The acceding countries' strategies towards ERM II …In 2004 all ECB publications will feature a...

OCCAS IONAL PAPER S ER I E SNO. 10 / F EBRUARY 2004

THE ACCEDINGCOUNTRIES’STRATEGIES TOWARDSERM II AND THEADOPTION OF THEEURO: AN ANALYTICALREVIEW

by a staff team led by Peter Backé and Christian Thimannand includingOlga Arratibel,Oscar Calvo-Gonzalez,Arnaud Mehl and Carolin Nerlich

In 2004 all ECB publications will feature

a motif taken from the

€100 banknote.

OCCAS IONAL PAPER S ER I E SNO. 10 / F EBRUARY 2004

THE ACCEDINGCOUNTRIES’

STRATEGIES TOWARDSERM II AND THE

ADOPTION OF THE EURO: AN ANALYTICAL

REVIEW

by a staff team led by Peter Backé and Christian Thimann

and includingOlga Arratibel,

Oscar Calvo-Gonzalez,Arnaud Mehl and Carolin Nerlich

This paper can be downloaded from the ECB’s website (http://www.ecb.int).

© European Central Bank, 2004

AddressKaiserstrasse 2960311 Frankfurt am MainGermany

Postal addressPostfach 16 03 1960066 Frankfurt am MainGermany

Telephone+49 69 1344 0

Websitehttp://www.ecb.int

Fax+49 69 1344 6000

Telex411 144 ecb d

All rights reserved. Reproduction foreducational and non-commercial purposesis permitted provided that the source isacknowledged.

The cut-off date for the statistics in this paperwas January 2004.

ISSN 1607-1484 (print)ISSN 1725-6534 (online)

3ECB

Occa s i ona l Pape r No . 10February 2004

CONTENT SINTRODUCTION 4

1 THE ROLE OF ERM II 6

1.1 ACCEDING COUNTRIES’ ANDECB/EUROSYSTEM PERSPECTIVE 6

1.1.1 Acceding countries’ perspective 6

1.1.2 The Eurosystem position 8

1.2 RATIONALE AND FEATURES OF ERM II 13

1.3 ECONOMIC ASPECTS OF ERM II 14

1.3.1 Current exchange rate regimes 14

1.3.2 Regime changes in view ofERM II membership 18

2 CHALLENGES OF EURO ADOPTION 23

2.1 ACCEDING COUNTRIES’ PERSPECTIVE 23

2.2 A REVIEW OF THE PROGRESS MADEIN NOMINAL CONVERGENCE 25

2.3 A BROADER ECONOMIC ASSESSMENT 28

2.3.1 Economic structures 29

2.3.2 Trade integration 31

2.3.3 Labour market flexibility 33

2.3.4 Financial sector integration 39

2.3.5 Real convergence and structuraldifferences in economicdynamics 42

2.3.6 Adjusting interest rates to euroarea levels 51

2.3.7 External adjustments:the impact of catching-up oncompetitiveness 54

2.3.8 Fiscal consolidation 56

CONCLUDING REMARKS 60

REFERENCES 62

4ECBOcca s i ona l Pape r No . 10February 2004

The ten countries joining the European Union inMay 2004 are intensively discussing issuesrelated to their future participation in ERM IIand the subsequent adoption of the euro. Manyof them have tabled concrete strategies forfurther monetary integration within the EU,elaborating in particular on the intended timingof ERM II entry and euro adoption. In somecases strategies announced in 2001 or 2002have recently been modified or refined. Mostcountries appear to be opting for a relativelyshort participation in ERM II. However, somecountries have shifted their preferred euroadoption dates backward by one or two years,compared with their original plans, which inturn implies that ERM II entry is in these casestargeted only some time after EU accession.This paper presents an analytical review ofthe acceding countries’ strategies towardsfurther monetary integration, with the aim ofcontributing to the ongoing dialogue with theacceding countries’ central banks. In particular,the paper aims at examining the economicrationale of the strategies declared by mostacceding countries so that also potential risksinvolved can be identified. It is worth notingthat this paper does not explore in much detailthe benefits of euro adoption per se, as it takesthe institutional framework as given, whichforesees the introduction of the single currencyin the acceding countries at some future point(with no opt-out clauses). Rather, the paperfocuses on the timing of euro adoption, i.e. onthe economic merits and risks of an earlyadoption of the euro compared with those ofintroducing the euro at a later stage, in each ofthe country cases under consideration.

The paper should not be seen as a form ofconvergence assessment, but rather as aconceptual framework to study the accedingcountries’ declared strategies. Judgements aretentative and may change over time, as newinformation becomes available and additionallyas the acceding countries’ strategies continue toevolve.

While at the current juncture the policy focuslies on ERM II issues, the analysis needs to be

seen in light of the countries’ medium to longer-term policy plans up to the adoption of the euro,as these plans set the context for the timing andthe possible modalities of ERM II participation.For countries that can realistically aim atintroducing the euro in the medium term, issuesrelated to ERM II participation may be quitedifferent than the issues facing countries forwhich the prospect of euro adoption is moreremote. Indeed, a full examination of theconsistency of ERM II strategies has to extendto the question of whether these strategies are inline with the eventual endpoint of monetaryintegration. In a way, one could therefore seethe optimal policy planning as being solvedbackwards. This paper, however, takes amore chronological approach, following thesequence outlined in the Treaty, and reviewsfirst the considerations underlying ERM IIparticipation, and then turns to issues related tofull monetary integration in the euro area.

Almost all of the ten acceding countriesoriginally declared their intention to adopt theeuro a few years after EU entry; since then,some have more recently moved towards asomewhat more extended timeline. Followingthe procedures laid down in the Treaty, someaim at introducing the euro at the beginning of2007 (or perhaps already in the later part of2006), subsequent to a two-year period withinERM II starting around mid-2004 and a positiveconvergence assessment thereafter. The mainmotives for the intended early adoption of theeuro are to advance economic and financialintegration with the euro area, to anchordomestic policies and to lower risk premia,thereby fostering economic growth and real

1 Special contributions by Ralph Süppel during his stay at the ECB aregratefully acknowledged, together with contributions by Iikka Korhonen(from the Bank of Finland) and Cezary Wojcik, who provided updatedempirical results, and Stefan Wredenborg, who helped with thequantitative research.

2 This paper has been discussed extensively in the ECB/ESCB and hasbenefited from substantive comments in the process, which are gratefullyacknowledged. The authors would like to thank in particular TommasoPadoa-Schioppa, Pierre van der Haegen, members of the InternationalRelations Committee and representatives of the acceding countries’central banks for their helpful suggestions. It is worth recalling that theviews expressed in this paper are those of the authors and do notnecessarily reflect those of the ECB.

I N T RODUCT I ON 1 ,2

5ECB


INTRODUCTION

convergence. Those countries that have recentlybegun to envisage a somewhat longer processbefore adopting the euro, on the other hand,mostly refer to 2008, 2009 or 2010 as targetyears for joining the euro area.

This paper reviews the economic considerationssurrounding entry into ERM II for the accedingcountries as well as those relating to theadoption of the euro. Wherever necessary andappropriate, the paper differentiates amongacceding countries, given the significant degreeof diversity among them. The considerations ofthis paper assume that the main interest of theECB is in ensuring an enlargement process thatis smooth and follows Treaty procedures,thereby protecting the credibility of the ERM IImechanism as well as the stability of the euroarea. This implies that the timing for ERM IIentry and euro adoption need to be broadlyappropriate and based on economic logic.

The analysis of the paper is necessarilytentative, as the underlying theory, notably onoptimum currency areas (OCA) and the choiceof exchange rate regimes, provides only limitedguidance. Indeed, both the costs and benefits ofmonetary union – let alone their balance andaspects related to timing – are difficult toquantify. Moreover, the time horizon ofavailable data is short and is influenced byseveral important shocks that the central andeastern European economies have experiencedin recent years. Still, the paper may be useful inclarifying some of the issues at stake and inproviding a conceptual framework of how toanalyse and review the acceding countries’strategies towards ERM II and the euro.

The paper is organised as follows. Part 1explores the issues related to ERM II. It recallsthe main views of the acceding countries and theECB/Eurosystem on the Exchange RateMechanism, and examines the constraints onexchange rate policies within ERM II, togetherwith the likely changes in de facto exchange ratepolicy for individual countries. Part 2 analysesissues related to a future adoption of the euro,with a particular focus on aspects that are

relevant for timing. It deliberately looks beyondconsiderations relating to the Maastrichtcriteria, and includes standard OCA indicatorsas well as a discussion of catching-up-related issues, focusing in particular onthe implications of structural differences ineconomic dynamics. Some tentative conclusionsare offered at the end of the paper.


This section reviews some of the economicissues related to the acceding countries’participation in ERM II. It consists of threeparts: an overview of the current policystatements regarding ERM II of both theacceding countries and the Eurosystem; a shortreview of the main features of ERM II, as setout in the Resolution by the European Council;and an exploration of the economic issuesrelated to participation in ERM II.

I.1 ACCEDING COUNTRIES’ ANDECB/EUROSYSTEM PERSPECTIVE

1.1.1 ACCEDING COUNTRIES’ PERSPECTIVE

Most acceding countries regard ERM II as aninstitutional requirement for the adoption of theeuro that cannot be avoided, but whoseappropriateness as an exchange rate policyframework is questionable. Overall, theyperceive ERM II as a “waiting room” that offersat best little value-added and may even entailcertain risks. In their policy statements, manyacceding countries have suggested that theirparticipation in ERM II will be characterised bythe following features:

– The duration of participation will be limitedto only two years.3

– Actual exchange rate management shall leantowards a narrow band from the verybeginning of participation. The reason forthis is that the relevant interpretation of theMaastricht criterion is seen as pointing to themaintenance of the exchange rate “close tothe central rates”, which in turn is interpretedas de facto meaning a narrower (possibly±2.25%) range, with perhaps someflexibility upward but not downward.4

– The central rate should be chosen so that itfacilitates the kind of exchange ratemanagement described before, and shouldideally be the future conversion rate.

1 TH E RO L E O F E RM I IThe rationale for such an interpretation isbased on a negative assessment of ERM II in thefollowing four main policy areas: (i) discipline,(ii) credibility, (iii) adjustability, and(iv) multilateralism.

– Discipline: Acceding countries value thedisciplinary impact of ambitious deadlinesfor euro adoption on macroeconomic andstructural policies, but see little disciplinaryeffect arising from ERM II. Moreover, interms of fiscal policy, they often refer to thedisciplinary impact of the multilateralsurveillance exercise that will be appliedupon EU accession in any case. Anydisciplinary impact in addition to theMaastricht criteria for euro adoption and theStability and Growth Pact (SGP) isconsidered negligible.

– Credibility: Acceding countries regardERM II as an intermediate exchange rateregime, subject to risks of speculativeattacks. The ERM crisis in 1992/93 and therecent experience of Hungary, which faced apolicy dilemma between its inflation targetand its exchange rate objective, are seen asexamples of these risks. In addition, thecurrent monetary arrangements – rangingfrom currency boards arrangements (CBAs)to free floats with inflation targeting – aregenerally seen as credible overall, asevidenced in mostly low inflation rates,declining policy interest rates and well-entrenched expectations of low inflation.Some officials argue that either their countryis ready to adopt the euro, in which caseERM II is not needed as the exchange rateshould be stable anyway, or it is not, inwhich case ERM II is not a sufficiently safeframework for exchange rate andmacroeconomic policy management.

3 Hungary intends to participate in ERM II for a somewhat longer timeperiod (see also Box 1).

4 Poland would prefer to make use of the standard fluctuation bandwhile participating in ERM II (see Box 1).

7ECB


1 The roleof ERM II

Box 1

SELECTED STATEMENTS BY ACCEDING COUNTRIES ON ERM II (EXTRACTS)

Cyprus “It is envisaged that the current ERM II-type exchange rate frameworkshall continue to be in place until the eventual adoption of the euro.” Theintention of the authorities in Cyprus is “an early participation in the ERM IImechanism” which derives from “the authorities’ aim to join the euro zoneas early as possible after accession.” Pre-accession Economic Programme(PEP) 2003.

Czech Rep. “The mere participation of a currency in [the] ERM II regime does not eliminatemonetary turbulence (unlike irreversible fixation of the exchange rate within theframework of a monetary union). Therefore the participation in ERM II can beperceived only as a prerequisite for joining the euro zone and the central bankdoes not consider a longer-than-necessary stay in ERM II to be desirable. In linewith this view the Czech Republic should enter the ERM II only after conditionshave been created that will enable it to introduce the euro at the time of theassessment of the exchange-rate criterion (two years after joining the ERM II).In view of the development of the general government deficit expected within theframework of proposed public finance reform, the koruna would thereforeremain outside [the] ERM II system, even for some time after the accession ofthe Czech Republic to the EU.” PEP 2003.

Estonia “Estonia will present an application to join the ERM II exchange ratemechanism immediately after EU accession in 2004… Estonia wants to keep afixed exchange rate and the financial framework supporting the currency boardsystem up to euro area accession and as a part of the ERM II framework,taking it as a unilaterally binding obligation to hold exchange rate stability.”PEP 2003.

Hungary “The Government, in concert with the central bank, has expressed itscommitment to joining the ERM II… as soon as possible after we become [a]Member State of the EU. Unlike in the case of the vast majority of accessioncountries, this exchange rate mechanism is very similar to the Hungarianexchange rate system, which makes ERM II entry easier. Participation in ERM IIhas substantial benefits while it has no major risks. Its benefits include greatercredibility and the fact that, as no unilateral decision can be taken in ERM II, therisk premium expected by foreign investors because the exchange rate policydecisions of a given country, in our case Hungary, are not fully predictable, willbe reduced significantly.” PEP 2003.

Latvia “Upon reaching a mutual agreement between the Bank of Latvia and all involvedEU institutions, the Bank of Latvia plans that Latvia will join ERM II togetherwith changing the currency peg on 1st January 2005… Taking into account therequirement in the area of meeting the currency stability criterion and the so farsuccessful practice, the Bank of Latvia has no plans for using the exchange rate


fluctuation limitation boundaries in the amount of ±15% of the central parity…fully, but, in case of necessity, by using unilateral interventions, intends to limitthe lat fluctuations against the euro to a narrower fluctuation corridor withinthe framework of ERM II.” PEP 2003.

Lithuania “It is purposeful to join the ERM II, at the same keeping all the features of thecurrent exchange rate system.” PEP 2003. “We will propose to the governmentto jointly start talks on accession to ERM II… the earliest possible date foracceding to ERM II could be 1st May 2004, when Lithuania will join the bloc. Itmay also happen a little later, for instance, on 1st July in the coming year… Whileacceding to ERM II, Lithuania will seek to keep the current litas-euro rate butfluctuation margins will be zero, i.e. there will be no fluctuations in the litas ratewith respect to the euro.” Governor Sarkinas, May 2003.

Malta “It would be appropriate to [apply] soon after membership next May toparticipate in the ERM II by early 2005.” Minister Dalli, November 2003.

Poland “Poland should join the Economic and Monetary Union (EMU) so soon aspossible if only the macroeconomic conditions make it possible. During a two-year participation in the ERM II the zloty should be allowed to deviate from thecentral parity within a standard fluctuation band (±15%).” PEP 2003.

Slovakia “The time spent inside the ERM II should be as short as possible ... the countrycould join the ERM II in 2005.” Strategy of the Slovak Republic for Adoptionof the Euro, June 2003.

Slovenia “Slovenia intends to enter into the ERM II in the first half of 2005.” PEP 2003.Subsequently, in November 2003, the Slovenian government and the central bankagreed on a joint monetary integration strategy according to which Sloveniawould intend to join ERM II “by the end of 2004”.

– Adjustability: Acceding countries believe thatthe standard fluctuation band of ±15%, plusthe general possibility of upward realignments,might signal that the exchange rate couldappreciate by as much as 15% or even moreand thereby contribute to higher exchange ratevolatility. Moreover, efforts to stabilise theexchange rate within the band could causespeculative attacks and thus lead to excessiveexchange rate volatility rather than stability.

– Multilateralism: Acceding countries areuncertain about the involvement of the ECBwith respect to coordinated intra-marginalinterventions and even interventions at the

margin. As a result, they fear higher foreignexchange market uncertainty andspeculation.

1.1.2 THE EUROSYSTEM POSITION

The Eurosystem position regarding ERM II isset out in the “Policy position of the GoverningCouncil of the European Central Bank onexchange rate issues relating to the accedingcountries”, published on 18 December 2003.5

This document builds on earlier statements

5 See http://www.ecb.int (‘Press release’ section).

9ECB


1 The roleof ERM II

made by the Eurosystem on ERM II. Box 2contains a selection of these statements. In itscommunication, the Eurosystem put forward theview that ERM II should be seen as a usefulregime in its own right, arguing that a numberof policy challenges faced in the run-up to theeuro may well be tackled best within thatframework.

The Eurosystem has defended the rationale forERM II membership by referring to argumentsthat can be seen in the same four categories asthose of acceding countries, albeit with apositive assessment of the value of ERM II.

– Discipline: By requiring consistenteconomic policies, ERM II could help inproviding a more stable macroeconomicenvironment and could moreover act as acatalyst for structural reforms.

– Credibility: ERM II, with its announcedcentral parity, would provide guidance toparticipants in foreign exchange markets,and may thereby contribute to greaterexchange rate stability. Moreover, byanchoring inflation expectations andreducing exchange rate volatility, ERM IImay also contribute to lowering the level andvolatility of inflation. Unlike otherintermediate regimes, ERM II entailsultimate exit into the euro area, thus makingthe system more resilient than otheralternative exchange rate regimes.

– Adjustability: The standard fluctuation bandwould leave sufficient room to adjust toasymmetric shocks and structural changes inthe economy. Moreover, in the event that thecatching-up process is faster than expected, arevaluation of the central rate would bepossible.

– Multilateralism: The multilateral character ofERM II would be a feature that wouldenhance the credibility of the framework, asall parties would be engaged in monitoringeconomic and policy developments, and

assessing market reactions and possiblyultimately co-ordinating actions, if required.

The position of December 2003 recalls the dualrole of ERM II, which acts as an arrangementfor managing exchange rates between thecurrencies of Member States participating in themechanism and the euro, while at the same timeplaying a role in the convergence criteria forjoining the euro. It lays out the main featuresof the mechanism (see also Section 1.2 below)and, in doing so, maintains that the standardfluctuation band foreseen in ERM II isappropriate for Member States that are engagingin a convergence process. The position paperthen turns to the issues related to entry intoERM II. A key element in this context is that toensure smooth participation in the mechanism,major policy adjustments – for example withregard to price liberalisation and fiscal policy –have to be undertaken prior to participation inthe mechanism, and a credible fiscalconsolidation path needs to be followed. Thisnotwithstanding, entry into ERM II is notsubject to a set of pre-established criteria.

Regarding the length of ERM II participation,the document recalls the minimum period ofERM II membership of two years prior to theconvergence assessment, highlighting that thelength of participation in ERM II should beassessed in terms of what is most helpful toaccompany the convergence process, rather thanin terms of the required minimum period. Asacceding countries differ greatly in theireconomic structures, exchange rates andmonetary regimes, as well as in the degree ofnominal and real convergence already achieved,no single path towards ERM II can be identifiedand recommended. For some new MemberStates, it might be appropriate to considerapplication for ERM II participation only after afurther degree of convergence has beenachieved. This is particularly advisable when anearly rigidity in the exchange rate couldprecipitate disorderly realignments withpotentially disruptive economic consequences,including for the credibility of the mechanism


as a whole. For other new Member States thathave implemented significant structural reformsand have shown the ability to advanceconvergence through sound economic policiesand an exchange rate regime that is in principlecompatible with ERM II, entry into themechanism can take place soon after accession,provided that there is mutual agreement on thecentral parity. The position of December 2003also maintains that, in certain cases, newMember States may consider it desirable toenvisage a longer stay in ERM II while furtherconvergence takes place.

The position paper ends with a section on theadoption of the euro, which includes anexplanation of how the exchange rate stabilitycriterion has been and will continue to beapplied. The criterion refers to participation inERM II for a period of at least two years priorto the convergence assessment without severetensions, in particular without devaluingagainst the euro. The assessment of exchangerate stability against the euro will focus on theexchange rate being close to the central rate,while also taking into account factors that mayhave led to an appreciation, in line with whatwas done in the past. In this respect, the widthof the fluctuation band within ERM II shall notprejudge the assessment of the exchange ratestability criterion. Finally, the position paperdescribes how the ECB addresses the issue ofthe absence of “severe tensions”.

11ECB


1 The roleof ERM II

Box 2

SELECTED STATEMENTS BY THE EUROSYSTEM ON ERM II (IN REVERSE CHRONOLOGICAL ORDER)

“Achieving a high degree of nominal convergence and a significant degree of ‘institutional’convergence is essential for, first, smooth participation in the Exchange Rate Mechanism(ERM II) and, later on, successful membership in the euro area.” L. Papademos, Vienna,November 2003.

“The point [in time] of joining ERM II… should be consistent with having achieved a sufficientlevel of nominal and real convergence… An important ingredient of any monetary policy strategyafter EU accession is to guide the choice of when to enter ERM II and later EMU. Membershipof EU does not necessarily mean immediate entry in ERM II, although this may be an option forsome acceding countries… For the participation in ERM II to be successful, however, it iscrucial that the real and nominal convergence processes have advanced sufficiently and thateconomic policies and structures are consistent with this regime. It follows that prematurerigidity of the exchange rate could precipitate disorderly realignments with disruptiveeconomic consequences, including the credibility of the mechanism.” G. Tumpel-Gugerell,Frankfurt am Main, November 2003.

“Participation in ERM II may contribute to anchor expectations and support theimplementation of sound macroeconomic and structural policies, thus fostering real andnominal convergence. However, participation in ERM II must be compatible with all otherelements of a country’s macroeconomic policy framework, in particular with the monetary,fiscal and structural policies.” G. Tumpel-Gugerell, Frankfurt am Main, October 2003.

“The principal function of the ERM is to act as an instrument for consolidating economicpolicies designed to promote stability and convergence, both nominal and real. Membership ofERM II is itself a criterion for membership of Monetary Union, and should last for at least twoyears. There is also a qualitative requirement: this minimum period of membership must becompleted without excessive tensions arising and without the currency concerned beingdevalued [during this period] by the participating country. ERM II is often said to play a dualrole since it acts as an instrument of convergence and as a membership criterion in its ownright… ERM II must not be seen as an uncomfortable two-year waiting room for accessioncountries, but instead as a mechanism that combines compromise and flexibility and thatfacilitates stability and adaptation to a new environment.” E. Domingo Solans, Berlin,October 2003.

“First of all, ERM II is a voluntary arrangement, which means it can be joined at any timefollowing accession … Second, ERM II is an exchange rate arrangement with fixed butadjustable central parities and a ‘normal’ fluctuation band of ±15 % … which can help anchorthe nominal exchange rate and, crucially, expectations. At the same time, it also leaves someroom to accommodate upward pressure on the exchange rate... These phenomena have to betaken into account when deciding the modalities of ERM participation and … in the assessmentof exchange rate stability … Third, ERM II is a multilateral exchange rate arrangement.Decisions concerning central rates and fluctuation bands will be taken on a case-by-case basisand by mutual agreement of all the parties involved, including the ECB … If inconsistent nationalpolicies threaten to make the central parity unsustainable, the parties involved will first urge


the country concerned to mend its ways, to adjust its policies. If this is not feasible, arealignment can be negotiated.” L. Papademos, Dublin, March 2003.

“It is important that any decision to join ERM II is consistent with an adequate level of nominaland real convergence with the euro area … Once in ERM II, countries will be expected tocontinue their convergence process until the sustainable achievement of the Maastricht criteria… For some countries the benefits of staying longer in ERM II could more than offset theopportunity costs …Optimally choosing the timing of adoption of the euro also implies reducingthe differences in per capita income levels.” O. Issing, Budapest, February 2003.

“Upon accession to the EU, countries are expected to intensify preparations for fullparticipation in EMU. This process will include, at some point, joining the exchange ratemechanism (ERM II) … In this context, it was discussed in particular how to use ERM II notsimply as a ‘waiting room’ before joining the euro, but as a meaningful framework to deal withthe challenges related to further nominal and real convergence. Looking ahead, manyaccession countries expressed an interest in joining ERM II upon or soon after EU entry. At thesame time, it was also stressed that the optimal time to join the mechanism and eventually adoptthe euro could vary from country to country.” Press release, Genval Seminar on the EUaccession process, December 2002.

“ERM II will indeed be a framework that provides both stability and flexibility – a combinationthat is likely to be beneficial for many of the current accession countries in their real andnominal convergence process… ERM II should not be seen as a mere waiting room prior to theadoption of the euro… Instead, it should be regarded as a meaningful and flexible framework forcombining nominal and real convergence, and for tackling the challenges faced by theaccession countries in the run-up to the adoption of the euro… In addition, ERM II may providean appropriate framework to avoid major misalignments when choosing the conversion ratesto the euro.” T. Padoa-Schioppa, Vienna, November 2002.

“ERM II membership needs neither to happen immediately after EU accession in all cases, norto be limited to only two years, which is the minimum for adoption of the euro. It would bemisleading to consider ERM II as a mere waiting room before the euro. On the very contrary,ERM II would allow countries to retain some limited exchange rate flexibility during thecatching-up process.” J.-C. Trichet, Vienna, June 2002.

“ERM II membership does not need to happen immediately after EU accession in all cases, nordoes ERM II membership need to be limited to only two years, which is the minimum foradoption of the euro. A longer membership of ERM II may, in some cases, be helpful since itwould allow countries to retain the exchange rate as an instrumental policy variable during thecatching-up process.” W. Duisenberg, Frankfurt, November 2001.

13ECB


1 The roleof ERM II

1.2 RATIONALE AND FEATURES OF ERM II

To the extent that it is part of the Treatyobligations to adopt the euro, ERM IImembership also imposes a legal requirementon Member States with a derogation. Theexchange rate mechanism itself is defined in the“Resolution of the European Council on theestablishment of an exchange rate mechanism inthe third stage of the EMU” (hereafter, theResolution).6 The Resolution states:

– The main objective of ERM II is to support astable economic environment needed for thegood functioning of the single market. Inparticular, real exchange rate misalignmentsand excessive nominal exchange ratefluctuations between the euro and other EUcurrencies must be avoided. ERM II isexpected to provide Member States outsidethe euro area with a reference for theirconduct of sound economic policies,particularly in the monetary field, designedto foster real convergence and support theirefforts to adopt the euro, and to help protectparticipants from unwarranted pressures inforeign exchange markets.

– The main features of ERM II include: (i) acentral rate against the euro; (ii) a standardfluctuation band of ±15% around the centralrate; (iii) obligatory interventions at themargins, which are in principle automaticand unlimited; and (iv) availability for veryshort-term financing. The Resolutionemphasises, however, that the ECB and theparticipating NCBs “could suspendintervention, if this were to conflict withtheir primary objective.” Such a decision tosuspend intervention would “take dueaccount of all relevant factors and inparticular of the need to maintain pricestability and the credible functioning of theexchange rate mechanism.”

– Participation in the exchange rate mechanismwill be voluntary for the Member Statesoutside the euro area. Nevertheless, MemberStates with a derogation can be expected to

join the mechanism, while a Member Statewhich does not participate from the outset inthe exchange rate mechanism may do so at alater date. The Resolution makes no mentionof entry conditions and, although a commonaccord on the central parity and fluctuationbands needs to be reached, provides nogrounds for a refusal of the application.However, “all parties to the mutualagreement, including the ECB, have the rightto initiate a confidential procedure aimed atreconsidering central rates”.

As follows from the above, ERM II is meant tohelp participating Member States orient theirpolicies to stability and to foster convergence.Furthermore, participation in ERM II plays arole in the convergence criteria for joining theeuro, because participation in it for a period ofat least two years without severe tensions andwithout devaluing at the country’s owninitiative is one of the preconditions for joiningthe euro. This dual role of ERM II will haveimplications for the exchange rate policies ofacceding countries upon ERM II entry.

Whether these two purposes can be achievedsimultaneously or whether they imply asequence, with a period in which thestabilisation role of ERM II is prevalent,followed by a convergence test period, willdepend on country-specific characteristics andwill, therefore, vary among acceding countries.Section 1.3 explores the risks associated withjoining ERM II right upon EU accession, whilePart 2 deals with issues that are relevant for thetiming of euro adoption, thereby delineatingpossible endpoints of ERM II participation.Indeed, if the timing of ERM II is not carefullyassessed, the management of exchange ratepolicy may be complicated and even subject tosignificant risks. In particular, in someacceding countries, nominal and real interestrates are still higher than in the euro area. Incombination with tightly managed exchange

6 Resolution of the European Council on the establishment of an exchangerate mechanism in the third stage of Economic and Monetary Union,Amsterdam 16 June 1997.


rates, this could trigger portfolio inflows linkedto so-called convergence trades in financialmarkets, particularly in the context of anERM II participation with very small exchangerate movements which limit the potential lossesof speculative attacks. On the other hand,adjusting interest rates too early to euro arealevels may run counter to disinflation progressand domestic stabilisation. Containingexchange rate appreciation in a context ofpersisting capital inflows through interventionswould, if not sterilised, risk the prospect ofdisinflation or, if sterilised, be fairly ineffectiveand entail potentially significant sterilisationcosts. Finally, the substantial fiscal adjustmentsenvisaged in some of the larger accedingcountries also pose severe risks to themanagement of monetary and exchange ratepolicies. In particular, as privatisation receiptsare envisaged to decline, the authorities mayincreasingly need to resort to debt issuance tofinance fiscal deficits, adding further pressureon capital inflows and complicating themanagement of monetary policy.7

As for the convergence test regarding thefulfilment of the exchange rate criterion in thetwo years before the convergence assessment, akey reference for exchange rate management isthe Informal Ecofin document on “Accedingcountries and the ERM II” adopted on 5 April2003. In line with the ECB’s positionmentioned above, this document states: “Theassessment of the fulfilment of the Maastrichtconvergence criteria and the procedures to befollowed for the introduction of the euro willensure equal treatment between future MemberStates and the current participants in the euroarea. A minimum stay of two years in themechanism prior to the convergence assessmentwithout severe tensions is expected. Moreover,the assessment of exchange rate stabilityagainst the euro will focus on the exchange ratebeing close to the central rate, while also takinginto account factors that may have led to anappreciation, in line with what was done in thepast” (emphasis added). This would focus onthe implementation of sound monetary andfiscal policies, which nevertheless may lead to

an appreciation of the exchange rate that isintrinsic to the catching-up process. Suchdevelopments would also mirror what has beendone in the past in the cases of Ireland andGreece.

1.3 ECONOMIC ASPECTS OF ERM II

As it is a legal text, the Resolution does notassess the economic rationale for ERM IImembership. However, in order to qualifyfurther the acceding countries’ participation inERM II, the economic implications of ERM IImembership need to be reviewed. This sectionaims at reviewing some of the main economicconsiderations that arise in this context.

1.3.1 CURRENT EXCHANGE RATE REGIMES

Several of the acceding countries have alreadyessentially renounced an independent monetarypolicy, since they have traditionally maintainedexchange rate strategies based on externalanchors, mostly through hard pegs (see Table 1and also Chart 1 on pp. 16-17):

– The Cyprus pound has been unilaterallypegged, with a ±2.25% fluctuation band, tothe ECU since June 1992 and to the eurosince 1 January 1999, with the same centralparity as previously adopted for the ECU.On 1 January 2001 wider bands of ±15%were introduced, in order to absorb anyshocks from potentially destabilising capitalmovements and to deter speculative capitalflows. At the same time, the narrower“softer” bands of ±2.25% were temporarilymaintained to help anchor prices andexpectations. The narrower bands weredefinitively abolished on 13 August 2001,leaving only the ±15% margins currently inplace.

7 Alternatively, if the acceding countries were to resort to issuing short-termdebt denominated in domestic currency, which is traditionally absorbed byplayers in the domestic market, this might have a crowding-out effect onprivate investment.

15ECB


1 The roleof ERM II

– Estonia has followed a currency boardarrangement since 1992. Originally pegged tothe Deutsche Mark, the Estonian kroon was re-pegged to the euro on 1 January 1999.

– Since February 1994, Latvia has followed ade facto peg to the Special Drawing Right(SDR), which was officially formalised in1997. The weight of the euro in the SDRcurrently stands at around 35%.

– Lithuania has followed a currency boardarrangement since April 1994. Althoughoriginally pegged to the US dollar, thecurrency board of the litas was re-pegged tothe euro in February 2002.

– Malta has followed a currency basket pegsince 1971. The weight of the euro in thecurrency basket has been regularly revised toreflect the trade relations of the Malteseeconomy, and currently stands at 70%.

– Since 1992, the Slovenian tolar has, for mostof this period, been on a moderate andsmooth depreciation trend against theDeutsche Mark and, since 1 January 1999,against the euro.

Exchange rate strategy 1) Currency Features

Cyprus Peg to the euro, with Cyprus pound The Cyprus pound has de facto fluctuated within a narrow range± 15% fluctuation bands

Czech Republic Managed float Czech koruna Inflation targeting: 2%-4% by end-2005

Estonia Currency board to the euro Estonian kroon Introduced in 1992

Hungary Peg to the euro, with Hungarian forint Exchange rate regime combined with inflation targeting: max. 4.5%±15% fluctuation bands by end-2003, 5.5% by end-2004 and 3%-5% by end-2005

Latvia Peg to the SDR Latvian lat Fluctuation band ±1%(euro weight 35%)

Lithuania Currency board to the euro Lithuanian litas Introduced in 1994; re-pegged from the US dollar to the euro inFebruary 2002

Malta Peg to a basket Maltese lira Currency basket (euro (70%), US dollar, pound sterling)

Poland Free float Polish zloty Inflation targeting: 2%-4% by end-2003; 1.5%-3.5% from 2004 onwards

Slovakia Managed float Slovak koruna Hybrid strategy, combined with implicit inflation targeting

Slovenia Exchange rates within Slovenian tolar Two-pillar strategy monitoring monetary, real, external and financialcrawling bands 2) indicators

Sources: IMF and ECB.1) Based on the IMF De Facto Exchange Rate Arrangements and Anchors of Monetary Policy as of 30 June 2003.2) The regime operating de facto in the country differs from its de jure regime, which is a managed float.

Table 1 Exchange rate strategies currently fol lowed by the acceding countries

(IMF classification)

Overall, this strategy has served most small andvery small acceding countries well, as it hassupported them in significantly advancingeconomic convergence in recent years. In mostcases, inflation and nominal interest rates standat around levels similar to those in the euro area,while fiscal accounts are presently in balance ordisplay relatively moderate deficits.8 However,several small acceding countries are currentlyregistering high current account deficits, whichwarrant continuous monitoring.

In the larger acceding countries, however, theexchange rate seems to play a macroeconomicrole as a stabilisation tool, dampening to someextent output variability. Therefore, in thesecountries, maintaining some degree of monetarypolicy independence and nominal exchange rateflexibility may still be useful at the current stageand possibly for some time to come. Indeed,this policy option would be consistent with thecountries’ own monetary policy choices atpresent, as the larger acceding countries have

8 The notable exceptions are Malta, with a projected budget deficit of 6.5%of GDP in 2003; Cyprus, with a projected budget deficit of 6.0% in 2003;and Slovenia, where HICP inflation stood at 5.6% in 2003. See Part B forfurther details, as well as on the challenges in the fiscal field.


Chart 1 Units of national currencies per euro including ±15% bands from average value

(inverted scales)

Source: ECB.

Cyprus

0.48

0.53

0.58

0.63

0.68

0.48

0.53

0.58

0.63

0.68Jan.1999 2000 2001 2002 2003

Jan.2004

Czech Republic

Jan.1999 2000 2001 2002 2003

Jan.2004

28

30

32

34

36

38

40

28

30

32

34

36

38

40

Estonia

Jan.1999 2000 2001 2002 2003

Jan.2004

13

14

15

16

17

18

13

14

15

16

17

18

Hungary

Jan.1999 2000 2001 2002 2003

Jan.2004

212

222

232

242

252

262

272

282

292

212

222

232

242

252

262

272

282

292

Latvia

Jan.1999 2000 2001 2002 2003

Jan.2004

0.5

0.6

0.7

0.5

0.6

0.7

Lithuania

Jan.1999 2000 2001 2002 2003

Jan.2004

3.0

3.5

4.0

4.5

3.0

3.5

4.0

4.5

17ECB


1 The roleof ERM II

felt the need to grant their currencies increasedflexibility in recent years.

– The Czech Republic adopted its currentstrategy, a free float combined with aninflation targeting framework, in 1998. Thisstrategy was introduced with the aim ofproviding a nominal anchor to monetarypolicy, after the severe currency crisis of1997 triggered the abandonment of thekoruna peg.9

– The Hungarian forint has moved away fromthe crawling fluctuation band introduced inMarch 1995, with the rate of the crawl beingreduced gradually and the band widened. InMay 2001, the crawl was abandoned and the

bands widened to ±15%. The choice of thisstrategy was quite timely, as it coincidedwith the liberalisation of short-term capitalflows. Moreover, it has been combined withan inflation-targeting framework. Morerecently, the central rate was devalued by2.26% in June 2003.

– Poland follows a free float with an inflation-targeting framework. Starting from acurrency basket peg introduced in 1991,Poland has gradually allowed increasingnominal exchange rate flexibility through a

Source: ECB.

Slovakia

Jan.1999 2000 2001 2002 2003

Jan.2004

36

38

40

42

44

46

48

50

36

38

40

42

44

46

48

50

Slovenia

Jan.1999 2000 2001 2002 2003

Jan.2004

180

200

220

240

180

200

220

240

Chart 1 (cont’)

(inverted scales)

Malta

Jan.1999 2000 2001 2002 2003

Jan.2004

0.34

0.38

0.42

0.46

0.34

0.38

0.42

0.46

Poland

Jan.1999 2000 2001 2002 2003

Jan.2004

3.3

3.5

3.7

3.9

4.1

4.3

4.5

4.7

3.3

3.5

3.7

3.9

4.1

4.3

4.5

4.7

9 The koruna was fixed against a currency basket composed of DEM (65%)and USD (35%), with a fluctuation band of ±0.5% from 1993 to February1996 and of ±7.5% from February 1996 to May 1997.


crawling band, introduced in 1995, with therate of the crawl being gradually reduced andthe band widened so as to finally let thecurrency float in 2000.

– Since the abandonment of the koruna peg in1998, Slovakia has followed a managed floatwith the euro as the main referencecurrency.10 This exchange rate frameworkhas been combined with an implicit inflation-targeting strategy.

These policy shifts towards greater exchangerate flexibility have often been triggered bystabilisation constraints involved in themanagement of more rigid regimes.11 They havealso been facilitated by the rapid growth offoreign exchange and local securities marketsresulting from capital account liberalisation,high FDI inflows and growing domesticbanking sectors.

1.3.2 REGIME CHANGES IN VIEW OF ERM IIMEMBERSHIP

For most of the small and very small accedingcountries, ERM II membership may notrepresent a de facto exchange rate regime shift.Indeed, euro-based CBAs and hard pegs arelikely to be maintained within ERM II. First,these countries lack a clear exit strategy fromtheir current regimes, other than the eventualadoption of the euro, and second, countriespursuing these strategies can, in principle, aim ata ERM II central parity equal to the one in theircurrent exchange rate regime, plus a very narrowfluctuation band unilaterally declared. Clearly,the decision on the exchange rate parity will haveto be made in the multilateral setting foreseen inthe ERM II Resolution, and there will be a case-by-case assessment of the compatibility of theeuro-based CBAs with ERM II and, later on,with the adoption of the euro.

– The Cyprus pound has continued to trademostly within the range of the previouslyexisting narrow band vis-à-vis the euroduring the last two years, without making

effective use of the official ±15% fluctuationband introduced in 2001.

– Conditional upon a positive assessment ofthe sustainability of the currency boardarrangement and on a multilateral agreementon the central rate as well, Estonia couldmaintain its euro-based currency board as aunilateral ex ante commitment to exchangerate stability, augmenting the exchange ratediscipline embedded in ERM II.

– Latvia also has a tradition of nominalexchange rate stability in the last decade.Still, before joining ERM II, Latvia wouldneed to revise its current currency basket,where the euro has, at current exchangerates, a weight of only around 35%, andmake the euro the sole reference currency.12

– Conditional upon a positive assessment ofthe sustainability of the currency boardarrangement and on a multilateral agreementon the central rate as well, Lithuania couldmaintain its euro-based currency board as aunilateral ex ante commitment to exchangerate stability, augmenting the exchange ratediscipline embedded in ERM II.

– Malta, which already operates a peg to a basketwith no fluctuation band, would need to reviseits current currency basket before joining ERM IIto make the euro the sole anchor currency(currently the euro’s weight is at 70%).

– Given that the tolar has been on a gentledepreciation path versus the euro, Sloveniawould have to shift from trend depreciationto horizontal bands around a fixed parity inorder to join ERM II.

10 Between 1993 and 1998, the Slovak koruna was pegged to a currencybasket with one devaluation of 10% in mid-1993. The basket comprisedfive currencies until mid-1994 and subsequently two currencies (60%DEM and 40% USD). In 1996 the fluctuation band was widened from±1.5% to ±7%. On 2 October 1998, the peg was abolished and replacedby a managed float.

11 For instance, the currency crises in the Czech Republic and Slovakia in1997 and 1998 respectively triggered the abandonment of the pegs.

12 The amount of euro in the SDR basket is EUR 0.426. Since the values ofthe amounts of each currency in the basket fluctuate along with theirexchange rate, the share of the euro fluctuates as well.

19ECB


1 The roleof ERM II

Maintaining the main features of the above-mentioned regimes until the adoption of theeuro may be justified on economic grounds.First, to the extent that the current exchange ratestrategy is generally credible and compatiblewith ERM II, it may not be desirable to advocatea double regime shift before the adoption of theeuro, as the latter may raise uncertaintyregarding the country’s medium-term monetarystrategy, thereby distorting economic decisionsand risking those stabilisation gains alreadyaccomplished. Second, the adoption of strongexternal nominal anchors has served well thesevery small open economies, with theirextremely thin foreign exchange markets,enabling them to achieve macroeconomicstabilisation, low inflation,13 and increasingtheir access to global financial markets. Theachievements in terms of nominal convergencehave not come at the expense of output growth,as structural reforms and macroeconomicpolicies have been set broadly in line with themaintenance of relatively rigid exchange rateregimes. Although some countries havewitnessed some banking sector problems andasset price bubbles in the past, associated withsizeable current account deficits, these deficitshave typically been coupled with economicrestructuring and have been largely covered byFDI flows. Finally, the existing regimes haveapparently enjoyed, by and large, credibility, asreflected in stable and declining long-terminterest rates on assets denominated in domesticcurrency and in the lack of an active use ofinterest rate policy or pressure-relatedinterventions to defend the exchange rate (seeChart 2).14

Unlike in the case of small and very smallacceding countries, which have adopted anexternal nominal anchor and have, thus,largely or entirely given up an independentmonetary policy, larger acceding countrieshave maintained some monetary policyindependence, and the exchange rate has beenused increasingly flexibly in recent years.Consequently, ERM II entry for these countrieswould imply a regime change, in particular forthose countries which have not yet taken on

explicit exchange rate commitments. Experienceshows that it is of crucial importance to designand manage regime changes in order to ensure asmooth course of developments and avoidwelfare losses. This relates, in particular, tochoosing the right timing for moving to a newregime, which requires special caution.

In principle, ERM II could well accommodatethe need for exchange rate flexibility in thelarger acceding countries. Provided that ERM IIentry is well-timed and that the entry parity ischosen in line with underlying fundamentals,the mechanism, with its ±15% fluctuation band,leaves significant room for adjustments toshocks and market developments. Thepossibility of changing the central parity undera multilateral agreement would add furtherroom for manoeuvre, for example, if thereal catching-up process is significantly fasterthan expected and leads to sustained upwardpressure on the exchange rate. Hence, a fulluse of the wide bands in ERM II and thecommunication of the possibility ofrealignments would, in principle, allow for aconsiderable margin of de facto exchange rateflexibility. An overly tight exchange ratemanagement “close to the central rates” ascurrently envisaged by some of these countrieswould, in all likelihood, be insufficient for theexchange rate to play a useful role as anadjustment tool.

The timing of ERM II entry also has to beconsidered carefully with a view to the fiscalposition of a given country. This againspecifically relates to the larger accedingcountries, which are currently registeringsubstantial fiscal imbalances (see also Part 2). Itis obvious that ERM II participation would besubstantially facilitated by a sound medium-term fiscal strategy, while at the same time alsounderpinning the credibility of the fiscalframework.

13 The exception is Slovenia in this group, where HICP inflation stood at5.6% in 2003.

14 It should be noted, however, that in the absence of market liquidity, itwould take a serious misalignment for interest rates to be indicative ofsuch pressures.


Chart 2 Interest rates and reserves

Cyprus

0

10

20

30

40

0

10

20

30

40

Jan.1999

Jan.20042000 2001 2002 2003

Long-term interest rate1)

Short-term interest rate2)

Reserves (% of GDP)3)

Czech Republic

0

10

20

30

40

0

10

20

30

40

Jan.1999 2000 2001 2002 2003

Jan.2004

Estonia

0

10

20

30

40

0

10

20

30

40

Jan.1999 2000 2001 2002 2003

Jan.2004

Hungary

0

10

20

30

40

0

10

20

30

40

Jan.1999 2000 2001 2002 2003

Jan.2004

Latvia

0

10

20

30

40

0

10

20

30

40

Jan.1999 2000 2001 2002 2003

Jan.2004

Jan.1999 2000 2001 2002 2003

Jan.2004

Lithuania

0

10

20

30

40

0

10

20

30

40

Sources: ECB, Eurostat, IMF and national sources.1) 5-year government bond yield (data series incomplete). 10-year government bond yield for Slovenia, 10-year kroon bankloans to households for Estonia.2) 3-month money market deposit rate.3) Total international reserves of the central bank (excl. gold).

21ECB


1 The roleof ERM II

A closer assessment of individual country casessuggests that early ERM II participation mayentail potential drawbacks for countries with afree or lightly managed float and a functioningand credible inflation targeting framework inplace. This is particularly true in cases wheresuch a policy set-up has been associated withsubstantial exchange rate volatility in the past.In principle, early entry into ERM II could beseen as less problematic for countries that havefollowed a unilateral “ERM II shadowing”strategy or witnessed more limited nominalexchange rate variability in recent years,provided that ERM II participation is consistentwith the overall policy stance. In this respect, itis noteworthy that fiscal consolidation appearsto be seriously off track in some of thesecountries (see Section 2.3.8).

Finally, a general caveat has to be made. Evenif, at the current juncture, there may be no prima

facie evidence that the current exchange rateregimes in some acceding countries have beeninappropriate from a stabilisation perspective orhave been subject to significant marketpressures, it would be premature to concludethat the countries are ready for membership ofERM II. Even if the current regimes may beregarded as appropriate, the choice of thecentral parity is an issue that needs carefulassessment, as decisions concerning centralrates are taken by mutual agreement of theparticipating members in ERM II, including theECB. Moreover, even if only minor de factochanges seem required in some small accedingcountries for participation in ERM II, apotential misalignment in these countries is arisk that cannot be ignored. More generally, theabsence of significant foreign exchange marketpressure in the past cannot be taken as implyingan absence of such pressure in the future. Riskscould, in principle, be relevant for all countries,

Jan.1999 2000 2001 2002 2003

Jan.2004

Slovakia

0

10

20

30

40

0

10

20

30

40

Chart 2 (cont’)

Jan.1999 2000 2001 2002 2003

Jan.2004

Malta

0

10

20

30

40

50

60

0

10

20

30

40

50

60

Jan.1999 2000 2001 2002 2003

Jan.2004

Poland

0

10

20

30

40

0

10

20

30

40

Slovenia

0

10

20

30

40

0

10

20

30

40

Jan.1999 2000 2001 2002 2003

Jan.2004


for example if EU entry were to prove a majorasymmetric shock to the economies of theacceding countries, or if individual country-specific developments were to derail the overallmonetary integration process. In such ascenario, the fact that financial markets areparticularly small in some acceding countriescould become a disadvantage, as the criticalmass needed to trigger policy changes is alsorelatively limited. Hence, if a majormisalignment threatening the stability of ERMII occurs, all participating members, includingthe ECB, have the right to initiate a procedureaimed at reconsidering central rates.

Against this background, the requirement for atwo-year participation in ERM II, which acts asa testing phase for the central rate as well as thesustainability of convergence in general, wouldhelp in identifying potential misalignments inthe central parity that have not been detectedbefore. Two main issues may have to beconsidered in this respect: first, whethernominal exchange rate stability itself hasaggravated stabilisation problems, reflected, forexample, in excessive current account deficitsor overheating. In this case, a change in thecentral parity could be warranted. Second,whether the growing size of the accedingcountries’ economies and their financialmarkets, in combination with structuraldifferences compared to the anchor country,may warrant greater exchange rate flexibility.Although at the current juncture neither severestabilisation problems nor strong reasons forgreater exchange rate flexibility may seempressing, these issues, which are explored infurther detail in Part 2 below, should not beruled out entirely.

23ECB


2 Challenges ofeuro adoption

2.1 ACCEDING COUNTRIES’ PERSPECTIVE

Almost all acceding countries initially indicatedtheir intention to join the euro area as soon aspossible. More recently, however, someacceding countries have begun to envisage asomewhat longer process before adopting theeuro. Box 3 presents a short overview of thetimeframes that are currently featuring in theacceding countries’ monetary integrationstrategies with respect to euro adoption.15 Froma technical point of view, the earliest possibledate for an acceding country to adopt the eurowould be on 1 January 2007, assuming that euroarea membership continues to occur at thebeginning of a calendar year.16 In such ascenario, the country would join the ERM IImechanism soon after EU accession in 2004.After spending a minimum period of two yearsin ERM II, the country’s convergence with theeuro area could be assessed in mid-2006 at theearliest. After a positive examination andfollowing some preparatory work, the countrywould enter the euro area in January 2007.

At a country level, the target dates range from2007 (with an occasional reference to the laterpart of 2006) to around 2010. In severalcountries the central banks initially seemed tofavour an earlier date than the government,mainly reflecting the difficult fiscal situation incountries concerned, but more recently suchdifferences in views about the optimal timingappear to be abating. This is inter alia due to thefact that in most acceding countries, thegovernment and the central bank haveformulated joint strategies on prospectivemonetary integration, or talks to this end havereached an advanced stage. The EconomicDialogue process between the EU and theacceding countries, in particular the Pre-accession Economic Programme (PEP)exercise, has apparently also helped thediscussion process in individual countries onmonetary integration matters. Most recently, forexample, such joint strategy documents havebeen released by Slovakia (June 2003),Hungary (July 2003), the Czech Republic(October 2003) and Slovenia (November 2003).

2 CHA L L ENGE S O F E URO ADOP T I ONThe stretching-out of euro adoption plans insome acceding countries clearly has to do withdifficulties caused by fiscal situations andconcerns that rapid consolidation may entailmajor short-term output losses or may lacksufficient political backing on other grounds,for example owing to perceived opposition tocuts in transfers or subsidies. Some accedingcountry central bankers have argued, however,that the sacrifice ratio may not depend on thespeed of fiscal consolidation, that rapidconsolidation may be less prone to reversals,and that an early meeting of the convergencecriteria would allow the benefits of monetaryunion to be reaped sooner rather than later. Thefiscal issues and challenges that the accedingcountries are facing will be discussed furtherin Section 2.3.8. At this point, it is worth notingthat in the cases of the Czech Republic, Polandand Malta, the fiscal trajectories enshrined inthe 2003 PEPs do not foresee that the budgetdeficit criterion will be met throughout thewhole programming period, which extendsfrom 2003 to 2006. In case the fiscal strategiesof these three countries materialise along thelines laid out in the 2003 PEPs, the budgetdeficit criterion would not be fulfilled beforethe year 2007 at the earliest. This in turn wouldimply that a positive convergence assessmentwould not be possible before 2008, and that theeuro could not be introduced before 2009 at theearliest, assuming that euro area membershipcontinues to take place at the beginning of acalendar year.

15 As the 2004 enlargement approaches, the debate about the optimal timingof adopting the euro is intensifying. In fact, this debate actually started acouple of years ago, when some acceding countries considered adoptingthe euro unilaterally as legal tender, i.e. before joining the EU. Followingstrong opposition by the ECB among others, the acceding countriesrefrained from taking such a step, but at the same time underlined theirintention to enter the euro area as soon as possible, while remaining fullyin line with the Treaty.

16 If this assumption were to be relaxed, an adoption of the euro as early asthe later part of 2006 could be technically feasible.


Box 3

STATEMENTS BY ACCEDING COUNTRIES ON THE TIMING OF EURO ADOPTION

Cyprus “The target date for adopting the euro remains the year 2007...notwithstanding the increased difficulties that have risen because of therecent fiscal slippages.” PEP 2003.

Czech Republic “The Czech Republic should join the euro zone as soon as plausibleeconomic conditions have been created.” PEP 2003. Most statementsthat address the timing issue more concretely point to the end of thedecade as a realistic time horizon for euro adoption.

Estonia “Estonia can obtain full European Union membership as soon as 2006.”PEP 2003.

Hungary “In order to realise the benefits of the adoption of the common currency,the Government decided to introduce the euro on 1 January 2008.” PEP2003.

Latvia “The earliest period when Latvia could really join the EMU is 1st January2008.” PEP 2003.

Lithuania “Realistically, the euro would probably replace the litas from the start of2007…” Governor Sarkinas, March 2003.

Malta “We can imagine we are ready to introduce the euro in the second half of2007 or in January 2008 at the latest.” Governor Bonello, February2004.

Poland “Poland should join the Economic and Monetary Union (EMU) so soonas possible only if the macroeconomic conditions make it possible ...Taking into account the condition of Poland’s economy and theGovernment’s projections of the general government deficit for theoncoming years, it should be expected that Poland will fulfil the nominalconvergence criteria stated in the Maastricht Treaty in 2007, so as tobecome a member of the Economic and Monetary Union in 2008 or2009.” PEP 2003.

Slovakia “The earliest realistic target for admission would be 2008.” Strategy ofthe Slovak Republic for adoption of the Euro, June 2003.

Slovenia “Both the Bank and the Government support adoption of the euro at theearliest opportunity and judge that it will be possible at the beginning of2007.” Joint programme of the Slovenian Government and Bank ofSlovenia for ERM II entry and adoption of the euro, November 2003.

25ECB



While in the initial stages of the discussion, thebenefits of an early adoption of the euro wereseen, by virtually all acceding countries, as faroutweighing any associated costs, the picturehas in the meantime become more nuanced. Asfor the benefits, acceding countries stress thatearly EMU membership would strengtheneconomic policy discipline, accelerate structuralreforms and raise the economies’ growthpotential through higher investment followingthe elimination of exchange rate uncertainty.More specifically, some acceding countryauthorities and officials stress that early EMUmembership “should have positive impacts ondomestic economic policy [through requiring]balanced public budgets” (Czech NationalBank, 2003b) and “would mobilise candidatecountries to complete their structural reforms”(Balcerowicz, 2002). In particular, the centralbanks stress the advantage of the EU policycoordinating framework and the multilateralsurveillance procedure to impose fiscaldiscipline on their governments. As regards theelimination of nominal exchange rateuncertainty, it is argued that this would implybenefits arising from lower transaction costs,17

enhance trade and investment, and lower therisks of financial sector disturbances. At thesame time, “the Stability and Growth Pact,coupled with a decline in the risk premium, …would lead to a stabilisation of long-terminterest rates at a low level” (Czech NationalBank, 2003a).

The costs of an early euro adoption are ingeneral not seen as significant in thosecountries that have no or only relativelymoderate fiscal deficits. In the other countries,the mainly medium and long-term gains of euroadoption have increasingly been contrasted withthe short-term costs of fiscal consolidation. Asfor other potential costs, a number of centralbanks have argued that the cost of relinquishingindependent monetary policy and flexibleexchange rates is limited owing to the highdegree of trade and financial integration withthe euro area (Balcerowicz, 2002). Moreover, ithas been maintained that the ongoingharmonisation in economic structures and

business cycles will diminish the importance ofasymmetric shocks over time (Szapary, 2002).

Overall, by taking all these arguments together,an early euro adoption is expected to boostgrowth in acceding countries through higherinvestment and trade, and thereby to contributeto real convergence with the EU countries. Astudy by Magyar Nemzeti Bank estimates thateuro area membership will boost GDP growthin the long run by 0.6% to 0.9% per year,whereas in the short term the benefits and costsare seen to cancel each other out (MagyarNemzeti Bank, 2002).

A number of academics openly support theacceding countries’ strategies for an earlyadoption of the euro. They mainly refer to thefact that the capital markets of accedingcountries are small and liberalised, which couldmake them vulnerable to financial crises ortrigger excessive exchange rate volatility (seee.g. Buiter and Grafe, 2002; Coricelli, 2002;Eichengreen, 2003). The adoption of the eurocould also eliminate the interest rate premiumand reduce interest rates (Coricelli, 2002).18

Finally, some authors argue that an ambitioustimetable for the adoption of the euro wouldtrigger earlier reforms of public finances andwould thereby contribute to higher growth(Center for European Policy Studies, 2002).

2.2 A REVIEW OF THE PROGRESS MADE INNOMINAL CONVERGENCE

Achieving a high degree of sustainable nominalconvergence, as laid down in the convergencecriteria listed in the Treaty, is a necessarycondition for an EU Member State with aderogation to enter the euro area. The rationalebehind these criteria is to ensure a high degreeof sustainable convergence with the euro area,assessed on the basis of inflation developments,long-term interest rates, exchange rate stability

17 In Hungary, reduced transaction costs were estimated to increase thelevel of GDP with a one-off effect of 0.18 to 0.30 percentage points(Magyar Nemzeti Bank, 2002).

18 See also Mundell (2002) and Gros (2000).


and the fiscal position in terms of public deficitand debt. In addition, the Treaty mentions“other factors” that should be taken into accountwhen assessing the degree of convergence suchas “the results of the integration of markets,the situation and development of the balancesof payments on current account and anexamination of the developments of unit labourcosts and other price indices” (Art. 121.1). Onthe basis of the convergence criteria and theseother factors, the Commission and the ECB willexamine the degree of sustainable convergenceachieved by each Member State. Taking intoaccount these reports, the European Council,acting by qualified majority on arecommendation from the Commission andhaving sought the opinion of the EuropeanParliament, will decide whether or not toabrogate the derogation of the Member Statesconcerned.

This section reviews the current performance ofthe acceding countries against the backdrop ofnominal convergence. The purpose of thissection is to inspect where acceding countries

Table 2 Economic indicators in view of convergence cr iter ia

Sources: ECB, Eurostat, Pre-accession Economic Programmes 2003 and Bloomberg.Note: Reference values are calculated for current EU Member States and in line with the ECB Convergence Reports.1) Period average; CPI for Malta. The EU countries with the lowest inflation rate in 2002 were Belgium, Germany and the UK.2) Period average.3) Based on the IMF De Facto Exchange Rate Arrangements and Anchors of Monetary Policy as of 30 June 2003.4) Weighted by nominal GDP in 2002.5) 10-year kroon bank loans to households6) Bank of Slovenia perceives the interest rate on RS44 government bonds issued in November 2002 as the best approximation of long-term bond interest rates in Slovenia.7) For Slovenia, the regime operating de facto in the country is different from its de jure regime, which is a managed float.

HICP Fiscal balance Public debt Long-terminflation1) (% of GDP) (% of GDP) yields 2) Exchange rate

(2002) (2002) (2002) (2002) regime 3)

Cyprus 2.8 -3.5 59.7 5.1 5-y ERM II shadowCzech Republic 1.4 -6.7 26.9 4.8 10-y managed floatEstonia 3.6 1.3 5.8 7.4 10-y 5) CBA (euro)Hungary 5.2 -9.2 56.3 7.1 10-y ERM II shadowLatvia 2.0 -3.0 14.6 5.3 5-y peg to SDRLithuania 0.4 -1.7 22.7 5.2 5-y CBA (euro)Malta 2.2 -6.2 66.6 5.4 5-y peg to basketPoland 1.9 -3.8 41.8 7.3 10-y floatSlovakia 3.3 -7.2 44.3 7.0 5-y managed floatSlovenia 7.5 -2.4 27.8 6.7 10-y 6) crawling bands 7)

AC-10 4) 2.7 -5.1 39.9 6.6

Reference value 2.9 -3.0 60.0 6.9

currently stand in terms of the Maastrichtcriteria, and to highlight the main possiblechallenges that individual acceding countriesface in meeting the criteria in the future. Thissection does not intend to provide or “pre-empt”any convergence assessment in these terms, butrather to help in assessing whether the strategyof adopting the euro a few years after EU entryis indeed a realistic policy option from today’sperspective.

When looking at economic indicators in view ofthe convergence criteria, it appears that a fewacceding countries currently already complynumerically with several criteria (obviouslywith the exception of the two years of ERMparticipation). With the caveat in mind that datastill have to be fully harmonised with ESA 95standards (in particular in the fiscal area), ananalysis of the most recent data availablesuggests the following (see Table 2).19

19 In the following analysis, reference is made to weighted averages for theacceding countries as a group. Clearly, the convergence examinationswill be undertaken on a country-by-country basis. The averages arepresented for reasons of illustration.

27ECB



– In 2002, the weighted average of the HICPinflation rate in acceding countries stood at2.7%, thus even slightly below thecorresponding hypothetical reference valueof 2.9% for 2002.20 There are six accedingcountries with an inflation rate currentlybelow the reference value: Cyprus, the CzechRepublic, Latvia, Lithuania, Malta andPoland.

– As regards the fiscal criteria, deficits inacceding countries amounted on average to5.1% of GDP in 2002, well above thethreshold of 3% of GDP.21 The figuresvaried significantly across countries, withonly Lithuania and Slovenia having fiscaldeficits of below 3% of GDP and Estoniaeven a surplus, while the Czech Republic,Hungary, Malta, Poland and Slovakia hadfiscal deficits well above the threshold. Thedebt-to-GDP ratio stood on average ataround 40% of GDP. With the exception ofMalta, the ratio in all countries was below60% of GDP.

– Concerning interest rates, long-term interestrates in acceding countries stood on averageat 6.6% in 2002, compared with 6.9%according to the calculated reference value.22

In fact, interest rates were below thereference value in most acceding countries,but slightly higher in Estonia, Hungary,Poland and Slovakia. However, it should benoted that in 2002 10-year bond instrumentswere not available in all acceding countries.

Looking ahead, the budget deficit criterion maywell pose difficulties for some countries,especially for the Czech Republic, Hungary,Poland and Slovakia, as the fiscal situation hasdeteriorated in these countries, oftensignificantly, in recent years. The deficit is alsoclearly above the 3% threshold in Cyprus andMalta. Moreover, fiscal pressures may increasein other acceding countries that have smallerdeficits at this stage. Section 2.3.8 explores thisissue in greater detail, with a particular focus onthe challenges that have arisen in the context ofachieving sustained fiscal consolidation and

restraint, while at the same time highlighting thecurrent and prospective spending pressures theacceding countries are facing (e.g. in the area ofacquis implementation and EU membershipcontributions). In addition, the remainder ofthis paper examines some factors in greaterdepth, such as current account developments,several price and competitiveness indicators,and overheating issues, which may help inassessing the acceding countries’ medium-termorientation towards price stability and sustainedgrowth.

It should be noted that the reliability of data,especially fiscal data, remains a matter ofconcern. While it is generally difficult toprovide reliable fiscal data, problems inacceding countries are even more pronounced,for various reasons. In methodological terms,the full implementation of ESA 95 standards isstill to be completed in these countries, and anumber of difficult classification issues stillhave to be fully resolved over the next months.Furthermore, there may still be some remainingproblems with providing fully consolidatedaccounts that incorporate all budgetary activitiesand display all obligations appropriately.Furthermore, in some acceding countries, theassessment of the fiscal situation may also beblurred by considerable contingent liabilities. Fullcompliance with EU rules in the provision ofstatistics will help to reduce these problem spotsand is therefore of key importance.

Interestingly, a comparison between theacceding countries in 2002, five years beforetheir intended adoption of the euro, and thecurrent Member States five years before theyqualified for EMU, i.e. 1996 for Greece and1994 for the other euro area countries, shows

20 According to the ECB Convergence Reports, the reference value forthe inflation criterion is based on taking the annual unweighted averageof the rate of inflation in the three EU countries with the lowest inflationrates, and adding 1.5 percentage points.

21 Again, it should be noted that while these figures are based on the Pre-accession Economic Programmes prepared by the acceding countries,they are not yet fully adjusted to ESA 95 standards.

22 According to the ECB Convergence Reports, the reference value forthe interest rate criterion is based on taking the arithmetic average of thelong-term interest rates of the three countries with the lowest inflationrates, and adding 2 percentage points.


that acceding countries have come much closerto nominal convergence. HICP inflation in 2002was in most acceding countries far lower than inGreece, Portugal and Spain five years beforethey entered the euro area. As regards long-terminterest rates, Greece, Italy, Portugal and Spainhad, five years before they entered the euroarea, much higher interest rates than accedingcountries have today. Moreover, while thefiscal situation was rather mixed for both setsof countries, most acceding countries havelower debt ratios.

2.3 A BROADER ECONOMIC ASSESSMENT

The preceding analysis has been based on theMaastricht Treaty framework, which forms thebasis for future convergence assessments ofnew Member States, in line with what has beendone in the past, and based on the equaltreatment principle between prospective andcurrent participants in the euro area.

The approach taken in the remainder of thispaper is a broader one, in line with the basic aimto get an analytical handle on the accedingcountries’ monetary integration strategies and,in particular, to examine in depth whetherthe economic logic of these strategies is soundfrom today’s perspective. To avoid anymisunderstandings, it should be underlined thatthis broader approach is not at all directed atidentifying new criteria or “reinterpreting” theframework enshrined in the Treaty on theEuropean Union. Nor is it the purpose of thepaper to endorse or support, in any form, theannounced monetary integration strategies ofacceding countries. Rather, the basic aim is toprovide an informed and balanced review of theeconomic wisdom of these strategies, in orderto prepare a view of the Eurosystem on theunderlying issues and foster the dialogue withacceding countries’ central banks.

An obvious starting point for such a broaderanalysis is the optimum currency area (OCA)theory, which is the standard reference point interms of economic theory for many current

discussions about the acceding countries’prospective readiness to join the euro area.According to the OCA theory, countries can beconsidered as part of an optimum currency areaif they fulfil certain criteria, which determinethe symmetry of external shocks and thecapacity of a country to absorb shocks. Thesecriteria refer to the similarity of economicstructures, business cycle synchronisation, thedegree of trade and financial integration, theflexibility of goods prices and wages, as well asfactor mobility. The OCA theory suggests thatif these criteria are fulfilled, a country canabandon the exchange rate as an adjustmenttool.

Nevertheless, despite some important insights,the OCA theory has often been criticised forbeing inconclusive and offering only a limitedanalytical framework. For example, the theorydoes not provide definite answers to theeffective costs and benefits, their correlationand the optimal timing of monetary integration.This is also related to the endogeneity of theOCA criteria. In fact, by reducing transactioncosts and market imperfections, joining acurrency union could foster trade and financialintegration and increase goods and wageflexibility. Likewise, within a currency unioneconomic structures could adjust and businesscycles could become more synchronised acrosscountries. Empirically, there is some evidencethat supports the endogeneity of the OCAcriteria (Rose, 2000; Rose and Van Wincoop,2001; Melitz, 2001; Frankel and Rose, 2000),although the findings of this strand of theliterature have met with considerable criticism(see Nitsch, 2002; Honohan, 2001; Persson,2001; compare also Rose, 2001 for limitedapplicability in the case of EMU).23 Overall, theresults have therefore been interpreted andapplied with caution. This is particularly true

23 Moreover, one should not fully disregard Krugman’s (1993)specialisation hypothesis as a theoretically possible alternative outcome.According to Krugman, an increase in trade owing to the use of a commoncurrency does not necessarily mean a lower exposure to asymmetricshocks, but could result in countries becoming more specialised due toincreasing returns to scale. As a result, the sensitivity of countries toindustry-specific shocks could increase, resulting in less synchronisedbusiness cycles.

29ECB



with regard to the question as to whethermonetary union fosters structural reform. Atany rate, it takes time for any endogenouseffects to work their way through the economicsystem, and the dynamics of such a processare difficult to anticipate. Thus, the periodthat leads to the new equilibrium may well befairly lengthy, which implies that adjustmentmechanisms other than the exchange rate wouldbe particularly important in absorbing shocks.

Against this background, the following chapterdiscusses where acceding countries standcompared with the euro area with respect to themost important OCA criteria, and how thesecriteria are likely to develop over time owing toEU accession and the impact of furthermonetary integration. The criteria consideredhere are the similarity of economic structures,the degree of trade integration, the flexibility oflabour markets, and the degree of financialintegration. Furthermore, irrespective of thetraditional OCA criteria, other factors need tobe taken into account, as joining the euro areatoo early might entail substantial costs foracceding countries. Therefore, this chapteralso discusses the differences in economicdynamics in acceding countries compared withthe euro area (in terms of trend growth, outputvolatility and business cycle synchronisation),the risks for external competitiveness, theimpact of interest rates declining to the euroarea level, and the optimal pace of fiscalconsolidation.

2.3.1 ECONOMIC STRUCTURES

GDP income levels in acceding countries arestill well below those in the euro area, withGDP per capita standing at around 50% of theeuro area average in PPP terms, ranging from36% in Latvia to 75% in Slovenia (seeChart 3).24 Nevertheless, acceding countrieshave experienced strong economic expansion,with real GDP growing on average at 3.6%(weighted) over the period 1993 to 2002, whichis well above the average real growth rate ofaround 2% in the euro area. As most countries

started from a low income per capita level andexperienced a severe recession at the beginningof transition, the catching-up in income levelswith the euro area can be expected to take a longtime for most countries. Alongside this process,acceding countries will tend to experiencestructural differences in economic dynamics.These real convergence issues will be dealt within greater detail in Section 2.3.5.

Nevertheless, economic structures in accedingcountries have already been continuouslyadjusted towards those in the euro area. In fact,acceding countries have made substantialprogress in transition and have advanced bothin terms of institutional convergence and inbringing their economic structures broadly inline with those in the euro area. Althoughsimilar economic structures to the euro area arefar from being a sufficient condition to benefitfrom EMU membership, it could be interpretedas being favourable for the countries’ capacity

Chart 3 GDP per capita in PPP terms

(as a % of the euro area average)

Sources: Eurostat and European Commission.1) Weighted by nominal GDP in 2002.

0102030405060708090

0102030405060708090

1 Spain2 Portugal3 Greece4 Slovenia5 Cyprus

6 Czech Republic 7 Hungary 8 Malta 9 Slovakia10 Estonia

11 Lithuania 12 Poland 13 Latvia

19932002AC-10 average in 20021)

1 2 3 4 5 6 7 8 9 10 11 1312

24 Income levels are highest in Slovenia, Cyprus, the Czech Republic andHungary, and are close to those in Portugal and Greece, while per capitaGDP in most of the other acceding countries is about two-thirds this level.When they joined the euro area, Spain, Portugal and Greece registeredGDP per capita levels amounting respectively to 85%, 70% and 67% ofthe euro area average in PPP terms.


Chart 4 Institutional reform (2003)

(index: 100 = well-functioning market economy)

Sources: EBRD and ECB staff calculations.

01 2 3 4 5 6 7 8

20

40

60

80

100

0

20

40

60

80

100

1 Czech Republic2 Estonia3 Hungary

4 Slovakia5 Poland6 Lithuania

governance and enterprisesmarkets and competitionfinancial institutions

7 Latvia8 Slovenia

to absorb economic shocks and to face similarshocks.25 Moreover, EU accession is expectedto further foster convergence, as by that timeacceding countries will have to complete theadoption and, even more importantly, theimplementation of the acquis communautaire(apart from those areas where transitionarrangements apply and full harmonisation willtake place later). Furthermore, upon EUaccession, acceding countries will be includedin most of the procedures for multilateralsurveillance as laid down in the EU Treaty. Yet,stronger coordination of economic policies islikely to contribute to a further deepening ofeconomic integration with the current EUMember States.

The acceding countries’ progress in completingtransition was recently confirmed by theEuropean Commission, which judged each ofthe ten countries to have established afunctioning market economy.26 As regards theacceding countries’ capacity to withstandcompetitive pressure and market forces withinthe Union, this is already the case for Cyprusand Malta, while the other eight countries arejudged to be sufficiently capable uponaccession if they continue on their currentreform paths. Moreover, acceding countrieswere acknowledged to have reached a high levelof alignment with the acquis communautaire bycontinuously bringing institutions andregulatory frameworks into line with EUstandards. At the same time, implementationand enforcement still have to be furtherimproved in several areas by a number ofcountries. In this context, the Commission hasestablished a comprehensive monitoringmechanism and an action plan to ensurecompliance until accession.

This assessment is broadly in line with theEBRD transition indicators, which confirmnotable progress in the areas of privatisationand liberalisation of markets and prices,although further progress would be needed withregard to financial institutions in most centraland eastern European acceding countries.27

Some differences also exist across countries

with respect to reforms in enterpriseprivatisation and restructuring. In contrast,progress in the area of markets and competitionis fairly homogeneous across all eight countries(see Chart 4).28

Finally, the economic size of the three broadsectors and the distribution of employmentacross sectors have gradually convergedtowards those in the euro area, despitelarge differences across countries (see Table 3).This evidence has to be interpreted withcaution, however, since it does not take intoaccount differences that may exist at a moredisaggregated level. At the broad sectoral level,the current GDP shares of agriculture andindustry are still higher in acceding countriesthan in the euro area (3.5% and 31.6% comparedwith 2.0% and 27.0% in the EU), reflecting thestill ongoing restructuring process, while the

25 The OCA theory suggests that countries with similar economic structuresseem to be less affected by asymmetric shocks and respond to shocks ina similar way, so that business cycles are also likely to be moreharmonised across countries. Under such circumstances, independentinterest rate policy can be considered as being less relevant for outputstabilisation. See Corden (1972) and Tavlas (1994).

26 See the Regular Report 2002 for all acceding countries. RegularReports assess to which extent acceding countries already fulfil theCopenhagen criteria (i.e. EU membership criteria).

27 EBRD transition indicators do not exist for Cyprus and Malta.28 For further information on financial sector integration, see Part 2,

Section 2.3.4.

31ECB



services sector (64.9% of GDP) is somewhatsmaller than in the EU (71.0%).29 Yet, structuraldifferences with the EU are considerable whenlooking at the distribution of employmentacross sectors. In fact, the employment sharesin acceding countries are significantly larger inthe agriculture and industry sectors and lowerin the services sector compared with the EU.Acceding countries’ relative productivity inthe agriculture sector is clearly lower, andproductivity in the services sector higher,compared with economy-wide productivity thanin the EU, while relative productivity in theindustry sector seems comparable. Against thisbackground, further adjustments in economicstructures are conceivable during the catching-up process.

Moreover, countries display large differenceswith respect to sector shares and employmentdistribution. In particular, Latvia and Lithuaniahave a considerably larger agriculture sectorthan the EU, both in terms of GDP andemployment share. In Poland, the share ofemployees in the agriculture sector is more than20 percentage points higher than in the EU,while the economic size of the agriculture sectoris rather similar, pointing to a much larger

Economic size as a % of GDP Employment distribution as a % of total

Agricultural Industry and Services Agricultural Industry and Servicesconstruction construction

Cyprus 4.1 20.3 75.6 5.1 23.4 71.4Czech Republic 3.2 37.3 59.5 4.8 40.0 55.3Estonia 5.4 29.3 65.3 6.9 31.2 62.0Hungary 3.7 30.7 65.6 6.2 34.1 59.7Latvia 4.7 24.7 70.6 15.1 24.4 60.5Lithuania 7.1 30.5 62.4 17.4 27.4 55.2Malta 2.8 28.1 69.1 2.0 31.7 66.3Poland 3.1 30.3 66.5 26.3 26.2 47.5Slovakia 4.4 31.1 64.5 6.2 38.5 55.3Slovenia 3.3 36.0 60.7 11.0 37.0 52.0AC-10 3.5 31.6 64.9 15.8 31.2 53.0

Greece 7.0 22.3 70.8 15.3 24.2 60.4Portugal 3.5 28.0 68.5 12.0 34.0 54.0Spain 3.2 28.5 68.2 5.9 29.4 64.7EU 2.0 27.0 71.0 3.9 28.2 67.8

Table 3 Economic size and labour distribution of sectors

(2002)

Source: Eurostat.

productivity gap between the agriculture sectorand the economy as a whole in Poland than inthe EU.30 Moreover, in the Czech Republic andSlovenia the industry sectors are considerablylarger and the services sector smaller. Therelative share of employment in the servicessector is much lower in most accedingcountries, with the exception of Cyprus andMalta. Nevertheless, the above figures might beaffected by the fact that the shadow economystill plays a relatively larger role in mostacceding countries. In addition, a more refinedview of the similarity in economic structureswould require further analysis of moredisaggregated sectoral figures, considering alsothe differences in the quality of products.

2.3.2 TRADE INTEGRATION

Diversified export structures and strong tradelinks between countries, particularly when

29 The following analysis takes the EU as a reference point with a view todrawing an inclusive picture. An analysis based on figures for the currenteuro area would yield a very similar picture.

30 However, the demographic structure of the Polish agriculture sector,where the average age is close to 60, might also imply a gradual declinein the high employment share over time.


intra-industrial trade accounts for a high shareof trade, are expected to reduce the exposure toasymmetric shocks and to harmonise thecapacity to absorb shocks. As one of the mainelements of the OCA theory, strong trade linkssuggest that country-specific policies, and inparticular nominal exchange rate flexibility, areless needed.31 This view has been contested bysome authors, who argue that closer monetaryintegration would rather favour specialisationin production and would therefore reduce thesynchronisation of business cycles.32 However,it could be argued that specialisation is oftenrelated more to regions than countries.33

Moreover, several empirical studies haveindeed found a significant positive correlationbetween trade integration and shock symmetry,which is particularly strong with a higher shareof intra-industrial trade.34

With the partial exception of Poland, accedingcountries are small and highly open economies,with a degree of openness of around 80% ofGDP, compared with 58% for the weightedaverage of individual euro area countries ifintra-euro area trade is included (see Chart 5).The most open countries are Malta, Estonia,Slovakia, Hungary and the Czech Republic.

Acceding countries already have close traderelations with the EU, accounting for 67% oftotal exports and 60% of total imports (seeChart 6). Expressed in GDP terms, the accedingcountries’ EU trade accounts on average for52%. This compares well with the level of tradeintegration among the current EU members,whose exports and imports within the EU are onaverage around 60% of total trade. However,large differences exist across countries, alsocompared with the euro area, with severalacceding countries displaying somewhat lessintense trade relations with the EU, measured asa share of total trade, than most of the currentEU Member States. But if the process of tradeorientation of the acceding countries towardsthe EU continues to progress as in recent years,these countries will in a few years be as highlyintegrated into EU trade as the current EUmembers.

Chart 5 Degree of openness (2002)

(total exports and imports as a % of GDP)

Sources: IMF DOTS and Eurostat for weights.1) Weighted by nominal GDP in 2002.

0

40

80

120

160

200

0

40

80

120

160

200

euro area weighted average1)

AC weighted1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1 Belgium 2 Ireland 3 Luxembourg 4 Netherlands 5 Austria 6 Finland 7 Germany 8 Portugal

17 Czech Republic18 Slovenia19 Lithuania20 Latvia21 Cyprus22 Poland

9 France10 Spain11 Italy12 Greece13 Malta14 Estonia15 Slovakia16 Hungary

Interestingly, the countries that are currentlyhighly integrated with the EU, such as theCzech Republic, Hungary, Poland andSlovenia, are not those with the highest degreeof openness, while the most open economies,such as Estonia, Malta and Slovakia, are lessintegrated with the EU. There are specificreasons for this: for Estonia, trade with theother Baltic countries and Russia is stillrelatively important; Malta trades significantlywith Asia; and Slovakia maintains a substantialtrade share with the Czech Republic. Yet, thelower degree of trade integration with the EUmight suggest that these countries, which havesignificant trade with non-EU/non-accessioncountries, could be somewhat more exposed toexternal demand shocks originating from thirdcountries than the euro area.

Importantly, a large part (around 40%) of theacceding countries’ trade with the euro area isintra-industrial, most of which is classified asvertical intra-industrial trade, i.e. the exchange

31 McKinnon (1963), Kenen (1969).32 Krugman (1993).33 De Grauwe and Aksoy (1999).34 Frankel and Rose (1998), Maurel (2002), Fidrmuc and Schardax (2000).

33ECB



of similar goods of different quality (seeChart 7).35 This is important with respect to thequestion of euro adoption, as it suggests thatcountries with a high degree of intra-industrialtrade will be subject to similar shocks andpattern of industrial activity. At a country level,the role of intra-industrial trade seems to beparticularly high in those countries with strongtrade links with the EU. In fact, the CzechRepublic, Hungary, Poland and Slovenia have ahigh share of intra-industrial trade with the EU,ranging from 46% of total EU trade in Poland to65% in the Czech Republic. In contrast, intra-industrial trade with the EU is relatively low inLatvia and Lithuania, which have a share of lessthan 20% of total EU trade.36 The pattern ofintra-industrial trade partly mirrors the patternof trade specialisation and the role oftechnology-driven industries. For example, theimportance of technology-driven industries isrelatively high in the Czech Republic, Estonia,Hungary and Poland, while in Latvia andLithuania specialisation has taken place withrespect to labour-intensive industries.

Chart 6 Acceding countries’ trade with theEU

(exports and imports of goods to the EU as a percentage oftotal trade in goods)

Sources: IMF DOTS and Eurostat for weights.Note: Data for 1997 instead of 1995 for Belgium and Luxembourg.1) Weighted by nominal GDP in 2002.

0102030405060708090

100

0102030405060708090100

19952002EU average 20021)

AC-10 average 20021)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

1 Luxembourg 2 Portugal 3 Belgium 4 Denmark 5 Spain 6 Netherlands 7 Austria 8 France 9 Ireland

10 Sweden11 Italy12 Finland13 Germany14 UK15 Greece16 Hungary17 Poland18 Czech Republic

19 Slovenia20 Latvia21 Estonia22 Slovakia23 Malta24 Lithuania25 Cyprus

In sum, the Czech Republic, Hungary, Polandand Slovenia seem to be most integrated withthe EU in trade terms and have the highest shareof intra-industrial trade with the EU. Thissuggests that they are likely to be subject tosimilar patterns in industrial activity and shocksas current EU Member States. In contrast,Latvia and Lithuania stand out as being amongthe least integrated with the EU and also have alow degree of intra-industrial trade.

2.3.3 LABOUR MARKET FLEXIBILITY

Flexible labour and goods markets can facilitateadjustment to asymmetric shocks. According tothe OCA theory, the need for nominal exchangerate adjustments is lower the more prices andwages are flexible, particularly downwards, and

35 In comparison, the share of intra-industrial trade in the euro area is wellabove 50%.

36 However, the low degree of intra-industrial trade might also be due to thesmall size of the economies and their relatively lower degree ofdiversification.

Chart 7 Share of intra-industrial trade inacceding countries

(as a % of total EU trade, 2000)

Sources: UNECE (2002) and ECB staff calculations.

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

horizontalverticalEU averageAC-8 average

1 Czech Republic2 Hungary3 Slovenia

2 3 4 5 6 7 8 1

4 Poland5 Slovakia6 Estonia

7 Lithuania8 Latvia


the more production factors are mobile acrossand within countries.37 Moreover, labourmarket structures may have a bearing on howefficiently an economy adjusts to shocks. Theflexibility of labour markets is particularlydifficult to assess and quantify, but severalindicators may be useful, such as the degree oflabour turnover, the strictness of theemployment protection legislation in place, thegenerosity of unemployment benefit systems,the centralisation of wage bargaining, the levelof minimum wages and the flexibility ofnominal wages.

Labour market developments in accedingcountries generally reveal major futurechallenges: unemployment rates are high andrising, the share of long-term unemployment issubstantial, and labour force participation isdeclining. However, large differences existacross countries, with unemployment rates in2002 ranging from around 5% to 6% in Cyprus,Hungary and Slovenia to 19% in Poland andSlovakia (see Charts 8 and 9). Comparable dataavailable for the OECD countries show thatlong-term unemployment shares in all fourOECD members (Czech Republic, Hungary,Poland and Slovakia) are similar to those in theeuro area, despite a far shorter history ofunemployment in the transition economies.Labour force participation, although still higheras a legacy of the past regimes, continues todecline in practically all countries, and isgradually reaching euro area levels.

While part of the increase in unemployment maylargely be a consequence of the still ongoingtransition process, challenges arise mainly fromits persistence over time. The increase inunemployment in some acceding countriesduring the late 1990s despite rapid economicgrowth suggests that unemployment is mostly astructural problem rather than a cyclical one.This analysis seems to be confirmed by therelatively high incidence of long-termunemployment. In most acceding countries, theshare of those that have been unemployed for

Chart 8 Unemployment rate in the CentralEuropean acceding countries

Sources: Eurostat and IMF.

02468

101214161820

02468101214161820

1997 1998 1999 2000 2001 2002

HungaryCzech RepublicPolandSlovakiaSlovenia

Chart 9 Unemployment rate in the Balt ics ,Cyprus and Malta

Sources: Eurostat and IMF.

02468

101214161820

02468101214161820

MaltaCyprusEstoniaLatviaLithuania

1997 1998 1999 2000 2001 2002

Chart 10 Unemployment rate in the euroarea

Source: Eurostat.

02468

101214161820

02468101214161820

euro areaGreeceSpainPortugal

1997 1998 1999 2000 2001 2002

37 See Friedman (1953), Mundell (1961).

35ECB



more than a year in the total number ofunemployed is not only higher than the EUaverage, but has also been increasing in recentyears, in contrast with the experience in the EU.The lack of a distinctive correlation betweeneconomic growth and employment growth inacceding countries is also in contrast with theexperience across the EU (see Chart 11).

High rates of unemployment do not appear to becorrelated with specific exchange rate regimes.Poland, Slovakia and Lithuania, the threeacceding countries with the highestunemployment rates, themselves cover thespectrum from hard pegs to flexible exchangerate regimes. The widespread nature ofunemployment points to structural reasons(such as skills mismatches) as being a likelysource of these persistently high unemploymentrates.

This persistence can also be partly explained bythe rather low levels of inter-regional labourmobility in many countries. In fact, theincidence of high unemployment regions ismuch higher across acceding countries thanwithin the EU (see Charts 12 and 13).

Mobility across jobs and the creation of newjobs, as measured by the job turnover rate, havealso declined significantly since the early

Chart 13 Histogram of regionalunemployment rates in the larger CentralEuropean acceding countries1) (2001)(percentages)

1) Slovenia and the Baltics excluded from the analysis as they are oneNUTS-2 region each.

y-axis: relative frequencyx-axis: unemployment rate

0

5

10

15

20

25

0

5

10

15

20

25

0-2 2-4 4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20 20-22 22-24 24-26

range: 2.0-24.1mode: 20median: 14.2std. dev.: 7.2

Chart 12 Histogram of EU regionalunemployment rates (2001)

(percentages)

Sources: Eurostat, ECB staff calculations.

y-axis: relative frequencyx-axis: unemployment rate

0

5

10

15

20

25

0

5

10

15

20

25

0-2 2-4 4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20 20-22 22-24 24-26

range: 1.2-24.8mode: 4median: 6.4std. dev.: 4.6

1990s, when the initial stages of transition ledto a high turnover in jobs. Against thisbackground, mobility across sectors andregions cannot be regarded as a very effectivemechanism for absorbing idiosyncraticshocks.38

Other indicators of labour market flexibility,such as employment protection legislation, sheda more positive light on the situation inacceding countries. In fact, employment

38 See Fidrmuc (2002).

Chart 11 Correlation real GDP growth –employment growth, EU-15 and the Centraland Eastern European acceding countries(1997-2001)1)

Sources: Eurostat and ECB staff calculations.1) 1998-01 for Czech Republic, Poland and Estonia; 1999-01 for Latviaand Lithuania; and 2000-01 for Slovakia.

AC-8: R2 = 0.06 (n = 30)EU countries: Elasticity = 0.55, R2 = 0.52 (n = 75)employment growth rate (percent)real GDP growth rate (percent)

-6-4-202468

10

-6-4-20246810

y-axis:x-axis:

-2 0 2 4 6 8 10


Chart 14 Index of employment protectionlegis lat ion (1999)

Source: OECD (1999).

0

1

2

3

4

0

1

2

3

4

PT IT ES FR SW DE BE SK AT NL FI CZ PL HU DK IE UK

More restrictive

Less restrictive

legislation can be regarded as less strict than inthe euro area (see Chart 14).39

Nominal rigidities are difficult to measure andmay change as a consequence of changes in themonetary regime. Nominal wage growth in theacceding countries has on average moderatedalongside disinflation. This seems to suggestthat there is some degree of nominal wageflexibility, at least in the range of positivenominal wage growth. However, there havebeen two episodes of high nominal wagegrowth which do not fit this generally positivepicture, namely in the Czech Republic in themid-1990s and in Hungary since 2001. Itshould be borne in mind that so far there hasbeen little need for downward flexibility ofnominal wages in acceding countries owing tothe catching-up process on the one hand andhigher inflation until recently on the other. Anexception was the experience of some sectors inthe Baltic countries after the Russian crisis,where wages appeared flexible downwards.Yet, with acceding countries approaching a lowinflation environment, the need for downwardflexibility of nominal wages may become morefrequent and nominal downward rigidities morebiting, in particular if output fluctuations arepronounced.

Flexible and sufficiently differentiated wagesare of crucial importance for enhancing the

efficiency of labour markets in matching supplyand demand. In this context, whether wagebargaining takes place collectively, and if so, atwhich level, may have implications for wagegrowth. Another relevant piece of informationfor assessing nominal rigidities relates to theduration of nominal wage contracts and to theextent to which they are synchronised orstaggered. Overall, nominal wage contracts inacceding countries are typically of a relativelyshort duration (which may be related to the pastexperience of inflation persistence in somecountries). Moreover, there appears to be noparticular evidence that would point to anexceptionally high degree of staggering ofnominal wage contracts.

Wage-setting frameworks and the role of tradeunions in wage formation differ widely acrosscountries. Table 4 hereafter shows that whencollective bargaining takes place in accedingcountries, it does so mostly at the companylevel. In Slovakia, where collective bargainingtakes place predominantly at the sectoral level,there is no evidence that pay increases havebeen particularly hefty, as high and persistentunemployment has most probably dampenedwage demands. Acceding countries also exhibitrelatively low and declining rates of trade unionmembership.

Heterogeneity across acceding countries is alsoa feature of the level of the minimum wage.Excessively high statutory minimum wages maydeter employment, particularly of young agegroups. In January 2003, Slovenia had anational minimum wage broadly comparable tothat in Portugal (416 EUR per month) or inSpain (526 EUR per month), the two MemberStates with the lowest minimum wages. Allother continental acceding countries wereclearly below EU levels (see Table 5 hereafter).However, when comparing the minimum wagewith the average wage, most acceding countrieshave relatively generous minimum wages when

39 The availability of comparable cross-country information is very limited.Given the numerous changes in labour market legislation across countrieson an ongoing basis, assessment based on stock-taking exercisesconducted in the past must be particularly cautious.

37ECB



compared with Spain (where the minimum wagerepresents around 35% of the average wage)and Portugal (where the corresponding figure is44%).

The level of unemployment benefits asmeasured by the replacement ratio (i.e. initialbenefit over previous earned income) isrelatively modest in the Baltic countries, butrelatively generous in countries such asHungary, Slovenia and Slovakia. Comparingunemployment benefit systems across countriesusing indicators such as the replacement ratio isnevertheless subject to great uncertainty. Thisis because the parameters of unemploymentbenefits, such as the length of entitlement tobenefits, differ greatly across countriesaccording to workers’ contributive history, andthus the average unemployed person may wellreceive a very different benefit than thestatutory maximum. Thus, looking at the publicexpenditure on unemployment benefits as aproportion of GDP complements theinformation on the statutory generosity of theunemployment benefit system. This indicatesthat spending on unemployment benefits isparticularly low in the Baltic countries,especially in Estonia and Lithuania, while

Bargaining level Collective Average paybargaining increase in 2001 Usual weekly

Company Sectoral Intersectoralcoverage (percent, adjusted working hours

(percent of for inflation)workers)

Central Europe

Czech Republic *** * 25-30 3.9 42.3Hungary *** ** * 34 8.4 42.9Poland *** * 40 1.4 45.1Slovakia * *** 48 0.9 42.9Slovenia * ** *** ca. 100 3.5 39.7

Baltics

Estonia *** * 29 7.2 42.4Latvia *** * * < 20 3.7 44.3Lithuania *** * 10-15 0.8 44.8

Table 4 Wage-sett ing framework in acceding countries

(2001)

Sources: ECB staff compilation based on EIRO/ILO (2002) and ETUC (2002).Note: (*) Existing level of wage bargaining; (**) Important but not dominant level of wage bargaining; (***) Dominant level of wagebargaining.

earlier data show that central European accedingcountries typically spend more than the Balticsin unemployment benefits although, with theexception of Poland, they are still far below theEU average.

It should be noted that the issue of nominalrigidities not only pertains to labour markets,but equally to product markets. Nominal priceflexibility is typically closely linked to thedegree of product market regulation. OECDcomposite indicators on product marketregulation that relate to the status quo in the late1990s suggest that acceding countries havemoved on fairly quickly with product marketderegulation.40 The ongoing adoption of thesingle market acquis has been instrumental inadvancing this process. Still, a special issue foracceding countries in this regard should bementioned, namely that of administered prices,which have a relatively high share in theconsumer baskets of some of these countries.This may insert an element of inflexibility intoproduct markets. The analysis of productmarket flexibility would benefit from furtherdeepening in the future.

40 Unfortunately, no updates of these indexes are available.


Unemployment benefitsExpenditure on

Minimum Minimum Replacement Maximum Total active labourwage wage ratio1) duration2) expenditure market policies

(euro p/month) (% of av. wage) (%) (months) (% of GDP in 2001) (% of GDP in 2001)

Central Europe

Czech Republic 199 47 50 6-12 0.31 (1999) 0.19 (1999)Hungary 212 56 64 3-12 0.56 (1997) 0.40 (1997)Poland 201 39 40 12-24 1.71 (1996) 0.49 (1996)Slovakia 118 37 60 6-12 0.54 (1996) 0.56 (1996)Slovenia 451 43 63 3-24 0.89 (1998) 0.83 (1998)

Baltics

Estonia 138 42 50 6-12 0.13 0.06Latvia 116 40 50 9 0.50 0.14Lithuania 125 47 253) n.a. 0.15 0.12

EU average 60 1.73 1.16

Table 5 Labour market welfare systems in acceding countries

(January 2003, unless otherwise specified)

Sources: ECB staff compilation based on Eurostat (2003), OECD (2003), Riboud et al. (2002), Cazes (2002).1) Initial benefit level divided by previous earned income.2) Duration is typically a function of the worker’s contributive history.3) Lithuania’s unemployment benefits do not follow an insurance principle: they are either a flat rate (state supported income) or variableup to one-fourth of the average gross wage.

The degree of labour market flexibility differsacross acceding countries. Indeed, theassessment critically depends on whichindicators one chooses to focus upon. Drawingconclusions from labour market flexibilityindicators is, moreover, particularly difficult inthe case of acceding countries since the ongoingprocess of transition may have a significanteffect on these indicators. Despite thesecaveats, a number of common features emerge.Labour turnover, which had been particularlyhigh in the first half of the 1990s, has sincebeen decreasing in most cases to reach levelsthat are more comparable with those prevailingin current EU countries. Workers in manyacceding countries remain on average a shortertime in their jobs compared with workers in theOECD countries. This is particularly the case inEstonia, which displays the highest proportionof workers with a short job tenure (less thanone year) and the smallest share of workerswith long job tenure (more than ten years),while lengthy average job tenures point towardssome limitations to labour market flexibility inSlovenia and Poland.

At an individual country level, labour marketsin acceding countries differ in their degree offlexibility across the range of institutionalfeatures reviewed here. Estonia and Hungaryappear to have the most flexible labour markets,underpinned by relatively weak labourprotection legislation, the limited role of tradeunions, low levels of social protection and ahigh degree of wage flexibility. Hungary hasseen, however, substantial pay increases inrecent years and suffers from one of the highesttaxation rates on labour. In addition, the jobturnover rate is particularly high in Estonia. Incontrast, for the Czech Republic, Poland,Slovakia and to some extent also Slovenia,several indicators, such as the wage bargainingstructure and the generosity of their socialbenefit systems, suggest that their labourmarkets may be more rigid.

It is difficult to project future trends in labourmarket flexibility and mobility. Given the highshare of long-term unemployment, however, itis clear that the high levels of structuralunemployment in a number of accedingcountries are likely to persist for some time.The main question with regard to future

39ECB



developments appears to be to what extentacceding countries will experience strongerupward real wage pressures in the wake of EUentry, as regulations and labour standards willtend to become more comparable with the EU. Afurther issue is how prospective participation inthe euro area would affect wage developments,in particular in a catching-up setting. However,wage pressures within a monetary union are notlikely to be immediate, as the free movement ofpersons will be restricted for a period of up tofive (and in some cases seven) years after EUaccession for several countries. This issue willbe discussed in more depth later in the paper.41

Cross-border labour mobility between currentand new Member States is also to be limited by atransitional arrangement agreed in the context ofthe accession negotiations. For the first twoyears after accession, Member States will acceptworkers from the new Member States accordingto national rules (the transitional agreement doesnot apply, however, to Cyprus and Malta). Twoyears after accession, the Commission will reporton the situation and Member States will have toannounce the system they wish to use from thenon, which could be applied for a further threeyears. Following this period, the remainingMember States that restrict access to their labourmarkets will again be invited to open their labourmarkets entirely. Only if a Member State canshow serious disturbances in its labour market,or the threat of such disturbances, will it be ableto continue to require work permits for amaximum of a further two years. The right ofestablishment is not affected by this transitionalperiod, and people will also be completely free todeliver services across borders as self-employedpersons or as companies. Austria and Germany,however, have the right to apply measures toaddress serious disturbances or the threat thereofin specific sensitive service sectors on theirlabour markets, which could arise in certainregions from the cross-border provision ofservices.

At this stage it is difficult to foresee what thelikely impact of this transitional arrangementwill be on limiting cross-border labour

mobility, as it will ultimately depend on howliberal the national measures of the currentMember States turn out to be. In the past, whenSpain and Portugal entered the EU, the phasing-in period of seven-ten years was subsequentlyshortened. In fact, in some Member States fulllabour market access may occur immediatelyupon accession. Moreover, in a declarationattached to the Accession Treaty, currentMember States have committed themselves togranting increased labour market access undernational law, with a view to speeding up theapproximation to the acquis.

2.3.4 FINANCIAL SECTOR INTEGRATION

A high degree of financial integration has beenstressed by some authors as conducive to thesustainability of a monetary union as, incombination with sound and well-developedfinancial markets, it is expected to helpsmoothen the impact of asymmetric shocks byfacilitating cross-border flows of capital.42

Within the “new” OCA theory, financialintegration is often seen to include the similarityof financial structures, institutions andlegislation, because similar structural features,including the financing pattern of privateenterprises (i.e. share of bank loans to totalliabilities) and balance sheets of households(i.e. share of mortgage payments to totalpayments), are seen to increase and synchronisethe effectiveness by which monetary policy istransmitted to the real economy. Thus, financialintegration can also be expected to contribute tothe better functioning of a monetary union.43 Tothe extent that financial integration includes

41 See especially Section 2.3.7.42 See Ingram (1973), Scitovsky (1966). Mundell (1973) takes this argument

a step further by suggesting that, under full financial market integration,countries which are exposed to asymmetric shocks may particularly profitfrom monetary unification. The idea underlying this argument is that usinga common currency will facilitate portfolio diversification, which allowscountries to adjust more smoothly and at lower costs to asymmetric realshocks, owing to mutual claims on each other’s resources.

43 In this context, it should be noted that financial integration was regardedas necessary for the euro area to ensure a comparable implementation ofthe single monetary policy in EMU. For an overview of the empiricalliterature with respect to financial integration in the euro area, seeMongelli (2002).


similar supervisory and regulatory frameworks,it could also contribute to greater financialstability.

Overall, the degree of financial integrationbetween acceding countries and the euro areaappears to be high, although considerabledifferences exist across indicators andcountries. As for capital flows, virtually allacceding countries have experienced large andincreasing inflows in recent years. By far thelargest component of these flows is foreigndirect investment, which has accounted forapproximately two-thirds of all capital flows toacceding countries in recent years. The EUplays a very important role as a source of suchFDI: in 2000, approximately 80% of FDIinflows originated from EU Member States. Asa consequence, cross-border ownership hasincreased substantially both in the real sectorand in the financial sector (see also below).

Moreover, the convergence of interest rates tolevels prevailing in the euro area has beensignificant, and spreads between deposit andlending rates have declined to euro area levelsin recent years (see Chart 15). However,significant differences at the country level

Chart 15 Deposit and lending interest rates(2002)

(percentages)

Source: IMF (data for Portugal are not available).

deposit ratelending ratespread

14

12

10

8

6

4

2

0

14

12

10

8

6

4

2

0

1 Spain2 Euro area3 Greece4 Malta5 Hungary

6 Cyprus7 Slovakia8 Estonia9 Czech Republic

10 Latvia11 Slovenia12 Lithuania13 Poland

1 2 3 4 5 6 7 8 9 10 11 12 13

Chart 16 Size of the banking sector (2002)

(banking assets as a % of GDP)

1 Greece2 Spain3 Portugal4 Euro area5 Lithuania

6 Poland 7 Hungary 8 Estonia 9 Latvia10 Slovenia

11 Slovakia12 Czech Republic 13 Cyprus14 Malta

300

250

200

150

100

50

0

300

250

200

150

100

50

01 2 3 4 5 6 7 8 9 10 11 12 13 14

Sources: IMF and national central banks.

continue to exist. For example, deposit/lendingspreads are considerably higher in Latvia andSlovenia than in the euro area, and are close totwice as high in Lithuania and Poland.

Acceding countries also seem to be highlyintegrated when the similarity of their financialstructures and institutions is considered. This islargely a result of the implementation of theacquis communautaire and the high degree ofmarket integration.

The level of financial intermediation, however,remains low in most acceding countries (seeChart 16), although banking assets as a ratio toGDP have risen in recent years. Only Cyprusand Malta have reached a level comparable withthe euro area average. The low depth offinancial intermediation partly reflects moderateGDP per capita levels. More importantly,though, it is due to the relatively short historyof banking sectors and the transition process,which included bank consolidation, graduallyevolving track records of new companiesand private households, and a strong presenceof foreign-owned companies. Financialintermediation can be expected to rise moreswiftly in the period ahead, as the transition

41ECB



process nears completion and the catching-upprocess advances. At the current juncture,intermediation is driven in particular byincreased leverage of SMEs and households.

Banking sector soundness has substantiallyimproved in most acceding countries in recentyears. Capitalisation, profitability and assetquality have strengthened considerably, inparticular in the Czech Republic and inHungary. Poland is somewhat an exception tothe general picture, both in terms of profitabilityand with regard to the share of non-performingloans, which is apparently to a large extent dueto the very low growth phase in 2001 and 2002.Moreover, in some country cases, the existenceof currency mismatches does warrant attentionfrom a financial stability viewpoint before theultimate adoption of the euro.

Taken together, all these observations implythat monetary transmission through interest andcredit channels has become more effective inmost acceding countries owing to improvedbanking sector soundness, but is stillconstrained as a consequence of the low depthof financial intermediation.

Another basic feature of acceding countries’financial markets is the dominance of thebanking sector over capital markets, a featurewhich is pronounced in all ten countries. Again,this is both the result of development levels andtransition strategies that mostly focused onrehabilitating the banking system; attempts insome countries to foster capital marketdevelopment, however, were only partiallysuccessful. This is particularly true for thecorporate bond market, which is in its infancyin all acceding countries, and for the equitymarket, which has made limited headway onlyin the Czech Republic, Hungary and Poland.While the trading volume on these markets hasincreased, it is indicative that the volume of newequity capital raised through these markets hasremained negligible. Concerning other financialmarket segments, liquid foreign exchangemarkets have developed in four acceding

countries, namely in the Czech Republic,Hungary and Poland (both spot and forwardmarkets) as well as in Slovakia (spot market).Government bill and bond markets are welldeveloped in the Czech Republic, Hungary andPoland.44

A further characteristic of financial systems inacceding countries is substantial foreignownership, which can be observed in all marketsegments, but mostly in the banking sector.Foreigners, mostly from EU countries, own inseven out of ten countries more than 80% of thetotal assets of commercial banks, which is wellabove the share of cross-border ownership inthe EU. At a country level, Estonia, Lithuania,Malta and Slovakia have the highest foreign-owned share of total assets, compared withCyprus and Slovenia, which have the lowestshare (see Table 6). The strong presence offoreign-owned banks has been instrumental inbroadly improving the performance of banks inacceding countries. Foreign ownership willtransform but not necessarily mitigatesupervision challenges, in particular upon EUaccession, when the home country principletakes effect for branches established inacceding countries. In many cases, subsidiariesand branches in acceding countries only accountfor a small fraction of the balance sheets of theparent banks, while at the same time being ofsystemic relevance for the acceding countries’financial systems. This raises a need forstrengthened cooperation between home andhost supervisory authorities, for instance in theexchange of information and in potential crisismanagement. Foreigners, in particular from theEU, play a significant role in stock markets inthe acceding countries, partly because marketdeepening has typically been closely linked toprivatisation, a process in which foreigninvestors have been at the forefront. Similarly,euro area investors play a substantial role in thefixed income markets and hold a large share ofacceding countries’ government bonds.

44 It is probably fair to say that the development of equity and corporate bondmarkets has a less direct bearing for the timing of monetary unionparticipation. Still, this analysis is presented here to complete the picture.


Looking ahead, financial integration in accedingcountries is expected to increase further.Capital flows will increase as capital accountliberalisation is fully concluded and EUaccession increases confidence and reducesrisks. Interest rate spreads can be expectedto narrow, as financial sectors become moreefficient over time. Furthermore, with respect tofinancial structures and institutions, furtherprogress can be expected, with the adjustmentsarising from the acquis communautaire andcentral bank cooperation with the Eurosystemcertainly representing a driving force.

In this context, it is worth recalling that allacceding countries will implement and enforcethe acquis as from the date of accession, withonly a few, temporary exceptions. In particular,under the chapter “free movement of capital”,negotiations have been closed with sometransitional arrangements on real estateinvestments (Slovenia), secondary residenceinvestments (Czech Republic, Hungary,Poland, and Cyprus)45 and agricultural land andforest investments (Czech Republic, Slovakia,Hungary, Lithuania, Latvia, Estonia andPoland). Under the chapter “freedom to provideservices”, some specific transitionalarrangements have been granted in the financialsector regarding full compliance with the acquisof the cooperative credit institutions (Cyprus,Hungary and Poland), savings and loans

Share of totalNumber of assets beingcommercial Of which: foreign-owned

banks foreign-owned (as a %)

Cyprus 13 5 13Czech Republic 37 26 86Estonia 7 4 99Hungary 39 27 87Latvia 19 12 47Lithuania 14 4 94Malta 13 10 99Poland 59 45 78Slovakia 18 15 96Slovenia 22 6 35

Table 6 Foreign ownership of banks

(2002)

Sources: EBRD, national central banks and/or national supervisory authorities.

undertakings (Slovenia), deposit guaranteeschemes (Estonia, Latvia, Lithuania andSlovenia) and investor compensation schemes(Estonia, Hungary, Latvia, Lithuania, Poland,Slovakia and Slovenia). Owing to their specificnature as well as their limited duration andscope, the transitional arrangements will notgenuinely affect further the financial integrationof the acceding countries.

Finally, the experience of the current euro areacountries suggests that the eventual adoption ofthe single currency will give a further boost todeepening financial integration, increasing thehomogeneity of structures and the efficiency ofbanks and capital markets.

2.3.5 REAL CONVERGENCE AND STRUCTURALDIFFERENCES IN ECONOMIC DYNAMICS

Real convergence, i.e. the catching-up ofincome levels and the adjustment of realeconomic structures to those prevailing in theeuro area, is one of the key economiccharacteristics of acceding countries. Whiledifferences in income levels and somedifferences in economic structures can, inprinciple, be compatible with participation in a

45 Malta has been granted the right to maintain on a permanent basis itsnational legislation regarding the acquisition of secondary residences.

43ECB



monetary union, such differences can haveimportant implications for the appropriatenessof the single monetary policy to individualmembers. In particular, real convergence islikely to be reflected in differences in economicdynamics. Differences in economic dynamicsbetween members are increasingly regarded asone of the main potential costs of a monetaryunion. If such differences are substantialand persistent, abandoning an independentmonetary policy as a stabilisation instrumentmay entail welfare losses. Inappropriatemacroeconomic policies would exacerbateperiods of overheating or downturns, leadto boom-bust cycles, and may bring abouthigher average unemployment over timethrough hysteresis effects. This sectionexplores these considerations for the currentacceding countries.

Over the past 13 years, economic growth inmost acceding countries has developed quitedifferently from the euro area, with theexception of Cyprus and Malta.46 Followinginitial transformational recessions in the early1990s, most acceding economies have expandedfaster than the euro area, experienced sharpercyclical fluctuations and have been subject toseveral major idiosyncratic shocks, includingstabilisation crises. The important question iswhether these structural differences in thepattern of economic dynamics between accedingcountries and the euro area will remainsignificant beyond 2007-2009, which is thetarget period of many acceding countries foreuro area entry. The answer to this question ismostly empirical. The following analysisexamines the basic properties of output dynamicssince 1996 (the first half of the decade, bycontrast, was mostly shaped by systemictransformation) and aims to assess structuraldifferences vis-à-vis the euro area. Note thatowing to the short time series, all findings have tobe taken with caution and estimates warrantupdating as new data become available.

In this analysis, three features stand out, all ofwhich have a bearing on the choice of timing ofeuro area entry and which will be briefly

addressed in this section: growth rates arepersistently higher in acceding countries, as aregrowth fluctuations (i.e. amplitudes ofupswings and downturns), while businesscycles are not always closely synchronised withthe euro area.

A convenient way of condensing thesedifferences into a single indicator is Theil’sinequality coefficient. This coefficientmeasures a scaled root mean squared differencebetween two series. It takes values betweenzero (perfect fit) and unity. The statisticprovides two important benefits. First, it allowsthe comparison of different pairs of variables atdifferent scales, with respect to a broad conceptof inequality. Second, inequality of time seriescan be decomposed into its main statisticalfactors, i.e. mean difference, difference invariability and lack of correlation.

Chart 17 shows GDP growth inequality asmeasured by Theil’s inequality coefficientsrelative to the euro area for three groupsover the period 1996 to mid-2003, namely theCEE (Central and Eastern European) accedingcountries, euro area peripheral countries (Greeceand Portugal), and the so-called euro area pre-ins(United Kingdom, Sweden and Denmark).47

It turns out that inequality is on average higherfor acceding countries (mean coefficient 0.43)than for the peripherals (0.32) and thepre-ins (0.24). Moreover, there has beenconsiderable divergence in inequality amongacceding countries. The countries witheconomic dynamics most similar to the euroarea are Hungary (coefficient: 0.27), Slovenia(0.31) and Poland (0.40). Meanwhile, the CzechRepublic, Estonia, Latvia and Lithuania postthe largest differences. Chart 18 repeats theexercise using Germany as a reference. Resultson the relative inequality of the accedingcountries, peripherals and pre-ins are similar to

46 In this section the term “acceding countries” refers to the eight Centraland Eastern European acceding countries (thus not includingCyprus and Malta).

47 This part of the analysis excludes Ireland, since quarterly GDP datastretching back to 1996 are not available for this country.


Chart 17 Thei l ’s inequal ity coef f ic ients forgrowth between various countries and theeuro area(based on year-on-year real GDP growth rates, 1996-2003Q2)

Sources: ECB staff calculations.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.00

0.10

0.20

0.30

0.40

0.50

0.60

LT EE CZ LV SK PL SI HU GR PT SE DK UK

CEE acceding countries euro areapre-ins

euro areaperipherals

Chart 18 Thei l ’s inequal ity coef f ic ients forgrowth between various countries andGermany(based on year-on-year real GDP growth rates, 1996-2003Q2)


CEE acceding countries euro areapre-ins

large euro areacountries

euro area peripherals

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

LT LV EE SK CZ PL SI HU GR PT UK SE DK ES NL FR IT

those obtained against the euro area benchmark.However, two additional findings are ofinterest. First, inequality versus Germany isgreater than versus the euro area in all cases,suggesting that there is an issue of dispersionamong euro area countries in addition to thedifference between different “blocks”. Second,one can add France, Italy, the Netherlands andSpain individually as an additional control. Theinequality of these countries versus Germany iscomparable to that of the pre-ins, but wellbelow that of the acceding countries.

What explains these inequalities? The statisticalcomponents of different growth behaviour areshown in Charts 19 and 20. Technicallyspeaking, they reveal how inequality is relatedto different means, different variances and alack of covariance between the individualcountries and the euro area (or Germany). Whenlooking at inequality versus the euro area, it isstriking that for all country groups thecovariance of growth with the euro area isimperfect and roughly of a similar size. Thisdifference may reflect lack of cycle synchrony,perhaps arising from idiosyncratic shocks.There is a considerable difference in the meansand, to a lesser extent, variances of the separatecountry groups, however. Thus, for the pre-insneither means nor variances put economicgrowth far from that of the euro area. Theperipheral countries add the difference in mean

growth as a significant component ofinequality. The CEE countries exhibit an evenmore sizeable mean difference and areadditionally subject to a higher level ofvariance, suggesting that their trend growth andcycle amplitudes tend to be larger. Theinequality versus Germany features meangrowth difference as a prominent characteristicof all country groups, including the largecurrent “euro-ins”. However, similar to the euroarea benchmark, the acceding countries tend tohave a relatively higher variability of output.

- Higher growth rates

An obvious feature of the data is that real GDPhas expanded considerably faster in theacceding countries than in the euro area over thepast few years. On average, the accedingcountries posted GDP growth rates of 4.1%from 1996 to mid-2003, compared with 2.3% inthe euro area, 2.7% and 3.6%% in Portugal andGreece, respectively, and from 2.3% to 2.8% onaverage in Denmark, Sweden and the UnitedKingdom. Within the group of accedingcountries, the Czech Republic is a clear outlier,with real GDP expanding by only 1.7% onaverage per year due to the severe stabilisationcrisis of 1997-99 (see Chart 21).

Higher growth rates in acceding countriescompared with the euro area can mainly be

45ECB



Chart 19 Dif ferences in GDP growthcompared to the euro area

(Theil’s inequality coefficients and components, 1996-2003Q2, based on year-on-year GDP growth rates)


0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.4

due to covariance differencedue to variance differencedue to mean difference

CEE 8 peripherals pre-ins

Chart 20 Dif ferences in GDP growthcompared to Germany

(Theil’s inequality coefficients and components, 1996-2003Q2, based on year-on-year GDP growth rates)


due to covariance differencedue to variance differencedue to mean difference

CEE 8 peripherals pre-ins FR, IT, ES, NL0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

explained by the catching-up of theseeconomies, as well as initially by the recoveryfrom the “transformational recession” of thefirst half of the 1990s. While higher growth isneeded to converge with the per capita incomelevels in the euro area, such structuraldifferences in economic dynamics may increasethe stabilisation costs that an acceding countrywould incur if it abandoned its own monetarypolicy. Different long-term growth rates notonly imply higher inflation rates, for examplethrough the Balassa-Samuelson effect; there isalso a risk of inappropriately low nominal, andthus real, short-term interest rates. Incombination with a high marginal return oncapital, these lower interest rates couldpotentially fuel a credit boom that, owing toinevitable supply-side constraints facing theinvestment demands, would give rise to assetbubbles and boom-bust cycles (for furtherdetails, see also the next section).

- Higher output fluctuations

On the heels of faster economic growth,most acceding countries have also experiencedwider output fluctuations. The averagestandard deviation of real GDP growthwas 2.4 percentage points in the acceding

countries from 1996 to mid-2003, higher than inthe euro area and the euro area periphery (withrespectively 1.3 and 1.5 percentage points). Thevariance of individual countries was on averagelarger in acceding countries than in the euroarea, although the standard deviations arescattered in a broad range from 1.4% inHungary and Slovenia to 3.6% in Estonia andLithuania. The five central European economiestogether, namely the Czech Republic, Hungary,

Chart 21 GDP growth and standarddeviations in Europe

(GDP, annual percentage change, 1996-mid-2003)


0

1

2

3

4

5

6

0

1

2

3

4

5

6

LV EE LT PL SI SK HU CZ GR UK PT SE EA DK

acceding countries

averagest. dev.

euro area and other EU countries


Poland, Slovakia and Slovenia, posted onaverage a much smaller standard deviation (1.9percentage points) than the Baltic countries (3.3percentage points). This finding partly reflectsthe recession in the aftermath of the Russiancrisis of 1998 and the subsequent recovery, butmay also have to do with country size.

Higher output fluctuation in acceding countriescompared with the euro area can be mainlyexplained by the fact that the transition processand the implementation of structural reformshave followed a fairly uneven path. Moreover,acceding countries have experienced onlyimperfect access to international capital marketsas they have been exposed to stronger changesin investor sentiment. Most importantly, highinvestment ratios in most acceding countries,combined with the fact that capital spendingtends to be more cyclical than consumption,suggests that during the catching-up period,growth fluctuations will remain larger.Interestingly, the growth differential betweenthe euro area and acceding countries has notdiminished over the sample period. A period ofnarrowing growth differences from 1996 to1999 has given way to a renewed divergenceover the past few years (see Chart 22). Withrespect to monetary policy, large differences inoutput fluctuations could imply that in anenlarged euro area, monetary policy would notbe sufficiently counter-cyclical for thosecountries with higher fluctuations.

In this context, however, it must first beinvestigated whether monetary policy hasplayed a useful stabilisation role, or whether ithas instead added to cyclical fluctuations.Given the predominant importance of theexchange rate channel in the transmission ofmonetary impulses to the real economy (interestand credit channels play a limited role as a resultof the low degree of financial intermediation),the issue to be considered is whether exchangerate fluctuations have been a useful adjustmenttool in dealing with adverse shocks – in whichcase they would stabilise domestic outputgrowth – or whether they have disrupted tradeand financial relations and have thereby

possibly magnified domestic outputfluctuations. Obviously, only in the first casewould the abandoning of the exchange rate toolprove costly. Background analysis on thisparticular issue suggests that the exchange ratehas indeed played a role as a shock absorber,which has been larger in countries with flexibleexchange rate regimes than in countrieswith tight exchange rate management.48 Thisconclusion is based on the finding that incountries with more flexible exchange ratearrangements, both nominal and real exchangerates have reacted to shocks, rather than havingled to them. This might support the view thatflexible exchange rates are an efficient tool foradjusting to asymmetric shocks, also reflectingthe fact that countries with more rigid regimeswere less able to adjust the real exchange rate togrowth differentials and fluctuations.

- Business cycle synchronisation

While acceding countries have on averagehigher and more volatile growth rates than theeuro area, this does not necessarily implydivergent business cycles, which is another

Chart 22 GDP growth

(quarterly data, annual percentage changes)

Source: Eurostat.1) Weighted by nominal GDP in 2002, excl. Cyprus and Malta.

-1

0

1

2

3

4

5

6

-1

0

1

2

3

4

5

6

acceding countries average1)

euro area

Mar.1996 1997 1998 1999 2000 2001 2002

Mar.2003

48 Süppel (2003).

47ECB



Chart 23 Correlation of de-trended GDPgrowth with the euro area

(based on quarterly year-on-year growth rates over 1996-mid-2003)


-0.4

-0.2

0

0.2

0.4

0.6

0.8

-0.4

-0.2

0

0.2

0.4

0.6

0.8

HU SE UK DK SI PT LV EE PL CZ GR LT SK

important factor when discussing whethercountries are already well prepared to join themonetary union and to abandon flexibleexchange rates. Strongly correlated cyclicalswings across countries imply that countries areexposed to similar shocks and respond in asimilar way.

With regard to the symmetry of economicfluctuations, the assessment is diverse acrosscountries and benchmarks. Furthermore,correlation needs to be estimated and is thusmore subject to judgement. The followinganalysis uses again data from 1996 until mid-2003 as well as four different measures toestimate correlation, in particular de-trendedGDP growth, short-term trends in industrialoutput growth, estimated broad cyclecomponents, and supply and demand shocks asidentified by a structural VAR model. Theseapproaches have complementary advantagesand drawbacks, suggesting that results thatbroadly hold across methods might bereasonably robust and credible.49

The correlation coefficients of de-trendedannual GDP growth (at a quarterly frequency)are presented in Chart 23.50 The “euro pre-ins”(Denmark, Sweden and the United Kingdom)post the highest correlation with the euro area atan (un-weighted) average of 0.45. They arefollowed by a correlation coefficient of 0.21 forthe acceding countries and of 0.21 as wellfor the two peripheral euro area economies(Portugal and Greece). Among individualacceding countries, the high correlation ofHungary and Slovenia with the euro area standsout.

The disadvantage of using GDP correlation tocapture cycle synchrony is that even after long-term trend adjustment, the coefficients may bebiased owing to technical correlation. Inparticular, the correlation of the centralEuropean economies with Germany may reflectsimilar weather conditions and calendar factors.A standard way of avoiding the problem is touse filters to extract the short-term GDP trends(through moving averages or medians).

However, the available times series are tooshort (30 observations) to do this in ameaningful way. Monthly data, which providemore observations, are better suited for suchsmoothing. They additionally allow to captureshort-term dynamics that may be left unnoticedwith lower frequency data, such as GDP series,that are available on a quarterly or annual basis.

The most popular proxy for monthly activity isindustrial production. Data are more completeand have a longer history than GDP data. Inparticular, Ireland can now be included in thesubsequent analysis. In addition, industryrepresents a substantial share of GDP inthe acceding countries of central and easternEurope (about 32% on average in 2002), and istypically the most decisive sector for cyclicaldynamics. This series provides enoughobservations to extract a short-term trend frommonthly growth data using a Hodrick-Prescottfilter (with a smoothing factor of 100). Giventhe higher frequency of the data, this short-term trend is used as a proxy for cyclicalfluctuations.

49 It should borne in mind that estimated correlation coefficients forGreece, Ireland and Portugal with the euro area – hereafter reported –may be positively biased given that these countries are euro area MemberStates themselves. However, since they account for a small share of euroarea GDP, this bias is unlikely to be large. For a recent other study ofbusiness cycle synchronisation between euro area countries andacceding countries, see Darvas and Szapáry (2004).

50 GDP series are de-trended by subtracting a quasi linear trend estimatedby the Hodrick-Prescott filter with a smoothing parameter of 14,000.


Chart 24 shows correlation coefficients forthis measure. As in the GDP analysis, thecorrelation between the euro area and the three“euro pre-ins” is strongest with an average (un-weighted) coefficient of 0.79. It is about 0.39for the three euro area peripheral economies.The acceding countries from central and easternEurope post an even higher average correlationcoefficient of 0.47. However, the dispersion ofthe group is very wide. Hungary stands out witha coefficient of 0.94, which is the highest of allcountries in the panel. The industry short-termtrends of Slovenia, Poland and Estonia havealso been strongly correlated with the euro area.The Czech Republic and Latvia post somepositive correlation in their industry short-termtrend with the euro area. Interestingly, unlike inthe case of GDP, Slovakia is now positivelycorrelated with the euro area. However,Lithuania remains negatively correlated.

It comes as no surprise that industry cycles aremore closely aligned than GDP for all countrygroups. Merchandise trade integration betweenthe euro area and the acceding countries fromcentral and eastern Europe is high and the bulkof foreign direct investment from West to Eastis also mostly in the manufacturing sector.Finally, manufacturing activity across countriesis subject to global cycles, particularly ininventory and investment spending. However,for all of these reasons, correlation of industrydata may overstate the co-movements of theoverall economies. And it is the latter thatshould matter for monetary policy.

Therefore a broad indicator for the businesscycle is additionally estimated on a monthlybasis. Compared with simple GDP growth, thishas the advantage of avoiding correlation owingto joined quarterly volatility and of examiningdynamics at a higher frequency. In addition,compared with the monthly industrialproduction series, it incorporates the dynamicsof more sectors. Indeed, non-tradables sectors,such as retail services and construction, oftenfollow dynamics that are more dependent onidiosyncratic domestic growth factors, such asmonetary conditions or fiscal policy.

Three separate monthly indicator sets are usedfor each country to distil from them a joinedcyclical factor: the annual growth of industrialproduction, the annual growth of retail salesvolumes, and the annual growth of constructionoutput. In some countries where not all datawere available, surveys have been used tocapture retail and construction activity.51 Thejoined cyclical component has been estimatedby using a state-space model of the Stock andWatson (1991) type that identifies the cycle as ajoined linear component of all sectoral cycles.This component can then be smoothed again bya Hodrick-Prescott filter to rid it of short-termvolatility.

The importance of looking at the broad cycleestimates rather than industry alone can bedemonstrated by the example of the CzechRepublic (see Charts 25 to 28 below). The chartof Czech and euro area industry growth showsthat both economies have been at least broadlycorrelated over the sample period. The chart ofthe estimated joint cycle component of industry,construction and retail sales, however, suggestsotherwise and is more in line with popularperceptions or GDP data. While the CzechRepublic suffered a deep downturn in 1998-

51 Series are reduced by dividing them by their respective standarddeviation.

Chart 24 Correlation of short-termindustrial output growth trends with theeuro area(based on quarterly year-on-year growth rates over 1996-mid-2003)


-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

HU SE IE UK SK DK SI PL EE GR CZ LV PT LT

49ECB



1999 that affected all sectors, the euro area’sdip was limited to industry, and the economy asa whole remained close to a cyclical high. Thenin 2000-2001, when the euro area cycleweakened, the Czech Republic recovered on abroad basis.

Thus the smoothed broad cycle estimates seemto be a particularly useful tool for estimating thesymmetry of economic fluctuations. Correlationcoefficients have been computed and are

Chart 25 Czech and euro area industrialproduction

(year-on-year growth rates, as a % and normalised)

Source: Datastream.

-3

-2

-1

0

1

2

3

-3

-2

-1

0

1

2

3

Jan.1996 1997 1998 1999 2000 2001 2002

Jan.2003

euro areaCzech Republic

Chart 26 Czech and euro area broad cycle

(Kalman filter estimates, smoothed and normalised)

-3

-2

-1

0

1

2

3

-3

-2

-1

0

1

2

3

Jan.1996 1997 1998 1999 2000 2001 2002

Jan.2003

euro areaCzech Republic

bankingcrisis

Source: ECB staff calculations.

presented in Chart 29. They deliver severalimportant messages:

The (un-weighted) average of the correlationcoefficients of the central and eastern Europeanacceding countries with the euro area fallssubstantially to just 0.13. The coefficient of the“euro pre-ins” stands at 0.43, while theperipheral euro area countries still show a highcorrelation of 0.51. This difference relative toindustry data is not too surprising, however.

Chart 27 Sector growth and broad cycle inthe Czech Republic

(year-on-year growth rates and HP smoothed estimates bythe Kalman filter for the broad cycle; as a % andnormalised)

Sources: Bloomberg, Datastream, Eurostat and ECB staff calculations.

-3

-2

-1

0

1

2

3

-3

-2

-1

0

1

2

3

broad cycle componentretail sales volumeretail salesindustrial production

Jan.1996 1997 1998 1999 2000 2001 2002

Jan.2003

Chart 28 Sector growth and broad cycle inthe euro area

(year-on-year growth rates and HP smoothed estimates bythe Kalman filter for the broad cycle; as a % andnormalised)

Sources: Bloomberg, Datastream, Eurostat and ECB staff calculations.

broad cycle componentindustrial productionretail salesconstruction

Jan.1996 1997 1998 1999 2000 2001 2002

Jan.2003

-3

-2

-1

0

1

2

3

-3

-2

-1

0

1

2

3


Given the geographic proximity, the GDP andindustrial production correlations between theeuro area and acceding countries from centraland eastern Europe have possibly been biased tothe high side by e.g. joined calendar andweather factors.

All the above analyses have looked at thecorrelation of various measures of aggregateoutput. A popular alternative is to use timeseries of both GDP and the GDP deflator inorder to distinguish and identify aggregatedemand and supply shocks via a structural VARmodel of the Blanchard-Quah (1989) type.52

The correlation of the estimated demand andsupply shocks for the CEE acceding countries,the “euro peripherals” and the “euro pre-ins”has been plotted in Chart 30. Countries whereboth demand and supply shocks are positivelycorrelated with the euro area are located in theupper right quadrant of the plot. This groupincludes four countries whose de-trended GDPgrowth and industrial output growth trendswere found to be positively correlated withthe euro area, namely the Czech Republic,Denmark, Latvia and Slovenia. Interestingly,Estonia, Hungary and Poland are found to bepositively correlated with the euro area in termsof their supply shocks, but not in terms ofdemand shocks. The euro peripherals,

particularly Greece, are found to be poorlycorrelated with the euro area, while the UK andSweden exhibit positive correlation for supplyshocks but negative correlation for demandshocks, similar to Lithuania and Slovakia.

To sum up, the correlation of economicfluctuations with the euro area seems overallweaker for the acceding countries from centraland eastern Europe than for the euro pre-ins.However, this correlation is by and large notsubstantially worse than that of the peripheraleuro area economies. Moreover, at theindividual country level, the synchronisationof business cycles varies considerably. As for

52 For an earlier analysis along these methodological lines, see Fidrmuc andKorhonen (2003), as well as Bayoumi (1992) or Bayoumi andEichengreen (1993) for an application to euro area countries. Comparealso Frenkel and Nickel (2002). Results reported are updates fromFidrmuc and Korhonen (2003) and kindly provided by Iikka Korhonen,whose contribution is gratefully acknowledged. Estimations are based onquarterly y-o-y real GDP and y-o-y GDP deflator growth rates (derivedfrom nominal GDP data). Data were taken from the InternationalFinancial Statistics database and sampled over 1996 to mid-2003. Theoptimal lag length of the bi-variate VARs was almost one (occasionallytwo). Given the short size of the sample, absence of cointegrationbetween the series was not tested. It is worth stressing that the Blanchard-Quah decomposition, while being frequently used in the empiricalliterature on optimal curreny areas, was seriously challenged by Lippi andReichlin (1993, 1994). Given the long-run neutrality condition, theseauthors showed that the decomposition of a VAR’s residuals intostructural shocks is unique only if all the information available to agentsis observable by the econometrician in the data.

Chart 29 Correlation of broad cycl icaltrends with the euro area

(correlation coefficients of cyclical components estimatedby the Kalman filter, monthly data, 1996-mid-2003)


-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

IE PL SE PT SI LV UK DK HU GR SK LT CZ EE

Chart 30 Shock symmetry with the euroarea

(correlation coefficients of estimated supply shocks (x-axis)and demand shocks (y-axis) from 1996 to mid-2003)

Source: Updated from Fidrmuc and Korhonen (2003).

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

Latvia CzechRepublic

Slovenia

DenmarkPortugal

GreeceLithuania

Slovakia

UKSweden

Hungary

Poland

Estonia

51ECB



Hungary, Latvia, Poland and Slovenia, outputfluctuations seem to be rather symmetric withthe euro area, and cycle correlation is usuallywithin the range of – and sometimes higherthan – that of the euro area peripherals, suchas Portugal and Greece. Business cyclesynchronisation with the euro area is, however,sometimes estimated to be lower in the CzechRepublic and Slovakia, mainly due to currencyturbulence and stabilisation episodes in the late1990s. Lithuania always shows a negativecorrelation of economic fluctuations or ofdemand shocks with the euro area, which mightbe explained by rather different economicstructures and the relatively low degree of tradeintegration with the euro area. Results forEstonia were not robust across methods.

Looking ahead, there are reasons to believe thatbusiness cycle synchronisation between thecentral and eastern European acceding countriesand euro area countries will improve over thenext few years, owing to further integrationwith the euro area, but that this will only be agradual development. Nevertheless, differenttrends in growth rates and output fluctuation arelikely to persist in the medium to longer term.

2.3.6 ADJUSTING INTEREST RATES TO EUROAREA LEVELS

In recent years, nominal interest rates have beengradually reduced in most acceding countries,mainly reflecting the decline in inflationoutcomes and expectations. Nevertheless,policy interest rates are still well above the levelof the euro area in most acceding countries (seeChart 31). In Hungary interest rate spreadscurrently amount to 1050 basis points, inSlovakia 400 basis points, in Poland 325 basispoints and in Slovenia 275 basis points. Only inthe Czech Republic is the spread zero.

The decreasing trend in policy interest rates haslargely come to a halt since mid-2003 in mostacceding countries, and Hungary has actuallywitnessed a significant reversal, as policy

interest rates were raised by 300 basis points to9.5% in June and by another 300 basis points to12.5% in November 2003.

Given the current policy interest rate spreads,most countries would have to substantially cutinterest rates towards the euro area level withina time span of a few years, if they were to jointhe euro area a few years after EU accession.The current difference in short-term interestrates between acceding countries and the euroarea thus raises a key issue: will and shouldacceding countries have the same nominalinterest rates as the euro area and, if so, howquickly should the convergence of nominalinterest rates proceed? The obvious implicationof monetary union and trend real exchange rateappreciation suggests that, in the wake ofnominal interest rate convergence, real interestrates may fall below euro area levels. A lowlevel of capital stock, presumed high returns onnew capital and (more empirically) theoutperformance of equity markets suggest thatreal rates should however be higher.

Chart 31 Key interest rates in the CentralEuropean acceding countries

(percentages)

Source: National authorities.

0

2

4

6

8

10

12

14

16

18

20

0

2

4

6

8

10

12

14

16

18

20

Jan. July2001

Jan. July2002

Jan. July2003

PolandSlovakiaHungaryCzech RepublicSloveniaeuro area


A rapid reduction in nominal interest rates inthe run-up to euro area membership might alsobe inappropriate for some acceding countries,as inflation developments and cyclicalconditions might instead call for tightermonetary conditions. In addition, the alignmentof interest rates with those in the euro areawould very likely exert, with a time lag, furtherupward pressures on inflation. In fact, a sharpreduction in real interest rates would potentially

53 See Kröger and Redonnet (2002). However, this scenario dependscrucially on the efficiency of the interest rate channel, which may becomparatively weak in most acceding countries. In this respect,downward pressures on nominal interest rates would further have to befully passed onto other interest rates that have a bearing on domesticinvestment and consumption, such as those on domestic loans.

trigger – through lower costs of borrowing –higher investment and greater consumption inacceding countries. To the extent that self-correcting mechanisms only become effective inthe medium term, lower interest rates may leadto an overheating of the economy, in particularas inflationary pressures would result in evenlower real interest rates. This, in turn, couldeasily translate into serious imbalances insectors such as real estate and equity markets.53

A further decline in interest rates is by nowwidely expected and discounted in the bondmarkets. In November 2003, while the 1-monthmoney market yields still showed an averagespread of 380 bps, the spreads were alreadynarrower at a mean of 180 bps for 5-yearmaturity bonds (see Charts 32 and 33).

Most importantly, the implied 5-year forwardinterest rate in five years has converged to euroarea levels in the Czech Republic and Slovakia,while trading at a spread of some 100-150 bpsin Hungary and Poland (see Chart 34). Thisdifference may well reflect a credit spread or a

Chart 32 1-month money market rates

(percentages)

Chart 33 5-year bond yields

(percentages)

Sources: ECB and Bloomberg. Sources: ECB and Bloomberg.

02468

1012141618

024681012141618

Jan. July2001

Jan. July2002

Jan. July2003

euro areaCzech RepublicHungaryPolandSlovakia

Jan. July2001

Jan. July2002

Jan. July2003


02468

1012141618

024681012141618

Chart 34 Implied 5-year forward interestrate in 5 years

(percentages)

Sources: ECB, Bloomberg and ECB staff calculations.

Jan. July2001

Jan. July2002

Jan. July2003


0

2

4

6

8

10

12

0

2

4

6

8

10

12

53ECB



Chart 35 Yield curves

(percentages)

Sources: ECB and Bloomberg.

Czech Republic

0

2

4

6

8

10

12

0

2

4

6

8

10

12

3-month 6-month 12-month 5-year 10-year

July 2002January 2003July 2003November 2003

Hungary

0

2

4

6

8

10

12

0

2

4

6

8

10

12


Poland

2

3

4

5

6

7

8

9

10

11

12

2

3

4

5

6

7

8

9

10

11

12


Slovakia

0

2

4

6

8

10

12

0

2

4

6

8

10

12


premium on foreign exchange uncertainty. Inthis context it is worth noting that, since aroundmid-2003, there has been a pick-up in bondyields in several acceding countries (seeChart 35 on yield curves). This increase waspartly driven by international factors (the globalincrease in yields) but also aggravatedby regional concerns over fiscal policy, leadingto a more pronounced rise in yields (as wellas spreads compared with euro area bondyields) in some countries, seemingly inthe Czech Republic, Hungary and Poland.

Although yields in many acceding countries arestill at a low level compared with otheremerging markets, recent developments showthat the convergence of yields is not necessarilyan irreversible and smooth process, and thatexpectations may change in response tounderlying fundamentals as well as to regionaland global developments.

As regards the implications of interest rateconvergence, the experience of catching-upeconomies in the euro area may be of relevance.


For example in Ireland, monetary conditionshad already eased in the run-up to the adoptionof the euro, which resulted in strong privatesector credit growth and a tangible increase inhousing prices. Moreover, inflation increasedconsiderably, at times three percentage pointsabove the euro area average. As fiscaltightening is the only macroeconomicinstrument available in a monetary union,Ireland maintained large budget surpluses ofaround 4.7% of GDP in 2000 to dampen GDPgrowth. It is questionable whether in anacceding country such a surplus would beconceivable, given the substantial pressure forpublic investment spending and for cushioningtransition-related adverse social consequencesand unemployment (see separate sub-section onfiscal issues below).

2.3.7 EXTERNAL ADJUSTMENTS: THE IMPACT OFCATCHING-UP ON COMPETITIVENESS

Acceding countries are likely to be faced withthe challenge of maintaining competitivenessand sustainable current account deficits in thecatching-up process over the medium andlonger term. As external adjustments vianominal exchange rates are no longer possiblein a monetary union, competitiveness concernscould arise if countries were confronted with anexcessive increase in their relative price level(e.g. through inappropriately low interest ratesand resulting overheating). The ensuing lossesin competitiveness could no longer be correctedthrough devaluation and would instead have tobe corrected through a phase of lower orperhaps negative wage and price growth, whichmay be difficult owing to downwards nominalrigidities, as discussed above.

The process of catching-up is usuallyassociated with appreciating real exchangerates, following relatively high productivitygrowth. Real appreciation comes about throughhigher inflation, nominal appreciation or acombination of both. Continuous realappreciation of acceding countries’ exchangerates is not a macroeconomic problem, if

matched by productivity growth. However, anumber of acceding countries have exhibitedlarge current account deficits since the mid-1990s, even though deficits have tended to fallsomewhat in recent years. In a catching-upprocess, current account deficits may bejustified from the perspective of a rationalintertemporal transfer of resources, allowing acountry to finance higher investments, in turnjustified by higher expected returns, or tosmoothen consumption. However, large currentaccount deficits could also signal potentialcomplications in terms of deteriorating pricecompetitiveness. Furthermore, depending onthe financing and indebtedness structure, theymay also be indicative of vulnerability of thecurrency to a confidence crisis.

As regards the acceding countries, currentaccount deficits have been on averageconsiderably higher in those countries withfixed exchange rate regimes than withflexible regimes (see Charts 36 and 37). Toassess whether and, if so, to what extentcurrent account deficits reflect competitivenessproblems, or whether these imbalances aredriven by more benign factors, it is worthlooking at different real exchange ratemeasures, developments in savings andinvestment, as well as shares in export markets.In particular, it would be potentiallyproblematic if current account deficits wereaccompanied by decreasing savings rates ratherthan by high investment ratios, which in turnentail healthier prospects for long-termsustainable growth. In this context, mostacceding countries have relatively highinvestment ratios, on average around 22-24%of GDP in 2002. In the Baltic countriesinvestments have been financed, on average, bya national savings ratio of around 17% of GDP,resulting in a current account deficit of 7%of GDP. In the larger acceding countries,investments have been covered, on average, bynational savings of around 18% of GDP,resulting in a current account deficit of 4% ofGDP.

55ECB



Chart 36 CEE5 – Savings and investments

(as a percentage of GDP)

Chart 37 Baltics – Savings andinvestments


Source: European Commission. Source: European Commission.

0

5

10

15

20

25

30

1995 1996 1997 1998 1999 2000 2001 20020

5

10

15

20

25

30

current account balancegross national savingsgross fixed capital formation

0

5

10

15

20

25

30

0

5

10

15

20

25

30

1995 1996 1997 1998 1999 2000 2001 2002

current account balancegross national savingsgross fixed capital formation

Chart 39 Market share in euro area imports

(country share in total euro area imports; percentages)

Source: Eurostat.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

19972002

1 Cyprus 2 Czech Republic 3 Estonia 4 Hungary

1 2 3 54 6 7 8 9 10

5 Latvia 6 Lithuania 7 Malta 8 Poland

9 Slovakia10 Slovenia

Chart 38 Market share in world exports

(ratio of national exports to total world exports;percentages)

Source: IMF (DOTS).

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

19972002

1 Cyprus 2 Czech Republic 3 Estonia 4 Hungary

1 2 3 54 6 7 8 9 10

5 Latvia 6 Lithuania 7 Malta 8 Poland

9 Slovakia10 Slovenia

In addition, despite the persistence of relativelylarge current account deficits, accedingcountries have been able to expand theirpresence in external markets, as evidenced bythe increase in their share of total world exportsfrom 1.7% in 1997 to 2.3% in 2002. Yet, asmentioned above, this increase appears to havebeen more moderate in countries with fixedexchange rate regimes than in countriesallowing for exchange rate flexibility (seeChart 38). The evolution of acceding countries’market share in euro area imports is similar or

perhaps even more encouraging, rising from3.2% in 1997 to 4.7% in 2002 (see Chart 39).

Looking ahead, it would be desirable thatprogress in transition be in due timeaccompanied by a recovery in the contributionof external demand to GDP growth. Such arecovery was timidly observed in 1999-2000but has recently reversed (see Chart 40). Thiscan, however, be explained to a large extent byweak economic activity in the euro area. Still, itgoes without saying that current account


Chart 40 Contribution of net exports toGDP growth

(percentage points)

Source: European Commission.1) Weighted by nominal GDP in 2002.2) Projection.

-5

-4

-3

-2

-1

0

1

2

-5

-4

-3

-2

-1

0

1

2

1997 1998 1999 2000 2001 2002 20032) 20042)

Baltic StatesCzech Republic, Hungary, Poland, SlovakiaAC-10 average1)

deficits need to be carefully looked at whenassessing their medium-term sustainability,particularly so in countries with fixed exchangerate regimes.

Moreover, upon euro area entry, accedingcountries might be confronted with excessiveaggregate demand dynamics owing to decliningshort-term nominal and real interest rates,which could result in inflationary pressures andan ensuing loss of export competitivenessleading to current account imbalances. Risingprice levels could also spur wage developmentswhich, if not matched by productivity growth,would lead to higher unit labour costs in thetradable sector. In particular, upon euro areaentry, nominal wages will become more easilycomparable across Member States. Workers inacceding countries may well push for nominalwages closer to the level in core Member States,regardless of actual productivity developments.While wage pressures in countries such asPortugal and Spain were rather limited after thestart of EMU, the situation might be differentfor most acceding countries, given that they willhave common borders with high-incomecountries such as Germany and Austria.However, it should be borne in mind that thefree movement of persons may be restricted fora period of up to five (and in exceptional cases

seven) years after EU accession, with theconcrete decisions regarding this to be made atnational level.54 This may dampen pressures forwage convergence.

As a consequence of wage pressures in a settingof buoyant domestic demand spurred by low –and possibly negative – real interest rates,acceding countries could experience a loss ofcompetitiveness after joining the euro area.While participation in monetary union will easethe external financing constraint, a sustainedwidening of external imbalances may lead to abuild-up of external debt that may beunsustainable in the longer run. In this context,it is worth noting that the scope of debt-creatingfinancing of external imbalances might widenfor acceding countries in the medium term, asprivatisation revenues are set to dry up.Furthermore, foreign direct investment, whichis becoming increasingly dependent onreinvested profits (as opposed to privatisation),may be subject to global business conditionsand investor sentiment and, in some cases, anincreasing share of profits accruing from FDImay be transferred abroad in the future ratherthan being reinvested in the country concerned.

2.3.8 FISCAL CONSOLIDATION

Fiscal policy is likely to be a decisive issue forseveral acceding countries, in particular withrespect to their strategies that aim at adoptingthe euro over the medium term. Upon EUaccession, countries are subject to the EUTreaty (in particular the excessive deficitprocedure) and the Stability and Growth Pact.55

Additionally, countries have to comply with theMaastricht criteria before joining the euro area,according to which their fiscal deficit may notexceed the threshold of 3% of GDP. However,currently more than half of the accedingcountries have a fiscal deficit well above this

54 See Section 2.3.3.55 It should be noted, however, that the EU’s fiscal rules do not foresee

to impose sanctions on non-euro area Member States with excessivedeficits. The rules are less explicit in terms of the pace of fiscalconsolidation. For current euro area Member States, the target is toachieve cyclically adjusted fiscal positions close to balance by 2006.

57ECB



threshold, amounting on average to 5.1% ofGDP in 2002. Moreover, in some countries thefiscal situation has deteriorated in 2002, withHungary being the most prominent example. Ofcourse, all figures should be interpreted withcaution, given that they are not yet fullyadjusted to ESA 95 standards and are therebynot fully comparable.

Against this background, several accedingcountries would have to consolidate their fiscalbalances at a fast pace if they wanted to meet thefiscal criterion in 2005 as required for anadoption of the euro in 2007. According to the2003 PEPs, most acceding countries haverelaxed their medium-term fiscal strategiescompared with 2002’s plans, thereby taking amore gradual approach than envisaged earlier,or even postponing significant fiscalconsolidation further into the future. Still, sixcountries, namely Cyprus, Estonia, Hungary,Latvia, Lithuania and Slovenia, envisage beingin a position to meet the Maastricht fiscalcriteria by 2005, and Slovakia now intends toreduce its fiscal deficit to below 3% of GDP in2006. The remaining countries, the CzechRepublic, Poland and Malta, intend to continuerunning public sector deficits of more than 3%of GDP throughout 2006.

The relaxation of medium-term fiscal strategiesas laid out in the 2003 PEPs raises questionsnot only concerning the credibility of theseplans but also as to whether consolidationefforts are sufficiently ambitious in allcountries concerned. Weak fiscal track recordsin some countries add further concerns. The2003 PEPs show that fiscal slippages arecontinuing in a number of countries and thatoverruns in 2003 may be considerable in a fewcountry cases. This pertains mainly to Cyprus,Malta and the Czech Republic, where fiscalslippages are in the order of 3.5 percentagepoints (ppts), 3.1 ppts and 1.8 ppts of GDPrespectively. Slovenia also expects an overrunof 0.6 ppt of GDP, which however still impliesa relatively contained deficit of 2% of GDP andan improvement of 0.6 ppt compared with 2002.Hungary will also have a budget overrun;

however, its size is unclear, with thegovernment expecting a slippage of 0.3 ppt ofGDP, while Magyar Nemzeti Bank foresees anoverrun of more that 1 ppt of GDP. In itsAutumn 2003 forecasts, the EuropeanCommission expects slippages of about 0.9 pptof GDP on average for all acceding countries.It is worth noting that overruns in 2003 arepartly due to overly optimistic underlyingassumptions, and partly a consequence of moreexpansionary fiscal policy stances than laid outin the budget frameworks.

Against this backdrop, there is clearly noground for concerns that fiscal consolidationmight entail excessive short-term costs in termsof output losses and temporarily slow realconvergence with the euro area. While it is truethat, in the medium term, fiscal consolidationmight have a positive impact on growth, theshort-term implications of a fast consolidationon growth are uncertain and could be negative.However, at the current stage, only Hungaryand Slovakia are planning substantial cuts infiscal deficits over the next few years, and theirambitious programmes are still facing the test ofimplementation. An improving global and inparticular European growth environment in2004 and beyond could be seen as conducive tostepping up fiscal consolidation efforts inacceding countries, as increasing externaldemand would help set off public consumptionand investment demand. Still, a rapid fiscalconsolidation strategy may face design andimplementation challenges, as accedingcountries are confronted – at the current stageand in the near future – with many competingdemands on spending and, in some cases, withpressure to cut direct taxes.

In the coming years, acceding countries will beexposed to continuing expenditure pressuresarising from the completion of the transitionprocess, the implementation of the acquiscommunautaire (especially in the area of theenvironmental acquis, where spending needswill remain particularly high for a number ofyears), the budgetary requirements relatedto EU and NATO membership, and pending


reforms of the health and pension systems.Against this backdrop, public investment isexpected to pick up in a number of countries. Atthe same time, prospective windfall gains frominterest rate convergence appear to be limited;even for countries with intermediate levels ofpublic debt, future savings on interest paymentswill be much smaller than in the case of severalof the EU countries which joined the euro areain 1999. Moreover, given the high tax burden insome countries, significant tax reforms areplanned, which would have a bearing ongovernment revenues and fiscal deficits as well.

It should additionally be noted that fiscaldeficits in acceding countries with major fiscalimbalances seem to be mainly of a structuralnature. This is also indicated by the high shareof mandatory expenditures and by the limitedrole of automatic stabilisers in most countries.Consequently, a recovery in growth, asexpected in most acceding countries from 2003onwards, would not contribute greatly to areduction in fiscal imbalances. Instead, giventhe structural nature of fiscal deficits,substantial budgetary reforms appear to beneeded to achieve ambitious consolidationtargets (including greater fiscal transparency,increased efficiency and less mandatoryexpenditures). The implementation ofconsolidation plans, however, may be furthercomplicated owing to strong social preferencesin central European acceding countries forretaining the welfare state. On the other hand,there is in most cases no room to undertake cutsin public investment, given the challengesoutlined.

At a country level, fiscal imbalances are mostprofound in the Czech Republic, Hungary,Malta, Poland and Slovakia, with deficits wellabove the 3% of GDP threshold (see Chart 41).Hungary has the highest deficit, which stood at9.2% of GDP in 2002 after almost doublingfrom 4.7% of GDP in 2001. Moreover, itshould be noted that in some of the otheracceding countries, fiscal deficits are presentlynot far below or are at the 3% threshold, namelyin Latvia, Lithuania and Slovenia. The fiscal

accounts of Latvia and Lithuania are moreoverprofiting, to some extent, from strong growthperformance. While fiscal challenges areparticularly pronounced for countries that arecurrently registering high imbalances, it shouldnot be overlooked that containing deficits willbe a demanding task in the other accedingcountries as well, given the expenditurepressures and the need to reduce high taxburdens. Temporary increases in budget deficitsto accommodate such demands may easily turninto more permanent imbalances, unless strongand coherent medium-term fiscal frameworksare in place.

Concerns that targeting a fast adoption of theeuro may lead some acceding countries tofollow a sub-optimal path of fiscalconsolidation are related to a scenario in whichfiscal consolidation in high-deficit countrieswould stall while current target dates for euroadoption would be retained, so that a massivefiscal correction would have to be undertaken ina short period of time a few years down the line.In this context, it should also be borne in mindthat upon EU accession, new entrants shouldbring their deficits well below the 3% threshold

Chart 41 General government budgettargets 2002-2006


Source: 2003 Pre-accession Economic Programmes.Note: Estonia targets balanced budgets from 2004 onwards.

-10

-8

-6

-4

-2

0

2

-10

-8

-6

-4

-2

0

2

CZ MT CY SK HU PL LV LT SI EE

20022003200420052006

59ECB



and move towards cyclically adjusted fiscalpositions close to balance in the medium term toincrease the room for manoeuvre in terms offiscal policy. Only under such circumstanceswould fiscal policy become an importantadjustment tool for smoothing the businesscycle after joining the euro area. This would beparticularly relevant for catching-up economies,given the potential for pronounced cyclicalfluctuations in these countries.


This paper has reviewed the exchange rate andmonetary integration strategies announced byacceding countries for the period immediatelyafter EU entry, with a special focus onparticipation in ERM II, and on their intentionsconcerning the adoption of the euro at a laterstage. As this study aims at exploring keyissues at stake, outside the framework of anysurveillance procedure and convergenceexamination, this paper has benefited fromseveral different approaches, including a shortreview of nominal convergence and a moreextensive optimum currency area perspective.An important aspect of the analysis has beendevoted to the implications of real convergence– i.e. catching-up growth in income andadjustment of the real economic structurestowards those prevailing in the euro area – onthe patterns of economic dynamics in accedingcountries. The fact that per capita income levelsare substantially below those in the euro area,and that some segments of the structure of thereal sectors are still affected by the transitionprocess, are among the key economiccharacteristics of acceding countries. Whiledifferent income levels can be, in principle,compatible within a monetary union, such realconvergence may imply differences in economicdynamics – including the level of growth andthe magnitude of fluctuations – that could makea single monetary policy inappropriate for somecountries. Other aspects covered are the risksfor external competitiveness in the convergenceprocess and the appropriate pace of fiscalconsolidation.

Notwithstanding the preliminary nature of theanalysis, the complexity of the underlying issueand the fact that theory is in many aspectsinconclusive with regard to optimum currencyareas and monetary integration, the followingtentative findings emerge from a comprehensiveinvestigation based on a wide set of economicindicators.

The acceding countries display somecommonalities with the euro area that may bodewell for future monetary integration. In terms of

broad sectors, the economic structures ofacceding countries have become similar to thoseof the euro area. Furthermore, the degree ofopenness is high, and trade as well as financialintegration with the EU is well advanced inmost cases. At the same time, the paper findssignificant differences among the accedingcountries in terms of a range of nominal, realand structural conditions. This pertains, inparticular, to labour market features, policyinterest rates, external positions and fiscalperformance. Moreover, the degree ofintegration and cyclical harmonisation with theeuro area also differs considerably fromcountry to country. Against this background,the paper supports the view that the economicmerits of any particular strategy towards ERMII and the later adoption of the euro will need acareful case-by-case assessment.

For some acceding countries, maintainingexchange rate flexibility for some time after EUaccession may be important from a cyclicalstabilisation viewpoint. In such countries,retaining exchange rate flexibility can make avaluable contribution to smoothening outputvolatility, especially if GDP developmentsdisplay substantial fluctuations, as tends to bethe case for countries that have embarked on adynamic catching-up process. It should benoted that the standard fluctuation band ofERM II would seem to give considerable roomfor exchange rate variability and, thus, for a useof the exchange rate as a tool for cyclicalstabilisation.

Whether it is preferable to maintain a degree ofexchange rate flexibility within ERM II oroutside this mechanism depends on a number offactors and thus on the specific situation ofindividual countries. Two important aspects inthis context are the monetary and exchange rateframework in place and the fiscal performance.For instance, countries that currently rely on amonetary policy strategy based on a domesticanchor would need to switch to a combination ofa domestic and external anchor when joiningERM II, if they choose to maintain such domestic

CONC LUD I NG R EMARK S

61ECB


CONCLUDINGREMARKS

anchors. If and where such frameworks workwell, there may be good reasons to retain themfor some time after EU entry.

A more gradual approach towards ERM IIparticipation pertains to countries that arecurrently facing high fiscal imbalances. In thesecases, achieving a critical mass of fiscalconsolidation and putting a credible medium-term fiscal strategy in place should precedeERM II entry in order to promote a smoothparticipation in the mechanism. Furthermore,choosing the optimal path of fiscalconsolidation – in particular in a post-transitionand EU accession context – may not be fullyconsistent with the targeting of a fast adoptionof the euro.

Further monetary integration shall be facilitatedin countries where fiscal deficits and publicdebt are limited, stability-oriented policies arefully maintained and further structural policiesare implemented appropriately in order tosupport the specific economic setting. Labourmarket flexibility also remains crucial to adjustto possible differences in economic conditions.For acceding countries where these conditionsare in place, ERM II participation may turn outto be comparatively limited in time, providedthat the consistency of domestic macroeconomicpolicies and the sustainability of convergence ingeneral are fully preserved.

Nevertheless, in this context, several importantcaveats have to be made. Even if, for a givencountry, prima facie evidence at the currentjuncture would seem to support the case formaintaining the current exchange rate regime, itwould be premature to draw conclusions aboutthe country’s readiness for membership in ERMII and for a subsequent adoption of the euro.Moreover, the choice of the central paritywithin ERM II is a key issue, as decisions inthis regard will be taken by mutual agreement ofthe participating members in ERM II, includingthe ECB. In addition, even if only minorexchange rate regime changes are required insome cases for participation in ERM II, apotential misalignment in these countries is a

risk that cannot be ignored. More generally, theabsence of significant foreign exchange marketpressure in the past cannot be taken as implyingan absence of such pressure in the future.

These analytical findings are very preliminary,as the paper offers above all a conceptualframework designed to review the accedingcountries’ strategies towards ERM II and theadoption of the euro. This does not thereforepreclude that the analysis will evolve, as it isrefined over time and takes account of neweconomic developments and changing policystances in the acceding countries.


Balcerowicz, L. (2002): “The Way to EMU from a Candidate Country’s Perspective”, in: Latschet,A. and F. Plüger, Der Euro und Europa.

Bayoumi, T. (1992): “The Effects of the ERM on Participating Economies”, IMF Staff Paper,No. 39.

Bayoumi, T. and B. Eichengreen (1993): “Shocking Aspects of European Monetary Integration”,in Torres, F. and F. Gavazzi, Growth and Adjustment in the European Monetary Union ,pp. 193-230.

Blanchard, O. and D. Quah (1989): “The Dynamic Effects of Aggregate Demand and SupplyDisturbances”, American Economic Review, 79, 4, pp. 665-73.

Buiter, W. and C. Grafe (2002): “Anchor, Float or Abandon Ship: Exchange Rate Regimes forAccession Countries”, CEPR Discussion Paper, No. 3184.

Cazes, S. (2002): “Do Labour Market Institutions Matter in Transition Economies? An Analysis ofLabour Market Flexibility in the Late Nineties”, Discussion paper No. 140, InternationalInstitute for Labour Studies (ILO).

Centre for European Policy Studies (2002): “The Euro at 25: Special Report of the CEPSMacroeconomic Policy Group”, Brussels.

Corden, W. (1972): “Monetary Integration”, Essays in International Finance, No. 93, PrincetonUniversity.

Coricelli, F. (2002): “Exchange Rate Arrangements in Transition to EMU: Some Arguments inFavour of Early Adoption of the Euro”, in Tumpel-Gugerell, G., L. Wolfe and P. Mooslechner(eds): Completing Transition: The Main Challenges, pp. 203-14.

Czech National Bank (2003a): “The Czech Republic and the Euro-Draft Accession Strategy”,January, http//www.cnb.cz.

Czech National Bank (2003b): “The Czech Republic’s Euro Area Accession Strategy”, Jointdocument of the Czech Government and the Czech National Bank, 13 October 2003, http://www.cnb.cz/en/pdf/en_eurostrategie_09_2003.pdf.

Darvas, Z. and G. Szápany (2004): “Business Cycles Synchronization in the Enlarged EU:Comovements in the New and Old Members”, MNB Working Paper, February.

De Grauwe, P. and Y. Aksoy (1999): “Are Central European Countries Part of the EuropeanOptimum Currency Area?”, in: De Grauwe, P. and V. Lavrac (eds): Inclusion of CentralEuropean Countries in the European Monetary Union, pp. 13-36.

Eichengreen, B. (2003): Interview in Handelsblatt, 13 January.

EIRO/ILO (2002): “Industrial Relations in the Candidate Countries”, July.

ETUC (European Trade Union Confederation) (2002): “Wage Formation in the Central and EasternEuropean Candidate Countries”, April.

Eurostat (2003): “Minimum Wages, EU Member States and Candidate Countries, January 2003”,Statistics in Focus, Theme 3, 10/2003.

Fidrmuc, J. (2002): “Migration and Regional Adjustment to Asymmetric Shocks in TransitionEconomies”, ZEI, University of Bonn, mimeo.

Fidrmuc, J. and I. Korhonen (2003): “Similarity of supply and demand shocks between the euroarea and the CEECs”, Economic Systems, 27, pp. 313-334.

RE F ERENCE S

63ECB


REFERENCES

Fidrmuc, J. and F. Schardax (2000): “More ‘Pre-ins’ Ante Portas?”, Focus on Transition, 5, 2,pp. 28-47.

Frankel, J. and A. Rose (1998): “The Endogeneity of the Optimum Currency Area Criteria”,Economic Journal, 108, pp. 1009-1025.

Frankel, J. and A. Rose (2000): “Estimating the Effect of Currency Unions on Trade and Output”,CEPR Discussion Paper, No. 2631.

Frenkel, M. and C. Nickel (2002): “How Symmetric are the Shocks and the Shock AdjustmentDynamics between the Euro Area and Central and Eastern European Countries?”, IMF WorkingPaper, No. 222, December.

Friedman, M. (1953): Essays in Positive Economics, University of Chicago.

Gros, D. (2000): “One Euro from the Atlantic to the Urals?”, CESifo Forum, No. 2.

Honohan, P. (2001): “Discussion”, Economic Policy, 33, pp. 457-61.

Ingram, J. (1973): “The Case for European Monetary Integration”, Princeton University, Essays inInternational Finance, No. 98.

Kenen, P. (1969): “The Optimum Currency Area: An Eclectic View”, in: Mundell, R. and A.Swoboda (eds): Monetary Problems of the International Economy, Chicago.

Kröger, J. and D. Redonnet (2002): “Exchange Rate Regimes and Economic Integration: The Caseof the Accession Countries”, CESifo Forum, No. 1, pp. 6-13.

Krugman, P. (1993): “Lessons of Massachusetts for EMU”, in: Torres, F. and F. Giavazzi (eds):Adjustment and Growth in the European Monetary Union, pp. 241-69.

Lippi, M. and L. Reichlin (1993): “The Dynamic Effects of Aggregate Demand and SupplyDisturbances: Comment”, American Economic Review, 83, pp. 644-652.

Lippi, M. and L. Reichlin (1994): “VAR Analysis, Nonfundamental Representations, BlaschkeMatrices”, Journal of Econometrics, 63, pp. 307-325.

Magyar Nemzeti Bank (2002): “Adopting the euro in Hungary: Expected Costs, Benefits andTiming”, NBH Occasional Paper 24, A. Csajbók and A. Csermely (Eds).

Maurel, M. (2002): “On the Way to EMU Enlargement towards CEECs: What Is the AppropriateExchange Rate Regime?”, CEPR Discussion Paper, No. 3409.

McKinnon, R. (1963): “Optimum Currency Areas”, American Economic Review, 52, pp. 717-25.

Melitz, J. (2001): “Geography, Trade and Currency Union”, CEPR Discussion Paper, No. 2987.

Mongelli, F. (2002): “New Views on the Optimum Currency Area Theory: What Is EMU TellingUs?”, ECB Working Paper No. 138.

Mundell, R. (1961): “A Theory of Optimum Currency Areas”, American Economic Review, 51,pp. 657-65.

Mundell, R. (1973): “Uncommon Arguments for Common Currencies”, in: Johnson, H. and A.Swoboda (eds): The Economics of Common Currencies, Allen and Unwin, St. Leonards.

Mundell, R. (2002): “Exchange Rate Arrangements in Central and Eastern Europe”, in: AustrianMinistry of Economic Affairs: Eastern Enlargement: The Sooner, the Better?, Vienna.


Nitsch, V. (2002): “Honey, I Shrunk the Currency Union Effect on Trade”, World Economy,25, 4, pp. 457-74.

OECD (1999): “Employment Protection and Labour Market Performance”, Employment Outlook,Chapter 2, July.

OECD (2003): “Labour Market and Social Policies in the Baltic Countries”, April.

Persson, T. (2001): “Currency Unions and Trade: How Large Is the Treatment Effect?”, EconomicPolicy, 33, October, pp. 435-461.

Riboud, M., C. Sánchez-Páramo and C. Silva-Jáuregui (2002): “Does Eurosclerosis Matter?Institutional Reform and Labor Market Performance in Central and Eastern European Countriesin the 1990s”, Social Protection Discussion Paper Series, World Bank, March.

Rose, A. (2000): “One Money, One Market: The Effect of Common Currencies on Trade”,Economic Policy, 30, April, pp. 9-45.

Rose, A. K. (2001): “Currency Unions and Trade: The Effect Is Large”, Economic Policy, 33,pp. 449-61.

Rose, A. and E. van Wincoop (2001): “National Money as a Barrier to Trade: The Real Case forCurrency Union”, American Economic Review (Papers and Proceedings), 91, 2, pp. 386-90.

Scitovsky, T. (1966): “The Theory of Balance-of-Payments Adjustment”, Journal of PoliticalEconomy, 75, 4 (2), pp. 523-31.

Stock J. and M. Watson (1991): “A Probability Model of the Coincident Economic Indicators”, inK. Lahiri and G. Moore (eds.), “Leading Economic Indicators: New Approaches andForecasting Records”, pp. 63-89.

Süppel, R. (2003): “Economic Dynamics in EU Accession Countries: A Case for Exchange RateFlexibility?”, mimeo.

Szapáry, G. (2002): “Is Maastricht too Much for the Candidate Countries?”, mimeo.

Tavlas, G. (1994): “The Theory of Monetary Integration”, Open Economies Review, 5, 2, pp. 211-30.

UNECE (2002): Economic Survey of Europe, United Nations, 1.

65ECB


EUROPEAN CENTRAL BANKOCCASIONAL PAPER SERIES

1 “The impact of the euro on money and bond markets” by J. Santillán, M. Bayle andC. Thygesen, July 2000.

2 “The effective exchange rates of the euro” by L. Buldorini, S. Makrydakis and C. Thimann,February 2002.

3 “Estimating the trend of M3 income velocity underlying the reference value for monetarygrowth” by C. Brand, D. Gerdesmeier and B. Roffia, May 2002.

4 “Labour force developments in the euro area since the 1980s” by V. Genre andR. Gómez-Salvador, July 2002.

5 “The evolution of clearing and central counterparty services for exchange-tradedderivatives in the United States and Europe: a comparison” by D. Russo,T. L. Hart and A. Schönenberger, September 2002.

6 “Banking integration in the euro area” by I. Cabral, F. Dierick and J. Vesala,December 2002.

7 “Economic relations with regions neighbouring the euro area in the ‘Euro Time Zone’” byF. Mazzaferro, A. Mehl, M. Sturm, C. Thimann and A. Winkler, December 2002.

8 “An introduction to the ECB’s survey of professional forecasters” by J. A. Garcia,September 2003.

9 “Fiscal adjustment in 1991-2002: stylised facts and policy implications” by M. G. Briotti,February 2004.

10 “The acceding countries’ strategies towards ERM II and the adoption of the euro:an analytical review” by a staff team led by P. Backé and C. Thimann and includingO. Arratibel, O. Calvo-Gonzalez, A. Mehl and C. Nerlich, February 2004.

Date post:	21-Jun-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

The acceding countries' strategies towards ERM II …In 2004 all ECB publications will feature a...

Documents