ECB gets a score of zero), which has little to do with disclosure. Finally, it seems that Siklos has made a mistake in coding for the element “publication of a monetary policy strategy.” While in his book he (rightly) states that the ECB has published its strategy, in the coding the ECB receives a zero on this element. 8. Most indicators are based on activities rather than the contents of what is being released. It is more important that the information be disclosed with clarity and that the disclosure of data bring genuine understanding of monetary policy making to the public. 9. This discrepancy is not only evident in the total scores of the indicators but also in the scores of various subgroups of issues—like objectives strategy and communication in our index. 10. Hämäläinen (2001) acknowledges this: “It is true that we have not always been very successful in our communication despite ambitious intentions. But communication is not easy in a pan-European context in which differing cultures, languages, traditions and motives affect how messages are interpreted by the different counterparties involved.” 11. There is much research on the influence of media on voting behavior. Biases in news coverage can have a measurable impact on electoral preferences of voters near the center of the political spectrum, since their political attitudes are often mixed, weakly held, or nonexistent. See Gunther and Mughan (2000). 12. This part draws on De Haan and Eijffinger (2000) and De Haan et al. (1999). See also Amtenbrink (1999). 13. For an earlier attempt to quantify accountability, see Briault et al. (1996). Another indicator for central bank accountability is developed by Siklos (2002). See also Bini Smaghi (1998). 14. The information reported by Fry et al. is based on a survey among of 94 central banks. The index of transparency used by Cechetti and Krause is based on responses to three

234

Notes to Pages 112–131

questions relating to the degree and frequency at which each central bank provides reports on its policy decisions, assessments about the state of the economy, and public explanation of forecasts. The index is obtained as a simple average of these three criteria. Independence is determined on the basis of the responses of five questions (i.e., on the importance of price stability, goal independence, instrument independence, government’s reliance on central bank financing, and term in office of the governor). Accountability is determined on the basis of the role of government in determining the objective(s) of monetary policy and on monitoring by parliament and the government. The index of credibility used by Cechetti and Krause (2002) is determined on the basis of the difference between an assumed inflation objective of 2 percent and expected inflation, proxied by actual inflation over the period 1985 to 1989. 15. Cukierman (2002) argues that the relationship between social welfare and the sacrifice ratio is not necessarily negative. A full welfare analysis should also take two other considerations into account. First, a higher sacrifice ratio also means that when monetary policy is expansionary, it has a stronger positive impact on output, and second, the permanent benefits from stable prices. 16. Haupt and Waller (2000) examined whether the incorporation of WES data in econometric models significantly improves their ability to analyze and forecast economic developments. Using survey data on inflation, these authors determine that forecasting models that incorporate WES survey data clearly outperform models that don’t. The obvious conclusion is that WES respondents had relevant information about future developments. 17. We thank Alan Blinder for providing us with his survey. 18. We agree with the definition of credibility given by Cukierman and Meltzer (1986, p. 1108): i.e., “the absolute value of the difference between the policy maker’s plan and the public’s beliefs about those plans.” 19. This has been calculated as follows: Using the IMF’s classification of advanced economies (WEO), we calculated average inflation rates for the period from 1971 to 2000 (long term) and from 1990 to 2000 (medium term). For the long term, we obtained an average inflation rate of 7 percent, for the medium term of 3 percent. High-inflation countries had above average inflation in both periods.

Chapter 5 1. As of March 1, 2004, the staff of DG Economics numbered 159 and the staff of DG Research 53, while the total staff of the ECB had increased to 1,231. 2. This “one size does not fit all” problem can undermine political support for the monetary union. Nitsch (2004) shows how inflation differentials, which are clearly related to differences in business cycle positions, can lead to dissolutions of currency unions. 3. Their main aim is to test whether this voting behavior was influenced by economic differences across the German states. Therefore they first examine to what extent the economic situation in the various states was similar. Berger and De Haan (2002) find that there was considerable volatility of inflation across states despite the fact that since late 1948 the D-mark circulated in all states under the sole rule of the Bundesbank. Interestingly, despite an early trend toward convergence that lasted until the mid-1950s, infla-

Notes to Pages 131–144

235

tion differentials continued to exist during the entire sample period. Behind these differences were, among other things, continued local and state autonomy in fiscal matters as well as considerable variation in the underlying economic structure. There were also large differences between growth rates of real GDP of the various states, which proved to be quite persistent. 4. Heinemann and Huefner (2002) have estimated reaction functions for the ECB. As there are no voting records for the ECB Governing Council, these authors had to rely on indirect ways to examine whether diverging economic developments in the euro area have affected ECB policy making. In their ordered probit model, Heinemann and Huefner find that the difference between the euro area average and the median of the inflation and output gap is significantly different from zero, though only at the 10 percent level. 5. Presidents of regional central banks were nominated by the governments of the concerned Lands, while members of the Executive Board of the Bundesbank were nominated by the federal government. 6. Also note the similarity in the political backgrounds of the appointing government and the appointee. 7. Berger and De Haan (2002) show that the economic differences between the German states in their sample period are of the same magnitude as the current economic differences between the countries in the euro area. 8. Furthermore Carlino and DeFina (1998) showed that output sensitivity to monetary policy shocks differs significantly across US regions, which is explained by the concentration of small firms and the share of manufacturing in total production. 9. In writing this section, we drew heavily on Massmann and Mitchell (2003). 10. We will focus on studies analyzing the synchronization of business cycles in the euro area. For references to other studies, see Massmann and Mitchell (2003); see also the homepage of Euro Area Business Cycle Network (http://www.eabcn.org). 11. However, Agresti and Mojon (2001) find that the business cycle fluctuations of GDP, consumption, and investment of most euro area countries were, even before stage three of EMU, highly synchronized with fluctuations in the euro area. 12. Wynne and Koo (2000) show that the cross-correlations among the business cycles (of GDP, prices, or employment) of the 12 US Federal Reserve districts are much higher than the cross-correlation of the business cycles of the 15 EU countries. Also Clark and van Wincoop (2001) report that business cycles of US Census regions are substantially more synchronized than those of EU countries, both over the past four decades and the past two decades. 13. A simple OLS regression of cyclical correlation on trade intensity may be inappropriate. Countries may, for instance, choose to link their currency to that of their most important trading partners; the resulting closeness of policies creates a positive association of business cycles in these countries. Frankel and Rose (1998) therefore use instrumental variables (IV). They use distance, adjacency, and common language dummies as instruments, based on the success of these variables in explaining trade (in “gravity equation” models) and the presumption that they are otherwise unrelated to the business cycle. The IV estimates of Frankel and Rose are as much as three times the size of the corresponding OLS estimates. However, Gruben et al. (2002) argue that neighboring countries with a similar language may be more disposed toward similar monetary poli-

236

Notes to Pages 144–171

cies. So the instruments that Frankel and Rose use may ultimately be seen as capturing the effects of various influences, and thereby upwardly biasing the estimated impact of trade alone. A statistical test for overidentifying restrictions confirms this. Gruben et al. abandon IV estimation in favor of OLS and incorporate the instruments used by Frankel and Rose into the system as independent variables. This yields point estimates that are some 50 percent lower than those reported by Frankel and Rose (1998). 14. Frankel and Rose (2002) combine estimates of the trade-creating effects of common currencies with evidence of a link between trade and growth. Their results suggest that some countries could increase their per capita income by 20 percent over 20 years by dollarizing or adopting the euro. 15. The countries are Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, Norway, Spain, Sweden, Switzerland, the United Kingdom, and the United States. 16. Interestingly, the standard deviation of inflation rates across countries after the introduction of the euro is not much different from the levels reported among German states after the introduction of the D-mark. In 1999–2001, the average standard deviation of inflation rates in the euro zone is 0.87. This closely matches the 1950–51 average standard deviation of inflation rates across German states of 0.88 as reported by Berger and De Haan (2002). 17. Kieler (2003) provides some illustrative calculations of how average inflation rates may differ across euro area countries over the next 10 to 20 years, depending on how fast the remaining differences in price levels are reduced. His results suggest that the inflationary trend may vary from 1.25 percent in Germany and France to 2 to 3.25 percent in Greece, Portugal, and Spain. 18. The results of von Hagen and Hoffman (2003) suggest, however, that aggregate demand in the euro area countries is determined by the euro real interest rate, while the national real interest rate differentials do not appear to have a significant effect on national output gaps. This conclusion is based on estimates of a simple backward-looking IS-curve in which the ex post real interest rate, the real effective exchange rate, and the US output gap are the regressors for the output gap in the euro area countries. 19. The complete results have been published in the Working Papers 91–114 of the ECB. See especially Mojon and Peersman (2001) and Angeloni et al. (2001). For a discussion of the results of simulations with econometric models of national central banks, see Van Els et al. (2001). 20. For a further discussion, see also Eijffinger and De Haan (2000) and Kieler and Saarenheimo (1998).

Chapter 6 1. For extensive surveys, we refer to Eijffinger and De Haan (1996) and Berger et al. (2001a). 2. Surveys of the early empirical research on CBI in transition economies can be found in Maliszewski (2000) and Wagner (1999). 3. The index of Grilli et al. (1991) is based on 16 criteria, relating to economic and political independence.

Notes to Pages 172–204

237

4. Wagner (1999) argues that as long as CBI is only “legal,” meaning that it only exists on paper, it will not only be ineffective but may even be counterproductive as further institutional reforms then become necessary. 5. The earlier study was published under the author’s previous name (Riesinger). 6. Some observers have argued in favor of euroization: the acceding countries adopt the euro as their national currency without being members of the euro area. However, unilateral euroization has been ruled out by the Ecofin as it “would run counter to the underlying economic reasoning of EMU in the [EU] Treaty, which foresees the eventual adoption of the euro as the endpoint of a structured convergence process within a multilateral framework” (Ecofin 2000). 7. Natalucci and Ravenna (2002) use a model calibrated for the Czech Republic and show that monetary policy and the choice of exchange rate regime alone do not necessarily allow compliance with the EMU accession requirements if the Balassa-Samuelson effect is at work. Even in the absence of any business cycle shock and with full credibility, monetary policy is unlikely to ensure that the levels of inflation and exchange rate are stabilized within the required bounds. 8. Also some studies exist that refer to just one country; see Iara and Traistaru (2003) for a discussion. 9. In summarizing the results of this study, we only consider coefficients that are significantly different from zero. 10. The Grubel-Lloyd index is defined as follows: 100 * (1 - (sum, all sectors |(export, per sector - import, per sector)|/sum, all sectors(export, per sector + import, per sector))). 11. See Rose (2004) for a summary of the (huge) literature on the effects of currency union on trade. 12. Fidrmuc and Korhonen (2004) report a high correlation between supply and demand shocks in the new member countries and the European’s share in their exports and imports. They also find significant correlations of both supply and demand shocks with the levels of intra-industry trade. In contrast, they find only low correlation of GDP per capita with both types of shocks. Thus the level of economic development of the countries does not seem to influence the synchronization of business cycles. 13. For an earlier paper following a similar line of argument, see Boone and Maurel (1998), who calculate correlation coefficients between the cyclical components of industrial production and unemployment rates for seven new member countries against Germany and the European. They report a relatively high degree of business cycle correlation for the new member countries with Germany. A related study is Boone and Maurel (1999). 14. The weighted average of the demand and supply shocks coefficients of correlation are 0.24 and 0.52 for the euro area countries and 0.13 and 0.11 for the new member countries. Malta and Cyprus are not included in the Fidrmuc and Korhonen (2003) sample. 15. We thank Fidrmuc for providing the detailed figures of Firdmuc and Korhonen (2004). Süpel (2003) does not provide the details of the calculations, and they are therefore not included in the table. 16. Ganev et al. (2002) and Elbourne et al. (2003) provide reviews of the literature on the topic.

238

Notes to Pages 204–221

17. See Golinelli and Rovelli (2002) for a good example of a recent small structural model for Hungary.

Chapter 7 1. This part heavily draws on Berger (2002) and Berger et al. (2003). 2. A proposal along these lines has been put forward by Horn (2003). Horn suggests that the countries be aligned in eight groups. The first group (with four votes) would be comprised of those four countries with the largest weight according to share of GDP and the share of population. In the second group would be the three countries next in the ranking of economic strength. This group has three votes. The remaining member states would be divided into six groups each consisting of three member states. Each group would have only one vote. Since each group’s composition would be established by economic strength and regional closeness, the vote would be cast by only one member of the group, and that position would change according to a rotation procedure. The total number of members in the Governing Council under this proposal would be nineteen. 3. See ECB Press Release, Governing Council prepares for enlargement, 20 December 2002, downloadable at: www.ecb.int. 4. The first draft of the paper by Berger et al. antedates the ECB proposal. 5. Bofinger (2003) shares this critique.

References

Agresti, A.-M., and B. Mojon. 2001. Some stylised facts on the euro area business cycle. ECB Working Paper 95. Akerlof, G. A., W. T. Dickens, and G. L. Perry. 1996. The macroeconomics of low inflation. Brookings Papers on Economic Activity 1: 1–59. Akerlof, G. A., W. T. Dickens, and G. L. Perry. 2000. Near-rational wage and price setting and the long-run Phillips curve. Brookings Papers on Economic Activity 1: 1–44. Alchian, A., and B. Klein. 1973. On a correct measure of inflation. Journal of Money, Credit, and Banking 5: 173–91. Alberola, E., S. Cervero, H. Lopez, and A. Ubide. 1999. Global equilibrium exchange rates: Euro, dollar, ‘ins’, ‘outs’ and other major currencies in a panel cointegration framework. IMF Working Paper 99/175. Alberola E., S. Cervero, H. Lopez, and A. Ubide. 2002. Quo vadis euro. European Journal of Finance 8: 352–70. Alberola, E., and T. Tyrväinen. 1998. Is there scope for inflation differentials in EMU? An empirical evaluation of the Balassa-Samuelson model in EMU countries. Bank of Finland Discussion Paper 15. Alquist R., and M. Chinn. 2001. Tracking the euro’s progress. International Finance 3: 357–73. Amtenbrink, F. 1999. The Democratic Accountability of Central Banks. London: Hart Publishers. Angeloni, I. 1999. The role of a regional bank in a system of central banks: A comment. Carnegie-Rochester Conference Series on Public Policy 51: 73–77. Angeloni, I., V. Gaspar, and O. Tristani. 1999. The monetary policy strategy of the ECB. In D. Cobham and G. Zis, eds., From EMS to EMU. London: Macmillan. Angeloni, I., A. Kashyap, B. Mojon, and D. Terlizzese. 2001. Monetary transmission in the euro area: Where do we stand? ECB Working Paper 114. Arnold, I. J. M., and C. J. M. Kool. 2002. The role of inflation differentials in regional adjustment: Evidence from the United States. Paper presented at the European Monetary Forum workshop, November 30, 2002.

240

References

Arnold, I. J. M., P. J. A. van Els, and J. De Haan. 2003. Wealth effects and monetary policy. De Nederlandsche Bank Research Memeorandum WO 719. Artis, M. 2003. Is there a European business cycle? CESifo Working Paper 1053. Artis, M. J., and M. Ehrmann. 2000. The exchange rate—A shock-absorber or source of shocks? A study of four open economies. EUI Working Papers ECO 2000/38. Artis, M., M. Marcellino, and T. Proietti. 2002a. Dating the euro area business cycle. CEPR Discussion Paper 3696. Artis, M., H.-M. Krolzig, and J. Toro. 1999. The European business cycle. EUI Working Papers ECO 1999/24. Artis, M. J., and W. Zhang. 1997. International business cycles and the ERM. International Journal of Finance and Economics 2(1): 1–16. Artis, M. J., and Zhang, W. 1999. Further evidence on the international business cycle and the ERM: Is there a European business cycle? Oxford Economic Papers 51: 120–32. Backé, P., and C. Wójcik. 2003. Introduction—Monetary integration of central and east European accession countries: The pros and cons of speedy versus more gradual strategies. In L. Vinhas de Souza and B. van Aarle, eds., The Euro Area and the New EU Member States. Houndsmills: Palgrave. Backé, P., J. Fidrmuc, T. Reininger, and F. Schardax. 2002. Price dynamics in central and eastern European EU accession countries. Oesterreichische Nationalbank, Working Paper 61. Balassa, B. 1964. The purchasing power parity doctrine: A reappraisal. Journal of Political Economy 72(6): 584–96. Baldwin, R., E. Berglof, F. Giavazzi, and M. Widgren. 2001. Preparing the ECB for enlargement. CEPR Policy Paper 6. Ball, L., N. G. Mankiw, and D. Romer. 1988. The new Keynesian economics and the output-inflation trade-off. Brookings Papers on Economic Activity 1: 1–65. Ball, L., and N. Sheridan. 2003. Does inflation targeting matter? NBER Working Paper 9577. Bank for International Settlements. 1994. National Differences in Interest Rate Transmission. CB 394. Basle: BIS. Barr, D., F. Breedon, and D. Miles. 2003. Life on the outside. Economic Policy, no. 37: 573–613. Baxter, M., and A. C. Stockman. 1989. Business cycles and the exchange-rate regime: Some international evidence. Journal of Monetary Economics 23(3): 377–400. Bayoumi, T., and B. Eichengreen. 1993. Shocking aspects of European monetary unification. In F. Torres and F. Giavazzi, eds., Growth and Adjustment in the European Monetary Union. Cambridge: Cambridge University Press. Beetsma, R. M. W. J., and H. Jensen. 1998. Inflation targets and contracts with uncertain central banker preferences. Journal of Money, Credit, and Banking 30(3): 384–403. Beetsma, R. M. W. J., and H. Jensen. 2003. Why money talks and wealth whispers: Monetary uncertainty and mystique. Comment. Journal of Money, Credit, and Banking 35(1): 129–36.

References

241

Begg, D. et al. 2000. One Money, Many Countries. Monitoring the European Central Bank 2. London: CEPR. Beissinger, T., and C. Knoppik. 2001. Downward nominal rigidity in West-German earnings 1975–95. German Economic Review 2: 385–417. Benigno, P., and J. D. Lopez-Salido. 2002. Inflation persistence and optimal monetary policy in the euro area. Board of Governors of the Federal Reserve System. International Finance Discussion Paper 749. Berger, H. 2002. The ECB and euro-area enlargement. IMF Working Paper 02/175. Berger, H., and J. De Haan. 2002. Are small countries too powerful within the ECB? Atlantic Economic Journal 30(3): 263–82. Berger, H., and U. Woitek. 1999. Does conservatism matter? A time series approach to central banking. CESifo Working Paper 190. Berger, H., J. De Haan, and S. C. W. Eijffinger. 2001a. Central bank independence: An update of theory and evidence. Journal of Economic Surveys 15(1): 3–40. Berger, H., J. De Haan, and E. van Fraassen. 2001b. How to reduce Inflation: An independent central bank or a currency board? The experience of the Baltic countries. Emerging Markets Review 2: 218–43. Berger, H., J. de Haan, and R. Inklaar. 2004. Restructuring the ECB. In H. Berger and T. Moutos, eds., Managing European Union Enlargement. Cambridge: MIT Press, pp. 29–66. Bergsten, C. F. 2002. The dollar and the US economy. Testimony before the Committee on Banking, Housing and Urban Affairs United States Senate, May 1, 2002 (http://www.iie.com/papers/bergsten0502.htm). Bernanke, B. S., and M. Gertler. 1995. Inside the black box: The credit channel of monetary policy transmission. Journal of Economic Perspectives 9(4): 27–48. Bernanke, B. S., and M. Gertler. 2000. Monetary policy and asset price volatility. In Federal Reserve Bank of Kansas City. New Challenges for Monetary Policy, pp. 77–128. Bernanke, B. S., T. Laubach, F. S. Mishkin, and A. S. Posen. 1999. Inflation Targeting: Lessons from the International Experience. Princeton: Princeton University Press. Bhattacharya, U., and P. Weller. 1997. The advantage to hiding one’s hand: Speculation and central bank intervention in the foreign exchange market. Journal of Monetary Economics 39(2): 251–77. Bini Smaghi, L. 1998. The democratic accountability of the European Central Bank. Banca Nationale del Lavoro Quarterly Review, no. 205: 119–43. Bini-Smaghi, L., and D. Gros. 2000. Open Issues in European Central Banking. London: Macmillan. Blanchard, O. J., and L. H. Summers. 1986. Hysteresis in unemployment. European Economic Review 31: 288–95. Blanchard, O. J., and D. Quah. 1989. The dynamic effects of aggregate demand and supply disturbances. American Economic Review 79(4): 655–73. Blanchflower, D. G. 2001. Unemployment, well-being and wage curves in Eastern and Central Europe. Journal of the Japanese and International Economies 15(4): 364–402.

242

References

Blinder, A. S. 1997. What central bankers could learn from academics—And vice versa. Journal of Economic Perspectives 11(2): 3–19. Blinder, A. S. 2000. Central bank credibility: Why do we care? How do we build it? American Economic Review 90(5): 1421–31. Blinder, A. S., C. Goodhart, P. Hildebrand, D. Lipton, and C. Wyplosz. 2001. How Do Central Banks Talk? Geneva Reports on the World Economy 3. Geneva/London: ICMB/CEPR. Bofinger, P. 1999. The conduct of monetary policy by the European central bank according to Article 105 of the Treaty versus the real economy. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. April 1999. Bofinger, P. 2003. Consequences of the modification of the governing council rules. Briefing paper for the Monetary Committee of the European Parliament. February 2003. Boone, L., and M. Maurel. 1998. Economic convergence of the CEECs with the EU. CEPR Discussion Paper 2018. Boone, L., and M. Maurel. 1999. An optimal currency area perspective of the EU enlargement to the CEECs. CEPR Discussion Paper 2119. Bordo, M., and O. Jeanne. 2002. Boom-bust in asset prices, economic instability, and monetary policy. NBER Working Paper 8966. Borghijs, A., and L. Kuijs. 2004. Exchange rates in central Europe: A blessing or a curse? IMF Working Paper 04/02. Borio, C. E. V., and W. Fritz. 1995. The response of short-term bank lending rates to policy rates: A cross country perspective. In Financial Structure and the Monetary Policy Transmission Mechanism. Basle: BIS. Borio, C. E. V., and P. Lowe. 2002. Asset prices, financial and monetary stability: Exploring the nexus. BIS Working Papers 114. Borowski, D., and C. Courharde. 2000. Euro, dollar, yen: Pour une approache multilatérale des taux de change d’équilibre. Revue Economique 51(3): 671–81. Brand, D., W. Gerstenberger, and J. D. Lindbauer. 1997. Selected general indicators of the Ifo Institute for Economic Research. In K. H. Oppenländer, ed., Business Cycle Indicators. Aldershot: Avebury. Brainard, W. 1967. Uncertainty and the Effectiveness of Policy. American Economic Review 57: 411–25. Briault, C. B., A. G. Haldane, and M. A. King. 1996. Independence and accountability. Bank of England Working Paper 49. Bryan, M. F., S. G. Cecchetti, and R. O’Sullivan. 2001. Asset prices in the measurement of inflation. De Nederlandsche Bank Staff Report 62. Buiter, W. H. 1999a. Alice in euroland. Journal of Common Market Studies 37(2): 181–209. Buiter, W. H. 1999b. Optimal currency areas: Why does the exchange rate regime matter? Sixth Royal Bank of Scotland/Scottish Economic Society Annual Lecture. Edinburgh, October 26, 1999. Buiter, W. H. 2003. Deflation: Prevention and cure. NBER Working Paper 9623.

References

243

Buiter, W. H., and C. Grafe. 2002. Anchor, float or abandon ship: Exchange rate regimes for the accession countries. Banca Nazionale del Lavoro Quarterly Review, no. 221: 111–42. Buiter, W. H., and N. Panigirtzoglou. 1999. Liquidity traps: How to avoid them and how to escape them? NBER Working Paper 7245. Buiter, W. H., and A. Taci. 2002. Capital account liberalisation and financial sector development in transition countries. Mimeo. EBRD. Bun, M. J. G., and F. J. G. M. Klaassen. 2002. Has the euro increased trade? Tinbergen Institute Discussion Paper 2002-108/2. Buti, M., S. C. W. Eijffinger, and D. Franco. 2003. Revisiting the stability and growth pact: Grand design or internal adjustment? CEPR Discussion Paper 3692. Calza, A., and J. Sousa. 2003. Why has broad money demand been more stable in the euro area than in other economies? A literature review. In Background Studies for the ECB’s Evaluation of Its Monetary Policy Strategy. Frankfurt: ECB. Camba-Mendez, G. 2003. The definition of price stability: Choosing a price measure. In Background Studies for the ECB’s Evaluation of Its Monetary Policy Strategy. Frankfurt: ECB. Camba-Mendez, G., V. Gaspar, and M. Wynne. 2002. Measurement Issues in European Consumer Price Indices and the Conceptual Framework of the HICP. ECB/CEPR (downloadable at the ECB homepage, www.ecb.int). Campa, J. M., and L. S. Goldberg. 2002. Exchange rate pass-through into import prices: A macro or a micro phenomenon? NBER Working Paper 8934. Campa, J. M., and J. M. González Mínguez. 2002. Differences in exchange rate passthrough in the euro area. Mimeo. IESE Business School. Canzoneri, M., R. Cumby, B. Diba, and G. Eudey. 2001. Productivity trends in Europe: Implications for real exchange rates, real interest rates, and inflation. Review of International Economics 10(3): 497–516. Castelnuovo, E., S. Nicoletti-Altimari, and D. Rodríguez Palenzuela. 2003. Definition of price stability, range and point inflation targets: The anchoring of long-term inflation expectations. In Background Studies for the ECB’s Evaluation of its Monetary Policy Strategy. Frankfurt: ECB. Carlino, G., and R. DeFina. 1998. The differential regional effects of monetary policy. Review of Economics and Statistics 80(4): 572–87. Cecchetti, S. G. 1999. Legal structure, financial structure, and the monetary policy transmission mechanism. Federal Reserve Bank of New York Economic Policy Review 5(2): 9–28. Cechetti, S. G., and S. Krause. 2002. Central bank structure, policy efficiency, and macroeconomic performance: Exploring empirical relationships. Federal Reserve Bank of St. Louis Review 84(4): 47–59. Cechetti, S. G., H. Genberg, J. Lipsky, and S. Wadhwani. 2000. Asset Prices and Central Bank Policy. CEPR/ICMB Geneva Reports on the World Economy 2. Cecchetti, S., N. Mark, and R. Sonora. 2002. Price level convergence among United States cities: Lessons for the European central bank. International Economic Review 43: 1081–99. Chaney, E. 2003. Euroland: A weaker euro is critical for the recovery. Global Economic Forum. Morgan Stanley.

244

References

Chortareas, G., D. Stasavage, and G. Sterne. 2002. Does it pay to be transparent? International evidence from central bank forecasts. Federal Reserve Bank of St. Louis Review 84(4): 99–118. Chortareas, G., D. Stasavage, and G. Sterne. 2003. Does monetary policy transparency reduce disinflation costs? The Manchester School 71(5): 521–40. Christodoulakis, N., S. Dimelis, and T. Kollintzas. 1995. Comparisons of business cycles in the EC: Idiosyncrasies and regularities. Economica 62: 1–27. Christoffersen, P., T. Sløk, and R. Wescott. 2001. Is inflation targeting feasible in Poland? Economics of Transition 9(1): 153–74. Clark, T. E., and E. van Wincoop. 2001. Borders and business cycles. Journal of International Economics 55: 59–85. Clostermann, J., and B. Schnatz. 2000. The determinants of the euro-dollar exchange rate, synthetic fundamentals and a non-existing currency. Konjunkturpolitik 46(3): 274–302. Coenen, G. 2003. Downward nominal wage rigidity and the long-run Phillips curve: Simulation-based evidence for the euro area. In Background Studies for the ECB’s Evaluation of Its Monetary Policy Strategy Frankfurt: ECB. Coenen, G., A. Orphanides, and V. Wieland. 2003. Price stability and monetary policy effectiveness when nominal interest rates are bounded at zero. Mimeo. March 2003. Coevering, C. van de. 2003. Structural convergence and monetary integration in Europe. De Nederlandsche Bank. MEB series 2003-20. Coricelli, F., and B. Jazbec. 2001. Real exchange rate dynamics in transition economics. CEPR Discussion Paper 2869. Corsetti G. 2000. A perspective on the euro. CESifo Forum 2 (summer). Ifo Institute for Economic Research, pp. 32–36. Corsetti, G., and L. Dedola. 2001. Macroeconomics of international price discrimination. Mimeo. University of Rome. Corsetti, G., and P. Pesenti. 1999. Stability asymmetry and discontinuity: The launch of European Monetary Union. Brookings Papers on Economic Activity, no. 2: 295–358. Corsetti, G., and P. Pesenti. 2002. International dimensions of optimal monetary policy. Mimeo. Yale University. Cosimano, T. F., and J. B. Van Huyck. 1993. Central bank secrecy, interest rates, and monetary control. Economic Inquiry 31(3): 370–82. Cottarelli, C., and A. Kourelis. 1994. Financial structure, bank lending rates, and the transmission mechanism of monetary policy. IMF Staff Papers 41(4): 587–623. Cukierman, A. 1992. Central Bank Strategy, Credibility and Independence: Theory and Evidence. Cambridge: MIT Press. Cukierman, A. 2001a. Accountability, credibility, transparency, and stabilization in the Eurosystem. In C. Wyplosz, ed., The EMU and Its Impact on Europe and the World. Oxford: Oxford University Press. Cukierman, A. 2001b. Are contemporary central banks transparent about economic models and objectives and what difference does it make? Bundesbank Discussion Paper 05/01.

References

245

Cukierman, A. 2002. Does a higher sacrifice ratio mean that central bank independence is excessive? Annals of Economics and Finance 3: 1–25. Cukierman, A., and A. H. Meltzer. 1986. A theory of ambiguity, credibility, and inflation under discretion and asymmetric information. Econometrica 54(5): 1099–128. Cukierman, A., G. P. Miller, and B. Neyapti. 2002. Central bank reform, liberalization and inflation in transition economies—An international perspective. Journal of Monetary Economics 49: 237–64. Cukierman, A., S. B. Webb, and B. Neyapti. 1992. Measuring the independence of central banks and its effect on policy outcomes. World Bank Economic Review 6(3): 353–98. Darvas, Z., and G. Szapáry. 2004. Business cycle synchronization in the enlarged EU: Comovements in the new and old members. Central Bank of Hungary. Working Paper 2004/1. De Broeck, M., and T. Sløk. 2001. Interpreting real exchange rate movements in transition countries. IMF Working Paper 01/56. Decressin, A., and J. Decressin. 2002. On sand and the role of grease in labour markets: How does Germany compare? IMF Working Paper 02/164. De Grauwe, P. 2000. Exchange rates in search of fundamentals: The case of the euro-dollar rate. International Finance 3(3): 329–56. De Grauwe, P. 2003. The central bank that has missed the point. Financial Times, May 13: 2003. De Grauwe, P., and F. Skudelny. 2000. Inflation and productivity differentials in EMU. Centrum voor Economische Studiën, Leuven, Belgium. Discussion Paper 15. De Haan, J., and F. Amtenbrink. 2002. A non-transparent European central bank? Who is to blame? Mimeo. University of Groningen. De Haan, J., and S. C. W. Eijffinger. 2000. The democratic accountability of the European central bank: A comment on two fairy-tales. Journal of Common Market Studies 38(3): 393–407. De Haan, J., F. Amtenbrink, and S. C. W. Eijffinger. 1999. Accountability of central banks: Aspects and quantification. Banca Nazionale del Lavoro Quarterly Review, no. 209: 169–93. De Haan, J., R. Inklaar, and O. Sleijpen. 2002. Have business cycles become more synchronized? Journal of Common Market Studies 40(1): 23–42. De Haan, J., H. Berger, and D. Jansen. 2004. Why did the stability and growth pact fail? International Finance 7: 235–60. Demertzis, M., and A. Hughes Hallett. 2002. Central bank transparency in theory and practice. De Nederlandsche Bank. WO Research Memoranda 704. De Nardis, S., and C. Vicarelli. 2003. Currency unions and trade: The special case of EMU. Review of World Economics 139(4): 625–49. De Nicolo, G., P. Honohan, and A. Ize. 2003. Dollarization of the banking system: Good or bad? World Bank Policy Research Working Paper 3116. De Souza, L. V. 2002. Trade effects of monetary integration in large, mature economies: A primer on the European Monetary Union. Kiel Working Paper 1137.

246

References

Detken, C., A. Dieppe, J. Hendry, C. Marin, and F. Smets. 2002. Model uncertainty and the equilibrium value of the real effective euro exchange rate. ECB Working Paper 160. Devereux, M. B., C. Engel, and C. Tille. 1999. Exchange rate pass-through and the welfare effects of the euro. NBER Working Paper 7382. Dickens, W. T. 2001. Comment. In A. Garcia Herrero, V. Gaspar, L. H. Hoogduin, J. Morgan, and B. Winkler, eds., Why Price Stability? Frankfurt: ECB. Dominguez, K. M. 1990. Market responses to coordinated central bank intervention. Carnegie-Rochester Conference Series on Public Policy 32: 121–63. Dominguez, K. M. 2003. The market microstructure of central bank intervention. Journal of International Economics 59: 25–45. Dominguez, K. M., and J. A. Frankel. 1993a. Does Foreign Exchange Intervention Work? Washington, DC: Institute for International Economics. Dominguez, K. M., and J. A. Frankel. 1993b. Does foreign exchange intervention matter? The portfolio effect. American Economic Review 83(5): 1356–69. Döpke, J. 1999. Stylised facts of Euroland’s business cycle. Jahrbücher für National Ökonomie und Statistik 219: 591–610. Dornbusch, R. 1987. Exchange rates and prices. American Economic Review 77(1): 93–106. Dornbusch, R., C. Favero, and F. Giavazzi. 1998. Immediate challenges for the ECB: Issues in formulating a single monetary policy. Economic Policy 26: 15–64. Dotsey, M. 1987. Monetary policy, secrecy, and federal funds rate behavior. Journal of Monetary Economics 20(3): 463–74. Duarte, M. 2003. The euro and inflation divergence in Europe. Federal Reserve Bank of Richmond Economic Quarterly 89(3): 53–70. Durjasz, P. 2001. Experiences with the monetary transmission mechanism in Poland. Oesterreichische Nationalbank. Focus on Transition 2: 142–46. Duval, R. 2001. Estimation du taux de change reel d’equilibre de long term euro-dollar par une approach dynamique synthetique. Mimeo. Universite Paris-I Pantheon. Dvorsky, S. 2000. Measuring central bank independence in selected transition countries and the disinflation process. Bank of Finland. BOFIT Discussion Paper 13/2000. Ecofin. 2000. Exchange-rate strategies for accession countries. November 7, 2000. EFN (European Forecasting Network). 2003. Autumn 2003 Report (www.efn.unibocconi.it). Égert, B. 2002a. Estimating the impact of the Balassa-Samuelson effect on inflation and the real exchange rate during the transition. Economic Systems 26(1): 1–16. Égert, B. 2002b. Investigating the Balassa-Samuelson hypothesis in the transition: Do we understand what we see? A Panel Study. Economics of Transition 10(2): 273–309. Égert, B. 2003. Assessing equilibrium exchange rates in CEE acceding countries: Can we have DEER with BEER without FEER? A critical survey of the literature. Oesterreichische Nationalbank. Focus on Transition 2: 38–106.

References

247

Ehrmann, M. 1998. Will EMU generate asymmetry? Comparing monetary policy transmission across European countries. European University Institute Working Paper 98/28. Eichengreen, B. 1998. European Monetary Unification: A tour d’horizon, Oxford Review of Economic Policy 14(3): 24–40. Eichengreen, B., and F. Ghironi. 2001. EMU and enlargement. Mimeo. University of Berkley. Eichengreen, B., and C. Wyplosz. 1993. The unstable EMS. Brookings Papers on Economic Activity 1: 51–124. Eijffinger, S. C. W. 2002. How successful was the changeover to euro currency and what are the consequences for price transparency within the eurozone? Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. January. Eijffinger, S. C. W., and J. De Haan. 1996. The political economy of central-bank independence. Princeton Special Papers in International Economics, no. 19. Eijffinger, S. C. W., and J. De Haan. 2000. European Monetary and Fiscal Policy. Oxford: Oxford University Press. Eijffinger, S. C. W., J. De Haan, and K. Koedijk. 2002. Small is beautiful. European Journal of Political Economy 18(2): 365–74. Eijffinger, S. C. W., and P. Geraats. 2002. How transparent are central banks? CEPR Discussion Paper 3188. Eijffinger, S. C. W., and H. Huizinga. 1999. Should monetary policy be adjusted frequently? CEPR Discussion Paper 2074. Eijffinger, S. C. W., M. Hoeberichts, and E. Schaling. 2000. Why money talks and wealth whispers: Monetary uncertainty and mystique. Journal of Money, Credit, and Banking 21(2): 218–35. Eijffinger, S. C. W., E. Schaling, and W. Verhagen. 1999. A theory of interest rate stepping: Inflation-targeting in a dynamic menu cost models. CEPR Discussion Paper 2168. Eijffinger, S. C. W., and P. Stadhouders. 2003. Monetary policy and the rule of law. CEPR Discussion Paper 3698. Elbourne, A., J. De Haan, and B. M. C. Kiviet. 2003. Financial structure and monetary policy transmission in transition countries. Mimeo. University of Groningen. Emerson, M., D. Gros, A. Italianer, J. Pisani-Ferry, and H. Reichenbach. 1992. One Market, One Money: An Evaluation of the Potential Benefits and Costs of Forming an Economic and Monetary Union. Oxford: Oxford University Press. European Central Bank. 2001. The Monetary Policy of the ECB. Frankfurt am Main: ECB. European Central Bank. 2002. Economic Fundamentals and the Exchange Rate of the Euro. Monthly Bulletin (January): 41–53. European Central Bank. 2003a. Background Studies for the ECB’s Evaluation of its Monetary Policy Strategy. Frankfurt am Main: ECB. European Central Bank. 2003b. Inflation Differentials in the Euro Area: Potential Causes and Policy Implications. Frankfurt am Main: ECB.

248

References

European Central Bank. 2004. The Monetary Policy of the ECB. Frankfurt am Main: ECB. European Commission. 2002a. The EU economy 2002 review. European Economy, no. 6. European Commission. 2002b. Economic Reform: Report on the functioning of community product and capital markets. COM 743, Brussels. European Economic Advisory Group at CESifo (EEAG). 2002. Report on the European Economy 2002. Munich: CESifo. European Economic Advisory Group at CESifo (EEAG). 2004. Report on the European Economy 2004. Munich: CESifo. European Monetary Institute. 1997. The Single Monetary Policy in Stage Three: General Documentation on ESCB Monetary Policy Instruments and Procedures. Frankfurt am Main: EMI (downloadable at: www.ecb.int). Eurostat. 2001. Candidate countries: Labour productivity levels and remuneration levels just over 40% of EU average. News release 55. Fagan, G., J. Henry, and R. Mestre. 2001. An area-wide model (AWM) for the euro area. ECB Working Paper 42. Fatás, A. 1997. EMU: Countries or regions? Lessons from the EMS experience. European Economic Review 41(3–5): 743–51. Fatum, R., and M. M. Hutchison. 2002. ECB foreign exchange intervention and the euro: Institutional framework, news, and intervention. Open Economies Review 13: 413–25. Fatum, R., and M. M. Hutchison. 2003. Is sterilized foreign exchange intervention effective after all? An event study approach. Economic Journal 113: 390–411. Faust, J., and L. E. O. Svensson. 2000. The equilibrium degree of transparency and control in monetary policy. CEPR Discussion Paper 2195. Faust, J., and L. E. O. Svensson. 2001. Transparency and credibility: Monetary policy with unobservable goals. International Economic Review 42(2): 369–97. Feenstra, R. C., J. E. Gagnon, and M. M. Knetter. 1993. Market share and exchange rate pass-through in world automobile trade. Journal of International Economics 40: 187–207. Feige, E. L. 2003. Dynamics of currency substitution, asset substitution and de facto dollarisation and euroization in transition countries. Comparative Economic Studies 45: 358–74. Feldstein, M., ed. 1999. The Costs and Benefits of Price Stability. Chicago: University of Chicago Press. Fidrmuc, Jan. 2001. Migration and adjustment to shocks in transition economies. ZEI Working Paper B01-23. Fidrmuc, J. 2001. The endogeneity of optimum currency area criteria, intraindustry trade and EMU enlargement. Bank of Finland, BOFIT Discussion Papers 8, forthcoming in: Economic Systems. Fidrmuc, J., and I. Korhonen. 2003. Similarity of supply and demand shocks between the euro area and the CEECs. Economic Systems 27(3): 313–36. Fidrmuc, J., and I. Korhonen. 2004. The euro goes east: Implications of the 2000–2002 economic slowdown for synchronisation of business cycles between the euro area and CEECs. Comparative Economic Studies 46: 45–62.

References

249

Flam, H., and H. Nordström. 2003. Trade volume effects of the euro: Aggregate and sector estimates. Mimeo. IIES. Forni M., and Reichlin L. 2001. Federal policies and local economies: Europe and the US. European Economic Review 45(1): 109–34. Frankel, J. 1999. No single currency regime is right for all countries or at all times. Essays in International Finance 215. Princeton: Princeton University Press. Frankel, J. A., and Rose, A. K. 1998. The endogeneity of the optimum currency area criteria. Economic Journal 108: 1009–25. Frankel, J. A., and A. Rose. 2002. An estimate of the effect of common currencies on trade and income. Quarterly Journal of Economics 117(2): 437–66. Freedman, C. 2002. Panel discussion: Transparency in the practice of monetary policy. Federal Reserve Bank of St. Louis Review 84(4): 155–66. Frenkel, M., and C. Nickel. 2002. How symmetric are the shocks and the shock adjustment dynamics between the euro area and central and eastern European countries. IMF Working Paper 02/222. Frenkel, M., C. Pierdzioch, and G. Stadtmann. 2001. The foreign exchange market interventions of the European Central Bank. Banca Nazionale del Lavoro Quarterly Review, no. 218: 249–87. Friedman, M., and A. J. Schwartz. 1963. A Monetary History of the United States, 1867–1960. Princeton: Princeton University Press. Fry, M., D. Julius, L. Mahadeva, S. Roger, and G. Sterne. 2000. Key issues in the choice of monetary policy framework. In L. Mahadeva and G. Sterne, eds., Monetary Policy Frameworks in a Global Context. London: Routledge. Galati, G., and C. Ho. 2001. Macroeconomic news and the euro/dollar exchange rate. BIS Working Paper 105. Ganev, G., K. Molnar, K. Rybin´ski, and P. Woz´niak. 2002. Transmission mechanism of monetary policy in central and eastern Europe. Center for Social and Economic Research, Warsaw. Case Reports 52. Gaspar, V., G. Pérez Quirós, and J. Sicilia. 2001. The ECB monetary policy strategy and the money market. ECB Working Paper 69. Geraats, P. M. 2000. Why adopt transparency? The publication of central bank forecasts. CEPR Discussion Paper 2582. Geraats, P. M. 2001. Transparency of monetary policy: Does the institutional framework matter? Mimeo. University of Cambridge. Geraats, P. M. 2002. Central bank transparency. Economic Journal 112: F532–65. Gern, K.-J., C. Kamps, C. P. Meier, and J. Scheide. 2000. Euroland: Peak of the upswing— Little evidence of a new economy. Kiel Discussion papers 369. Gersbach, H. 1998. On the negative social value of central banks’ transparency. Mimeo. University of Heidelberg. Gersbach, H., and V. Hahn. 2001a. Should the individual voting records of central bankers be published? Bundesbank Discussion Paper 02/01.

250

References

Gersbach, H., and V. Hahn. 2001b. Voting transparency and conflicting interests in central bank councils. Bundesbank Discussion Paper 03/01. Gilchrist, S., and J. V. Leahy. 2002. Monetary policy and asset prices. Journal of Monetary Economics 49: 75–97. Glick, R., and A. K. Rose. 2002. Does a currency union affect trade? The time series evidence. European Economic Review 46: 1125–51. Goldberg, P. K., and M. M. Knetter. 1997. Goods prices and exchange rates: What have we learned? Journal of Economic Literature 35: 1243–392. Golinelli, R., and R. Rovelli. 2002. Painless disinflation? Monetary policy rules in Hungary, 1991–99. Economics of Transition 10(1): 55–91. Goodfriend, M. 1986. Monetary mystique: Secrecy and central banking. Journal of Monetary Economics 17: 63–92. Goodfriend, M. 1991. Interest rates and the conduct of monetary policy. Carnegie-Rochester Conference Series on Public Policy 34: 7–30. Goodfriend, M. 1999. The role of a regional bank in a system of central banks. CarnegieRochester Conference Series on Public Policy 51: 51–71. Goodfriend, M. 2000. Overcoming the zero bound on interest rate policy. Journal of Money, Credit, and Banking 32(4, pt. 2): 1007–35. Goodfriend, M., R. König, and R. Repullo. 2004. External Evaluation of the Economic Research Activities of the European Central Bank (downloadable at: www.ecb.int). Goodhart, C. A. E. 1996. Why do the monetary authorities smooth interest rates? LSE Financial Markets Groups and ESRC Research Centre. Special Paper series 81. Goodhart, C. A. E. 2001. What weight should be given to asset prices in the measurement of inflation? De Nederlandsche Bank Staff Report 65. Goodhart, C. A. E., and B. Hoffman. 2000. Do asset prices help to predict consumer price inflation? The Manchester School 68(suppl.): 122–40. Gordon, R. J. 1996. The macroeconomics of low inflation: Comment. Brookings Papers on Economic Activity 1: 60–6. Gormley, L., and J. de Haan. 1996. The democratic deficit of the European Central Bank. European Law Review 21: 95–112. Gottschalk, J., and D. Moore. 2001. Implementing inflation targeting regimes: The case of Poland. Journal of Comparative Economics 29: 24–39. Grilli, V., D. Masciandaro, and G. Tabellini. 1991. Political and monetary institutions and public financial policies in the industrial countries. Economic Policy, no. 13: 341–92. Gros, D. 2003. Reforming the composition of the ECB Governing Council in view of enlargement: How not to do it! Briefing paper for the Monetary Committee of the European Parliament. February. Gros, D., et al. 2000. Quo Vadis Euro? The Cost of Muddling Through. Brussels: CEPS. Gros, D., and L. Bini-Smaghi. 2001. Is the European Central Bank Sufficiently Accountable and Transparent? Brussels: CEPS.

References

251

Groshen, E. L., and M. E. Schweitzer. 1999. Identifying inflation’s sand and grease effect in the labour market. In M. Feldstein, ed., The Costs and Benefits of Price Stability. Chicago: University of Chicago Press. Gruben, W. C., J. Koo, and E. Millis. 2002. How much does international trade affect business cycle synchronization? Federal Reserve Bank of Dallas Working Paper 0203. Guiso, L., A. K. Kashyap, F. Panetta, and D. Terlizzese. 1999. Will a common European monetary policy have asymmetric effects? Federal Reserve Bank of Chicago. Economic Perspectives 23(4): 56–75. Gunther, R., and A. Mughan. 2000. Democracy and the Media: A Comparative Perspective. Cambridge: Cambridge University Press. Halpern, L., and C. Wyplosz. 2001. Economic transformation and real exchange rates in the 2000s: The Balassa-Samuelson connection. Mimeo. Hämäläinen, S. 2001. The ECB’s monetary policy—Accountability, transparency and communication. Introduction at The Old Age, New Economy and Central Banking Conference. Organized by CEPR/ESI and Suomen Pankki, September 14, 2001, Helsinki. Hamecz, I. 2001. Monetary transmission in Hungary. Oesterreichische Nationalbank. Focus on Transition 2001/2: 139–42. Hansen, J., and W. Roeger. 2000. Estimation of real equilibrium exchange rates. European Commission Economic Paper 144. Harding, D., and A. Pagan. 2001. Extracting, analysing and using cyclical information. Mimeo. University of Melbourne. Haupt, H., and S. Waller. 2000. Economic analysis and short-term forecasting with qualitative data from the Economic Survey International (ESI). In K. H. Oppenländer, G. Poser, and B. Schips, eds., Use of Survey Data for Industry, Research, and Economic Policy. Aldershot: Avebury. Heinemann, F., and F. P. Huefner. 2002. Is the view from the eurotower purely European? National divergence and ECB interest rate policy. Mimeo. Centre for European Economic Research, Mannheim. Hendrikx, M., and B. Chapple. 2002. Regional inflation divergence in the context of EMU. De Nederlandsche Bank. MEB Working Paper 19. Hochreiter, E., and T. Kowalski. 2000. Central banks in European emerging market economies in the 1990s. Banca Nazionale del Lavoro Quarterly Review, no. 212: 45–70. Hofmann, B. 2002. The pass-through of money market rates to business loan rates in the euro area countries. Mimeo. ZEI, University of Bonn. Honohan, P., and P. R. Lane. 2003. Divergent inflation rates in EMU. Economic Policy, no. 37: 357–94. Horn, G. A. 2003. Consequences of the modification of the Governing Council rules. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. February. Horvath, J. 2000. Supply and demand shocks in Europe: Large-4 EU members, Visegrad5 and Baltic-3 Countries. Mimeo. Central European University.

252

References

Horvath, J. 2003. Optimum currency area theory: A selective review. Bank of Finland. BOFIT Discussion Papers 15. Hughes Hallett, A. J., and N. Viegi. 2001. Credibility, transparency and asymmetric Information in monetary policy. CEPR Discussion Paper 2671. Hüfner, F. P., and M. Schröder. 2002. Exchange rate pass-through to consumer prices: A European perspective. ZEW Discussion Paper 02-20. Iara, A., and I. Traistaru. 2003. How flexible are wages in EU accession countries? ZEI Working Paper B25-2003. Imbs, J. 2004a. Trade, finance, specialization and synchronization. Review of Economics and Statistics 86: 723–34. Imbs, J. 2004b. The real effects of financial integration. Mimeo. London Business School. Inklaar, R., and J. De Haan. 2001. Is there really a European business cycle? A comment. Oxford Economic Papers 53: 215–20. International Monetary Fund. 1999. Greece—Selected issues. IMF Country Report 99/138. Washington, DC: IMF. International Monetary Fund. 2000a. Supporting Document to the Code of Good Practices on Transparency in Monetary and Financial Policies. Washington, DC: IMF. International Monetary Fund. 2000b. Monetary and Exchange Rate Policies in the Euro Area. IMF Country Report 00/46. Washington, DC: IMF. International Monetary Fund. 2002. Monetary and exchange rate policies of the euro area—Selected issues. IMF Country Report 02/236. Washington, DC: IMF, ch. 1. Issing, O. 1999a. The monetary policy of the eurosystem. Finance and Development 36(March): 18–21. Issing, O. 1999b. The eurosystem: Transparent and accountable, or “Willem in Euroland.” Journal of Common Market Studies 37(3): 503–19. Issing, O. 2000. Communication challenges for the ECB. CFS research conference on the ECB and its watchers. Frankfurt, June 26, 2000. Issing, O. 2001a. The euro area and the single monetary policy. Oesterreichische Nationalbank. Working Paper 44. Issing, O. 2001b. Why price stability? In A. Garcia Herrero, V. Gaspar, L. H. Hoogduin, J. Morgan, and B. Winkler, eds., Why Price Stability? Frankfurt: ECB. Issing, O. 2003. Foreword. In Background Studies for the ECB’s Evaluation of its Monetary Policy Strategy. Frankfurt: ECB. Issing, O., V. Gaspar, I. Angeloni, and O. Tristani. 2001. Monetary Policy in the Euro Area: Strategy and Decision-Making at the European Central Bank. Cambridge: Cambridge University Press. Ito, T. 2002. Is foreign exchange intervention effective? The Japanese experiences in the 1990s. NBER Working Paper 8914. Jaeger, A. 2003. The ECB’s money pillar: An assessment. IMF Working Paper 03/82.

References

253

Jakab, Z. M., and M. A. Kovacs. 1999. Determinants of real-exchange rate fluctuations in Hungary. NBH Working Paper 1999/6. Jansen, D., and J. de Haan. 2003. Statements of ECB officials and their effect on the level and volatility of the euro-dollar exchange rate. Journal of International Money and Finance, forthcoming. Jansen, W. J., and A. C. J. Stokman. 2003. Foreign direct investment and international business cycle comovement. De Nederlandsche Bank. MEB series 2003-10. Jensen, H. 2001. Optimal degrees of transparency in monetary policymaking. CEPR Discussion Paper 2689. Kalemli-Ozcan, S., B. E. Sorensen, and O. Yosha. 2001. Economic integration, industrial specialization, and the asymmetry of macroeconomic fluctuations. Journal of International Economics 55: 107–37. Kaufman, S. 2003. The business cycle of European countries: Bayesian clustering of country-individual IP growth series. Oesterreichische Nationalbank Working Paper 83. Kieler, M. 2003. The ECB’s inflation objective. IMF Working Paper 03/91. Kieler, M., and T. Saarenheimo. 1998. Differences in monetary policy transmission? A case not closed. European Commission. DG for Economic and Financial Affairs, Economic Papers 132. Kim, S. 2003. Monetary policy, foreign exchange interventions, and the exchange rate in a unifying framework. Journal of International Economics 60: 355–86. King, M. 1997. Changes in UK monetary policy: Rules and discretion in practice. Journal of Monetary Economics 39: 81–97. King, M. R. 2003. Effective foreign exchange intervention: matching strategies with objectives. International Finance 6(2): 249–71. Knetter, M. 1993. International comparisons of pricing to market behavior. American Economic Review 83: 473–86. Knoppik C., and T. Beissinger. 2003. How rigid are nominal wages? Evidence and implications for Germany. Scandinavian Journal of Economics 105(4): 619–41. Koen, V., L. Boone, A. De Serres, and N. Fuchs. 2001. Tracking the euro. OECD Economics Department Working Paper 298. Korhonen, I. 2003. Some empirical tests on the integration of economic Activity between the euro area and the accession countries: A note. Economics of Transition 11(1): 1–20. Kremers, J. J. M., D. Schoenmaker, and P. J. Wierts. 2003. Financial Supervision in Europe. Northampton, MA: Edward Elgar. Krueger, J. T., and K. N. Kuttner. 1996. The Fed Funds Futures rate as a predictor of Federal Reserve policy. Journal of Futures Markets 16(8): 865–79. Krugman, P. R. 1991. Geography and Trade. Cambridge: MIT Press. Krugman, P. R. 1998. It’s back: Japan’s slump and the return of the liquidity trap. Brookings Papers on Economic Activity 49(2): 137–205. Kuijs, L. 2002. Monetary policy transmission mechanisms and inflation in the Slovak republic. IMF Working Paper 02/80.

254

References

Kumar, M. S., T. Baig, J. Decressin, C. Faulkner-MacDonagh, and T. Feyzioglu. 2003. Deflation: Determinants, risks, and policy options. IMF Occasional Paper 221. Kuttner, K. N. 2001. Monetary policy surprises and interest rates: Evidence from the Fed Funds Futures market. Journal of Monetary Economics 47(3): 523–44. Kuttner, K. N., and A. S. Posen. 1999. Does talk matter after all? Inflation targeting and central bank behavior. Institute for International Economics. Working Paper 99-10. Kuttner, K. N., and A. S. Posen. 2001. Inflation, monetary transparency, and G3 exchange rate volatility. In M. Balling, E. H. Hochreiter, and E. Hennessy, eds., Adapting to Financial Globalization. London: Routledge. Lamfalussy, A. 1997. Securing the Benefits of EMU. Address at the London School of Economics (http://www.ecb.int/emi/key/key08.htm). Lebow, D. E., and J. B. Rudd. 2003. Measurement error in the consumer price index: Where do we stand? Journal of Economic Literature 41: 159–201. Lohmann, S. 1992. Optimal commitment in monetary policy: Credibility versus flexibility. American Economic Review 82(1): 273–86. Lorenzen, H., and N. Thygesen. 2000. The relation between the euro and the dollar. Paper presented at the EPRU Conference on Perspectives on Danish and European Economic Policy. Copenhagen. Loungani, P., and N. Sheets. 1997. Central bank independence, inflation, and growth in transition economies. Journal of Money, Credit, and Banking 29(3): 381–99. Lybek, T. 1999. Central bank autonomy, and inflation and output performance in the Baltic states, Russia, and other countries of the former Soviet Union, 1995–97. IMF Working Paper 99/4. Maeso-Fernandez, F., C. Osbat, and B. Schnatz. 2001. Determinants of the euro real effective exchange rate: A beer/peer approach. ECB Working Paper 85. Maliszewski, W. S. 2000. Central bank independence in transition economies. Economics of Transition 8(3): 749–89. Maliszewski, W. S. 2002. Monetary policy in transition: Structural econometric modeling and policy simulations. Center for Social and Economic Research, Warsaw. Studies and Analyses 246. Maurel, M. 2002. On the way of EMU enlargement towards CEECs: What is the appropriate exchange rate regime? CEPR Discussion Paper 3409. Massmann, M., and J. Mitchell. 2003. Reconsidering the evidence: Are Eurozone business cycles converging? ZEI Discussion Paper B05-2003. McCallum, B. M. 2000. Theoretical analysis regarding the zero lower bound on nominal interest rates. Journal of Money, Credit, and Banking 32(4, pt. 2): 49–72. McCarthy, J. 1999. Pass-through of exchange rates and import prices to domestic inflation in some industrialized economies. Mimeo. Bank for International Settlements, Basle. Meade, E., and D. N. Sheets. 2002. Regional Influences on U.S. monetary policy: Some implications for Europe. Centre for Economic Performance Discussion Paper 523, London, LSE; forthcoming in: Journal of Money, Credit, and Banking.

References

255

Melitz, J. 2001. Geography, trade and currency union. CEPR Discussion Paper 2987. Mélitz, J., and F. Zumer. 1999. Interregional and international risk-sharing and lessons for EMU. Carnegie-Rochester Conference Series on Public Policy 51: 149–88. Meltzer, A. H. 2001. The transmission process. In Deutsche Bundesbank, The Monetary Transmission Process: Recent Developments and Lessons for Europe. London: Macmillan. Micco, A., E. Stein, and G. Ordoñez. 2003. The currency union effect on trade: Early evidence from EMU. Economic Policy, no. 37: 315–56. Mishkin, F. S. 1997. The causes and propagation of financial instability: Lessons for policy makers. In Federal Reserve Bank of Kansas City, Maintaining Financial Stability in a Global Economy. Kansas City: Federal Reserve Bank of Kansas City. Mishkin, F. S. 2001. The transmission mechanism and the role of asset prices in monetary policy. NBER Working Paper 8617. Mojon, B. 2000. Financial structure and the interest rate channel of ECB monetary policy. ECB Working Paper 40. Mojon, B., and G. Peersman. 2002. A VAR description of the effects of monetary policy in the countries of the euro area. ECB Working Paper 93. Muscatelli, A. 1998. Optimal inflation contracts and inflation targets with uncertain central bank preferences: Accountability through independence. Economic Journal 108: 529–42. Natalucci, F. M., and F. Ravenna. 2002. The road to adopting the euro: Monetary policy and exchange rate regimes in EU candidate countries. Board of Governors of the Federal Reserve System, International Finance Discussion Papers 741. NBP. 1998. National Bank of Poland. Annual Report 1998. Nitsch, V. 2004. Have a break, have a . . . national currency: When do monetary unions fall apart? CESifo Working Paper 1113. Nolan, C., and E. Schaling. 1996. Monetary policy uncertainty and central bank accountability. Bank of England Working Paper 54. Obstfeld, M. 2002. Exchange Rates and Adjustment: Perspectives from the New Open Economy Macroeconomics. NBER Working Paper 9118. Ortega, E. 2003. Persistent Inflation Differentials in Europe. Banco de España Working Paper 0305. Otto, G., G. Voss, and L. Willard. 2001. Understanding OECD output correlations, Reserve Bank of Australia. Research Discussion Paper 2001-5. Oosterloo, S., and J. de Haan. 2003. A survey of institutional frameworks for financial stability. De Nederlandsche Bank, Occasional Study 4. Owens, R. E., and R. H. Webb. 2001. Using the Federal Funds Futures market to predict monetary policy actions. Federal Reserve Bank of Richmond Economic Quarterly 87(2): 69–77. Parsley, D., and S.-J. Wei. 1996. Convergence to the law of one price without trade barriers or currency fluctuations. Quarterly Journal of Economics 111: 1211–36.

256

References

Pautola, N., and P. Backé. 1998. Currency boards in central and eastern Europe: Past experience and future perspectives. Oesterreichische Nationalbank. Focus on Transition 1/1998: 72–113. Payne, R., and P. Vitale. 2003. A transaction level study of the effects of central bank intervention on exchange rates. Journal of International Economics 61: 331–52. Pelkmans, J., D. Gros, and J. Nunez Ferrer. 2000. Long-run economic aspects of the European Union’s eastern enlargement. WRR Working Document W109. The Hague. Persson, T. 2001. Currency unions and trade: How large is the treatment effect? Economic Policy, no. 33: 453–48. Poole, W., and R. H. Rasche. 2000. Perfecting the market’s knowledge of monetary policy. Journal of Financial Services Research 18(2): 255–98. Posen, A. S. 1998. Central bank independence and disinflationary credibility: A missing link? Oxford Economic Papers 50: 335–59. Posen, A. S. 2003. Six practical views of central bank transparency. In P. Mitzen, ed., Central Banks, Monetary Theory and Policy: Essays in Honour of Charles Goodhart, vol. 1. London: Edward Elgar. Radzyner, O., and S. Riesinger. 1997. Central bank independence in transition: Legislation and reality in central and eastern Europe. Oesterreichische Nationalbank. Focus on Transition 1/1997: 57–90. Rawaswamy, R., and T. Sløk. 1998. The real effects of monetary policy in the European Union: What are the differences? IMF Staff Papers 45: 374–96. Reifschneider D., and J. C. Williams. 2000. Three lessons for monetary policy in a low inflation era. Journal of Money, Credit, and Banking 32: 936–66. Remsperger, H. 2003. Inflation differentials in EMU: Causes and implications. Presentation at the CEPR/ESI Seventh Annual Conference on The Euro Area as an Economic Entity, Eltville, September 13, 2003. Rogers, J. H. 2001. Price level convergence, relative prices and inflation in Europe. Board of Governors, Federal Reserve System. International Finance Discussion Paper 699. Rogoff, K. 1985. The optimal degree of commitment to an intermediate monetary target. Quarterly Journal of Economics 100(4): 1169–89. Rose, A. 2000. One money, one market: The effect of common currencies on trade. Economic Policy, no. 30: 7–46. Rose, A. 2004. The effect of common currencies on international trade: A meta-analysis. In G. M. von Fürstenberg, V. Alexander, and J. Mélitz, eds., Monetary Unions and Hard Pegs: Effects on Trade, Financial Development, and Stability. Oxford: Oxford University Press. Rose, A., and E. van Wincoop. 2001. National money as a barrier to trade: The real case for currency union. American Economic Review 91(2): 386–90. Ross, K. 2002. Market predictability of ECB monetary policy decisions: A comparative examination. IMF Working Paper 02/233. Rother, C. P. 2000. The impact of productivity differentials on inflation and the real exchange rate: An estimation of the Balassa-Samuelson effect in Slovenia. IMF Country Report 00/56: 26–39.

References

257

Roubini, N. 1999. The case against currency boards: Debunking 10 myths about the benefits of currency boards. Mimeo, Stern NYU (http://www.stern.nyu.edu/~nroubini/asia/ AsiaHomepage.html). Rudebusch, G., and L. E. O. Svensson. 2002. Eurosystem monetary targeting: Lessons from U.S. data. European Economic Review 46: 417–42. Rudin, J. R. 1988. Central bank secrecy, “Fed watching”, and the predictability of interest rates. Journal of Monetary Economics 22(2): 317–34. Saacke, P. 2002. Technical analysis and the effectiveness of central bank intervention. Journal of International Money and Finance 21: 459–79. Samuelson, P. 1964. Theoretical Notes on Trade Problems. Review of Economics and Statistics 46: 145–54. Sarno, L., and M. P. Taylor. 2001. Official intervention in the foreign exchange market: Is it effective and, if so, how does it work. Journal of Economic Literature 39(3): 839–68. Schulmeister, S. 2000. Die Kaufkraft des Euro Innerhalb und Ausserhalb der Waehrungsunion. WIFO Studie. Sibert, A. 2003. The new monetary policy strategy of the ECB. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. May. Siklos, P. 2002. The Changing Face of Central Banking: Evolutionary Trends since World War II. Cambridge: Cambridge University Press. Sims, G. T., and D. Wessel. 2000. The European central bank can’t master communication. Wall Street Journal. April 27. Sinn, H.-W. 2001. Chancellor Schroeder’s “steady hand” must act now. Ifo Viewpoint August 29. Sinn, H.-W., and M. Reutter. 2001. The Minimum Inflation Rate for Euroland, NBER Working Paper 8085. Smets, F. 2000. What horizon for price stability? ECB Working Paper 24. Söderlind, P. 1995. Forward Interest Rates as Indicators of Inflation Expectations. CEPR Discussion Paper 1313. Söderlind, P., and L. E. O. Svensson. 1997. New techniques to extract market expectations from financial instruments. Journal of Monetary Economics 40(2): 383–429. Stein, J. 2001. The equilibrium value of the euro/$US exchange rate: An evaluation of research. CESifo Working Paper 525. Stiglitz, J. 1998. Central banking in a democratic society. De Economist 146: 199–226. Sturm, J.-E., and J. De Haan. 2002. Inflation in developing countries: Does central bank independence matter? New evidence based on a new data set. Ifo Studien 47(4): 389–403. Summers, L. H. 1991. How should long-term monetary policy be determined? Journal of Money, Credit, and Banking 23(August, pt. 2): 625–31. Svensson, L. E. O. 1993. Term, inflation, and foreign exchange risk premia: A unified treatment. NBER Working Paper 4544.

258

References

Svensson, L. E. O. 1997. Optimal inflation targets, “conservative” central banks, and linear inflation contracts. American Economic Review 87(1): 98–114. Svensson, L. E. O. 1999. Monetary policy issues for the Eurosystem. Carnegie-Rochester Conference Series on Public Policy 51: 79–136. Svensson, L. E. O. 2000. How should monetary policy be conducted in an era of price stability? NBER Working Paper 7516. Svensson, L. E. O. 2001. The zero bound in an open economy: A fool prove way of escaping from the liquidity trap. Bank of Japan Economic and Monetary Studies 19(S-1): 277–312. Svensson, L. E. O. 2001b. What is good and what is bad with the Eurosystem’s published forecasts, and how can they be improved? Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. February. Svensson, L. E. O. 2003. In the right direction, but not enough: The modification of the monetary-policy strategy of the ECB. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. Szapary, G. 2000. Maastricht and the choice of the exchange rate regime in transition countries during the run-up to EMU. NBH Working Paper 7. Tabellini, G. 1987. Secrecy of monetary policy and the variability of interest rates. Journal of Money, Credit, and Banking 19(4): 425–36. Tarkka, J., and D. Mayes. 1999. The value of publishing official central bank forecasts. Bank of Finland Discussion Paper 22/99. Taylor, M. P. 2004. Is official exchange rate intervention effective? Economica 71(1): 1–11. Teïletche, J. 2000. La parité euro/dollar durant les décennies 80 et 90: Peut-on trouver une spécification raisonable et à quel horizon? Mimeo. The Economist. 1998. Euro towers or Fawlty Towers? October 31. Thygesen, N. 2002. The path to the euro for enlargement countries. Briefing paper for the Economic and Monetary Affairs Committee, European Parliament. Second quarter 2002. Thornton, D. 2002. Monetary policy transparency: Transparent about what? Paper prepared for the MMF Monetary Policy Transparency conference. May 10. Bank of England, London. Toolsema, L. A., J.-E. Sturm, and J. De Haan. 2002. Convergence of monetary transmission in EMU, new evidence. Mimeo. University of Groningen (rev. CESifo working paper 465). Trichet, J.-C. 2001. The euro after two years. Journal of Common Market Studies 39(1): 1–13. Trichet, J.-C. 2002. The European monetary union and the EU enlargement. Speech delivered at the Munich Economic Summit, 8 June 2002. Ullersma, C. A. 2002. The zero lower bound on nominal interest rates and monetary policy effectiveness: A survey. De Economist 150(3): 273–97. Vaubel, R. 1997. The bureaucratic and partisan behavior of independent central banks: German and international evidence. European. Journal of Political Economy 13(2): 201–24.

References

259

Van Aarle, B., M. Boss, and J. Hlouskova. 2000. Forecasting the euro exchange rate using vector error correction models. Weltwirtschftliches Archiv 136(2): 232–58. Van Els, A. Locarno, J. Morgan, and J.-P. Villetelle. 2001. Monetary policy transmission in the euro area: What do aggregate and national structural models tell us? ECB Working Paper 95. Vickers, J. 1999. Monetary policy and asset prices. Lecture given at the Money, Macro and Finance Group 31st Annual Conference, Oxford University. September 22. Viñals, J. 2001. Monetary policy in a Low Inflation Environment. Banco d’España Working Paper 0107. Vitale, P. 1999. Sterilized central bank intervention in the foreign exchange market. Journal of International Economics 49(2): 245–67. Vitale, P. 2003. Foreign exchange intervention: How to signal policy objectives and stabilise the economy. Journal of Monetary Economics 50: 841–70. Von Hagen, J., and B. Hoffman. 2003. Macroeconomic implications of low inflation in the euro area. ZEI Working Paper B29-2003. Wagner, H. 1999. Central bank independence and the lessons for transition economies from developed and developing countries. Comparative Economic Studies 41(4): 1–22. Wallis, K. F. 1999. Asymmetric density forecasts of inflation and the Bank of England’s fan chart. National Institute Economic Review (January): 106–12. Walsh, C. E. 1995. Optimal contracts for central bankers. American Economic Review 85(1): 150–67. Walsh, C. E. 1999. Announcements, inflation targeting and central bank incentives. Economica 66: 255–69. Winkler, B. 2000. Which kind of transparency? On the need for clarity in monetary policymaking. ECB Working Paper 26. Wren-Lewis, S., and R. Driver. 1998. Real exchange rates for the Year 2000. Washington, DC: Institute for International Economics. Wyne, M. A., and J. Koo. 2000. Business cycles under monetary union: A comparison of the EU and US. Economica 67: 347–74. Wynne, M. A., and D. Rodríguez-Palenzuela. 2002. Measurement bias in the HICP: What do we know and what do we need to know? ECB Working Paper 131. Wyplosz, C. 2001. Do we know how low inflation should be? In A. Garcia Herrero, V. Gaspar, L. H. Hoogduin, J. Morgan, and B. Winkler, eds., Why Price Stability? Frankfurt: ECB. Wyplosz, C. 2003a. Exchange rate policy of the euro area. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. December. Wyplosz, C. 2003b. Accession countries and ERMII. Briefing paper for the Committee on Economic and Monetary Affairs of the European Parliament. September. Yates, T. 2002. Monetary policy and the zero bound to interest rates: a review. ECB Working Paper 190.

Index

Accountability of the European Central Bank, 108–12 Acquis communautaire, 170, 180 Asset price inflation, 74–78 Balassa-Samuelson effect, 42–43, 153–58, 201–205, 226 Blinder’s survey on credibility, 112–17 Boskin Report, 40 Business cycles differences and inflation, 158–60 implications of enlargement of the monetary union, 192–201 synchronization, 137–48, 192–201, 226–27 trade relationships and, 188–89 Central bank governor turnover rate, 172–76 Central bank independence (CBI), 170–77 Code of Good Practices on Transparency in Monetary and Financial Policies of the IMF, 3–4 Consumer spending asset price inflation and, 74–77 patterns, 149–50 Convergence criteria, 177–79 Credibility Blinder’s survey on, 112–17 of European Central Bank, 117–23 Currency boards, 180–83 Decentralization diverging business cycles and, 137–48 risks of, 130–37 structure of the European Central Bank, 5–8, 125–28

Decision making by the European Central Bank, 133–35 Deflationary policy, 39–47 Delors, Jacques, 2 Delors Committee, 2 Demand money, 56 shocks, 196–98 Dependence on trade, 150 Deposit facility, 15 Disclosure accountability and, 108–12 economic, 86, 89–90 effects of, 92–94 operational, 91–92, 98–99, 102–103, 105 policy, 86, 91 political, 85, 87–89, 94, 96–97 procedural, 86, 90–91, 97–98 various facets of, 85–87 Economic and Monetary Union (EMU), the, start of, 1–2 Economic disclosure, 86, 89–90 Economist, The, 5 EONIA (Euro Over-Night Index Average), 15 Euro, the and dollar exchange rate, 58–63, 104 exchange rate pass-through (ERPT), 59–63 external value of, 57–74 interventions in exchange rates of, 63–74 Over-Night Index Average (EONIA), 15 trade enhancing effect of, 189

262

European Central Bank (ECB). See also Monetary policy, European Central Bank accountability of, 108–12 centralization of, 215–16 credibility of, 117–23 decentralized structure of, 5–8 decision making by, 133–35 Executive Board, 11 favorable assessment of, 225 General Council of, 11 Governing Council, 3, 5, 9, 10–11, 13–14, 18, 56–57, 213–14 inflation targeting by, 225–26 interest decisions by, 16–19 Lombard facility, 15 marginal lending facility, 15 minimum reserve system, 16 money market, 15 M3, 12–18, 47–56 new EU members and, 6–7 newspaper reports on, 106–107 political independence of, 9, 221–22 political weight of members of, 213–14 reform options, 215–23 research output of, 129 standing facilities, 15 start of, 1–2 structure of, 10–11 transparency of, 3–5 European Forecasting Network, 193–94 European Monetary Institute, 56, 137 European Union, new members of the, 6–7, 169–70 business cycle synchronization by, 192–201 central bank independence and, 170–77 currency boards and, 180–83 implications of, 192–209 inflation among, 201–205 Maastricht criteria for, 177, 179–83 monetary transmission and financial structure in, 205–209 optimal currency area theory and, 183–92 quick entry for, 226 Exchange market interventions, 18, 63–74 Exchange rate optimal currency area theory and, 183–92 pass-through (ERPT), 59–63 as source of shocks, 190

Index

Exchange Rate Mechanism (ERM), 2, 6, 178–79 Executive Board of the European Central Bank, 11 Expectations, financial market, 32–36 Federal Reserve Bank, US, 5, 94, 96, 122, 126, 130 Financial market indicators, 13 expectations according to newspaper reports, 32–36 Financial Times, The, 106–107 First pillar of monetary policy strategy, 12–13, 47–56 Forecasts, inflation, 56–57 Forward rates, analysis of, 28–32 General Council of the European Central Bank, 11 Governing Council, European Central Bank, 3, 5, 9, 10–11, 13–14 and Boskin Report, 40–41 decentralization and, 126, 128, 131–32 decisions on interest rates, 18–19 evaluation of monetary policy strategy by, 21–25 inflation forecasts by, 56–57 options for reform of, 215–17 reforms proposed by, 219–22 rotation, 219–20 structure of, 213–14 voting structure reform of, 217–19 Greenspan, Alan, 40, 43 Harmonized Index of Consumer Prices (HICP), 2–3, 11–12, 13, 204 euro–dollar exchange rate and, 58–59 price index and, 37–39 Inflation, 12, 13–14 asset price, 74–77 Balassa-Samuelson effect on, 42–43, 153–58, 201–205, 226 below 2 percent, 39–47 central bank independence and, 173 differentials in the euro area, 148–60 forecasts, 56–57 measurement bias in, 40–41 monetary policy effect on, 163–67 money growth and, 54–55 new EU members and, 201–205 objective of price stability policy, 36–47

Index

Policy Target Agreement (PTA) and, 109 price index calculation of, 37–39 price level convergence and, 151–53 targeting, 225–26 temporary differentials in, 42–43 transparency and, 88–89, 93–94 wages and unemployment and, 44–47 Integration financial, 146–47 monetary, 144–45, 226–27 Interest rates forward rates and, 28–32 monetary policy and, 16–21 stepping versus smoothing, 78–80 Interregional migration, 188 Interventions, euro–dollar exchange rate, 63–74 Issing, Otmar, 4, 21–22 Labor market indicators, 13 Lafontaine, Oskar, 16 Lombard facility, 15 Maastricht Treaty, 2, 5, 9, 11, 169, 177, 179–83 Marginal lending facility, 15 Measurement bias, 40–41 Minimum reserve system, 16 Monetary policy, European Central Bank. See also European Central Bank (ECB) criticism of, 27 diverging business cycles and, 137–48, 226–27 effects on output and inflation, 163–67 evaluation of, 21–25 exchange rate pass-through (ERPT) and, 59–63 expectations according to newspaper reports, 32–36 interest rates and, 16–21 money demand and, 56 money growth pillar in, 12–13, 47–56 pass-through of, 161–63 predictability of, 27–36 responsibilities, 2–8 strategy, 11–16 transmission and financial structure, differences in, 160–65, 205–209 two pillars of, 12–14, 47–56, 57 Money growth, 12–13, 47–56 M3, 12–13, 47–56

263

Newspaper reports, 32–36, 106–107 Open market operations, 15–16 Operational disclosure, 86–87, 91–92, 98–99, 102–103, 105 Optimal currency area theory, endogeneity of OCA criteria, 183–92 Output, monetary policy effects on, 163–67 Pass-through of monetary policy decisions, 161–63 Pillars of monetary policy, 12–14, 47–56, 57 Policy disclosure, 86, 91 Policy Target Agreement (PTA), 109 Political disclosure, 85, 86, 87–89, 94, 96–97 Political independence of the European Central Bank, 9, 221–22 Predictability of ECB policy decisions, 27–36 Price Balassa-Samuelson effect and, 153–58 index, 37–39 level convergence, 151–53 stability, 2–3, 11–12, 13–14, 19 inflation objective of, 36–47 Procedural disclosure, 86, 90–91, 97–98 Refinancing operations, 15–16 Reform of European Central Bank to increase centralization, 215–17 options for, 215–19 political independence and, 221–22 2002 proposal for, 220–23 to reduce overrepresentation of some members, 213–14 rotation, 218–19 voting, 217–18 Research output of European Central Banks, 129 Risks of decentralization, 130–37 Rotation, European Central Bank Governing Board, 218–19 Second pillar of monetary policy strategy, 13–14 Shocks, demand and supply, 196–98 Specialization and business cycle synchronization, 144 Standing facilities, monetary policy, 15

264

Stepping versus smoothing, interest rate, 78–80 Structure of the European Central Bank, 10–11 Supply shocks, 196–98 Synchronization, business cycle, 138–48, 192–201, 226–27 Taylor rule, 44 Transmission, monetary policy, 160–65, 205–209 Transparency of the European Central Bank, 3–5 definition of, 83 effects of disclosure on, 92–94 importance of, 83–85 objectives and indexes of, 94–107 various facets of disclosure and, 85–94 Treaty on European Union (TEU), 109 Two-pillar strategy of the European Central Bank, 12–14, 24 Vector autoregressions (VARs), 205–207 Wages effect of inflation on, 44–47 flexibility in new EU member countries, 185–86 World Economic Survey, 117 Zero bound, 43–44

Index