external evaluation of imf surveillance · 2000. 11. 22. · part 2 external evaluation of...

ExternalEvaluation of

IMFSurveillance

Report by a Group ofIndependent Experts

International Monetary Fund

© 1999 International Monetary Fund

Design and composition: Alicia Etchebarne-BourdinCover design and charts: IMF Graphics Section

Librar y of Congress Cataloging-in-Publication Data

External evaluation of IMF surveillance : report / by a group of independent experts.p. cm.

Includes bibliographical references.ISBN 1-55775-863-8 (pbk.)1. International Monetary Fund. 2. Economic policy—Evaluation. I. International

Monetary Fund.

HG3881.5.I58 E951999332.1'52—dc21 99-088147

Price: US$19.00

Please send orders to:International Monetary Fund, Publication Services

700 19th Street, N.W., Washington, D.C. 20431, U.S.A.Tel.: (202) 623-7430 Telefax: (202) 623-7201

E-mail: [email protected]: http://www.imf.org

Glossary of Abbreviations iv

Part 1 Statement by Thomas Bernes, Chairman of the 3Evaluation Group of Executive Directors,on the Report of the External Evaluation of Surveillance, August 27, 1999

Summing Up by the Chairman of the Executive Board 4of Executive Board Meeting 99/100, on the External Evaluation of Fund Surveillance,September 8, 1999

Part 2 External Evaluation of Surveillance Report 7

Part 3 Statement by the Managing Director on the Report of 95the External Evaluators on Fund Surveillance,Executive Board Meeting, August 27, 1999

Staff Response to the External Evaluation of Fund 98Surveillance

Contents

iii

The following symbols have been used throughout this book:

. . . to indicate that data are not available;

— to indicate that the figure is zero or less than half the final digit shown, or thatthe item does not exist;

– between years or months (e.g., 1996–97 or January–June) to indicate the yearsor months covered, including the beginning and ending years or months;

/ between years (e.g., 1996/97) to indicate a crop or fiscal (financial) year.

“Billion” means a thousand million.

Minor discrepancies between constituent figures and totals are due to rounding.

The term “country,” as used in this book, does not in all cases refer to a territo-rial entity that is a state as understood by international law and practice; theterm also covers some territorial entities that are not states, but for which sta-tistical data are maintained and provided internationally on a separate and in-dependent basis.

APEC Asia-Pacific Economic CooperationASEAN Association of South East Asian NationsBIS Bank for International SettlementsCCL Contingent Credit LineCGER Coordinating Group on Exchange RatesCIS Commonwealth of Independent StatesECB European Central BankECOFIN European Council of Finance MinistersED Executive DirectorEMU Economic and Monetary UnionERM exchange rate mechanismEU European UnionEWS early warning systemFSSA Financial Sector Stability AssessmentG-7 Group of SevenG-22 Group of 22GCC Gulf Cooperation CouncilICMR International Capital Markets reportIIF Institute of International FinanceILO International Labor OrganizationIOSCO International Organization of Securities CommissionsMAE Monetary and Exchange Affairs DepartmentNGO nongovernmental organizationOECD Organization for Economic Cooperation and DevelopmentPDR Policy Development and Review DepartmentPIN Press/Public Information Notice RED Recent Economic Developments reportRES Research DepartmentWEMD World Economic and Market DevelopmentsWEO World Economic Outlook

Glossary of Abbreviations

iv

Part 1

Statement by Thomas Bernes, Chairman of the Evaluation Group of

Executive Directors, on the Report of the External Evaluation of Surveillance

Summing Up by the Chairman of the Executive Board of

Executive Board Meeting 99/100, on the External Evaluation of Fund Surveillance

I t is to be expected that any comprehensive evalua-tion of the Fund’s surveillance activities will touch

upon an impressive range of issues. In this regard, thereport of the external evaluators does not disappoint.On behalf of the Evaluation Committee, I wouldtherefore like to express my sincere appreciation toMessrs. Arriazu, Crow, and Thygesen for an out-standing effort that has resulted in the production of awell-written, insightful, and thought-provoking doc-ument. It is the mark of a good evaluation that ques-tions about the prevailing wisdom are raised and thestatus quo challenged. Indeed, were the evaluators tohave concluded that all was well and good at theFund, I would have had reservations as to whether ornot our money had been well spent. This is clearlynot the case here. That is not to say that the institutioncould—or should—adopt all recommended changes,but that we should seize on the opportunity providedby the evaluators’report to reassess long-standingpractices and the adequacy of institutional and proce-dural reforms presently under way.

It is worth noting that—with a topic as broad asthat addressed by the evaluators—there will in-evitably be exceptions to the conclusions drawn,even while the broad thrust of generalized observa-tions remains valid. This suggests that the relevanceof particular recommendations will vary by country,within the Fund’s institutional structure, and acrossstaff. Therefore, to make the most constructive useof this evaluation, we need to approach its observa-

tions with a considerable degree of honesty andopenness to criticism, drawing on the insights whereappropriate. The flip side of this is that we need toavoid excessive fixation on individual instanceswhere the criticisms may not be as directly relevant.

In terms of next steps, and consistent with existingprecedents, it would seem appropriate that the text ofthe report should be published along with the staff ’sresponse and the summing up from this discussion.

In the period after the Annual Meetings, and in lightof both the report and the Board discussion, I wouldsuggest that management prepare an “action plan” forconsideration by Directors on proposals to respond tothe report. A date within 12 months of the approval ofthis plan should be set for a “stocktaking” of our re-sponse. As with the external evaluation of ESAF, I be-lieve it is critical to provide some formalization of thefollow-up process to our evaluations.

More generally, we will then need to turn our at-tention—before the end of the year—to launching areview of our experience with independent evalua-tion. The terms of reference and modalities for sucha review should be discussed shortly after the An-nual Meetings. At the same time, I believe it wouldbe appropriate to undertake a further external evalu-ation—perhaps more modest in scope than the sur-veillance evaluation—since it would not be appro-priate or desirable for this institution to cease theconduct of independent evaluations while we assessour experience to date.

Statement by Thomas Bernes,Chairman of the Evaluation Group ofExecutive Directors, on the Report of theExternal Evaluation of Surveillance

August 27, 1999

3

E xecutive Directors welcomed the Report of Ex-ternal Evaluators on Fund Surveillance. They ex-

pressed their deep appreciation for the careful workand considered judgments of the panel. Directors con-sidered that the issues raised in the report would serveto stimulate debate within and outside the institution,and to motivate further discussion of a number of top-ics of importance to the work of the Fund.

Directors welcomed the comprehensive evalua-tion of the Fund’s surveillance work and the highconsideration in which the World Economic Outlook(WEO) and the International Capital Markets re-ports (ICMR) are held. They noted the value thatmembers place on the Fund’s surveillance of theirown economies. In this regard, the evaluators’obser-vation that Fund surveillance should be viewed as an“input” to a country’s policies—that could on occa-sion be significant—underscored the importance ofthe Fund’s analysis being first rate and of remainingfocused on issues of serious and immediate concern.

Directors underlined the substantial commonground between the evaluators’report and theFund’s own internal evaluations. They noted, in par-ticular, the need to (i) revisit the definition of thecore areas; (ii) give more explicit attention to inter-national aspects of a country’s macroeconomic poli-cies and spillover issues; (iii) focus more on cross-country comparisons and regional developments;(iv) devote substantially more attention to vulnera-bility analysis; and (v) give more emphasis to finan-cial sector and capital account issues.

On the focus of surveillance, Directors acknowl-edged that the issue remains a challenge for the or-ganization in light of the forces driving an expand-ing agenda. In this regard, they acknowledged thatissues identified in Fund surveillance as core hadchanged over time, moving from a narrow focus onexchange rate policy and the balance of paymentsand attendant monetary and fiscal policies to greater

emphasis on capital account, financial sector, andnonfinancial structural issues. Some Directorsstressed that the report was written at a time whenthe focus of surveillance was undergoing majorchanges within the Fund, largely as a result of thecrises of the last five years. Thus, as noted in theevaluation, some of the recommendations on thefocus of surveillance, including those relating to en-hancing surveillance of the financial sector, capitalaccount issues, and policy interdependence and con-tagion, are in the process of being implemented.

Nevertheless, Directors expressed a range of viewson which issues should be considered generally asnoncore issues in the context of Fund surveillance.Most Directors thought that one of the main recom-mendations of the report—that surveillance shouldfocus only on the core areas of exchange rate policyand directly associated macroeconomic policies—rancounter to the demands of the membership and theinternational community for increasing emphasis onthe interactions between macroeconomic, structural,and social policies. They viewed the broader focus ofsurveillance as appropriate in light of global develop-ments and the need for a surveillance process that re-mains relevant to the policy challenges faced byFund members. Nevertheless, a number of these Di-rectors saw scope for sharpening the focus of surveil-lance in the context of a country-by-country ap-proach: coverage of issues could differ depending onthe circumstances of a particular country, but the staffwould be expected to present a clear case for the con-sideration of any particular set of noncore issues asrelevant to the core concerns of the Fund. Other Di-rectors, however, felt that Fund surveillance hadmoved inappropriately beyond the original core is-sues, including into areas such as labor markets, pen-sion reform, social policy, and governance. Neverthe-less, most Directors agreed that, to the extentpossible, the Fund should make use of outside exper-

4

Summing Up by the Chairman of the Executive Board of Executive Board Meeting 99/100, on theExternal Evaluation of Fund Surveillance

September 8, 1999

Summing Up by the Chairman of the Executive Board

5

tise in areas beyond its conventional mandate andwhen it had little in-house expertise. In this regard,Directors stressed the importance of close coopera-tion with other international institutions, taking dueaccount of comparative advantage and expertise, andavoiding duplication of effort. Directors agreed to re-turn to the issue of the focus of surveillance moresystematically in the context of the internal review ofFund surveillance later in the year.

Directors supported increased attention to the in-ternational and regional aspects of surveillance.They saw the need for increased cross-countrycomparisons—in which the Fund has a unique advantage—in the context of greater emphasis onregional and international developments. They alsoendorsed the recommendation to heighten the inter-action between bilateral and multilateral surveil-lance, and looked forward to a better integration ofthe ICMR and WEO analysis with bilateral surveil-lance. However, Directors agreed that, while theArticle IV process should be enriched through theintegration of global and regional dimensions, itshould remain clearly focused on a country’s ownpolicies.

Directors emphasized their support for more ex-plicit attention to vulnerability issues in Fund surveil-lance: this would entail enhanced analysis of the capi-tal account, the financial sector, and the treatment ofcontagion. They agreed that, in an environment of in-creased financial and trade flows between countries,Fund surveillance at the country level should paygreater attention to the sequencing and the pace ofmoves toward capital account liberalization. Directorsobserved that the stepped-up level of staff work on fi-nancial sector issues in collaboration with the WorldBank (including through the Financial Sector Assess-ments Program), the Bank for International Settle-ments, and other international organizations wasbeing reflected in more comprehensive coverage ofvulnerabilities in this area. Directors agreed that sur-veillance should pay greater attention to policy inter-dependence and the risks of contagion, and they notedin this regard that multilateral surveillance has an im-portant role to play in identifying potential spillovereffects.

On surveillance procedures, Directors observedthat the strength of the Fund as an institution derivesfrom the symmetry with which countries are treatedin the surveillance process. Many Directors saw an-nual consultations as a cornerstone in ensuring thecontinuity of Fund surveillance. At the same time,Directors recognized the need for some flexibility inFund procedures given the institution’s strained re-sources. Against this backdrop, Directors agreedthat, for most industrial economies, in light of theirsystemic impact, annual consultations remain appro-priate. Most Directors thought that surveillance of

these countries should continue to focus on their do-mestic policies while also bringing to the table theinternational implications of those policies.

To ensure resource-efficient surveillance that, atthe same time, could be more continuous, some Di-rectors thought it should be possible in some cases tohave shorter annual consultation visits supplementedwith interim electronic communications. However,other Directors thought it important that this not leadto any diminution of the attention paid by nationalauthorities to the formal consultation discussion. Di-rectors asked the staff to come forward with propos-als in this area.

Most Directors viewed annual consultations withsmaller industrial countries—particularly those thatare members of the euro area—as providing a num-ber of critical advantages that could be lost with lessfrequent consultations, although a few Directors sawpotential advantages in reducing consultation fre-quency. Several Directors also pointed out that, in thecase of the euro area countries, fiscal policy remaineda national prerogative and many other policies con-tinued to be conducted at the national level; it wouldthus be impossible to cover these areas adequately inconsultation with the European Central Bank or Eu-ropean Union institutions. While several Directorsconsidered that there might be scope to reduce thesize and duration of missions to these countries asEuropean integration proceeds, other Directors werenot in favor of diminished attention to the euro area.

Directors were also not in favor of shifting sole re-sponsibility for the WEO projections to the ResearchDepartment. They felt that the area departments pro-vided critical input to these projections and that theprojections should remain a joint product of thestaff. They noted that the current practice of six-monthly WEO reports, with special issues whenwarranted by circumstances, had worked well andshould continue.

On organization, Directors considered with inter-est the proposal that all Article IV staff reports bediscussed in the first instance in a committee ratherthan by the full Board, but they thought this likely toresult in an increased work load for the organizationas a whole. In this regard, most Directors thought itdesirable to continue with the lead speaker experi-ment recently undertaken as a step toward more effi -cient and focused Board coverage of surveillance.Directors welcomed the observations of the evalua-tors on the internal review process. They urged man-agement and staff to pursue greater efficiencies inthis area, while preserving the valuable contributionof the process to the high quality of analysis.

Directors noted that the transparency of the Fundhad increased considerably in recent years and that apilot project for the voluntary release of staff reportshad been launched. They agreed that the review of the

SUMMING UP BY THE CHAIRMAN OF THE EXECUTIVE BOARD

pilot project would inform the development of a gen-eral publication policy for Article IV staff reports.

Looking ahead, Directors stressed that thestrengthening of Fund surveillance has been, and is,an ongoing process. In this regard, the report of theevaluators provides an informed outside perspectivethat should serve as an important input in delibera-tions on enhancing surveillance. Directors lookedforward in the period ahead to further considerationof many of the issues addressed in the evaluators’re-port. Directors considered that the key issues to re-

turn to later this year could include the focus of sur-veillance; the increased attention to international, re-gional, and cross-country issues; vulnerability analy-sis and early warning systems; and the coverage offinancial sector and capital account issues.

Management intends to come back to the Boardafter the Annual Meetings with precise suggestionson a program to deal with the issues raised by theExternal Evaluation Report. These issues will alsobe followed up in more detail in the Biennial Reviewof Surveillance scheduled for end-1999.

6

External Evaluation of Surveillance Report

Members of the Evaluation Team:

John Crow (Chairman) Ricardo ArriazuNiels Thygesen

Secretary to the Team: Jonathan Portes

Part 2

Preface 11

Executive Summary 13

I Introduction 17Objectives of Surveillance 17Origins and Development of Fund Surveillance 19

II Conduct and Methods 24Bilateral Surveillance 24Regional Surveillance 28Multilateral Surveillance 31Fund Internal Organization and Procedures 31Role of the Executive Board 33

III Substance 35Bilateral Surveillance 35Multilateral Surveillance 43Other Sources of Surveillance 46

IV Impact 48Impact by Country Groupings 48Other Elements Shaping Impact 49Impact in Four Countries 51Internal Reviews 53Transparency, Publicity, Communication 53Communication Within Government 57

V Conclusions and Recommendations 59Objectives and Priorities 60Substance 65Organization 69Communication 75Recent Developments 78

AppendicesI Terms of Reference 82II Articles of Agreement 84III List of Interviewees 86IV Recommendations 90V Confidential Exchange: An Elaboration 92

Boxes2.1 Chronology of an Article IV Consultation 252.2 Surveillance of the Euro Area 303.1 Early Warning Systems 423.2 Surveillance, Capital Flows, and Financial Crises 455.1 Possible Board Committee Structure for Surveillance 76

Contents

9

T his evaluation of IMF surveillance over members’policies under Article IV ofthe Articles of Agreement was commissioned by the Executive Board in June

1998. Our terms of reference are in Appendix I. The evaluation was conducted by John Crow (formerly Governor of the Bank of

Canada), who served as Chairman; Ricardo Arriazu (Bansud, Buenos Aires, and for-merly Alternate Executive Director at the Fund), and Niels Thygesen (Danske BankProfessor of International Economics at the University of Copenhagen). JonathanPortes (formerly United Kingdom Treasury) served as Secretary to the team.

In July 1998, the group held a seminar with the Executive Board, where we set outour general approach and presented our list of sample countries pursuant to the terms ofreference: Brazil, Chile, China, the Czech Republic, India, Japan, Korea, Saudi Arabia,South Africa, Sweden, Thailand, and the United States. We should note that, in accor-dance with our terms of reference, our examination of surveillance was confined to pe-riods when these countries made no, or at most sporadic, use of IMF resources.

Over the next months, we made country visits as well as taking advantage of the1998 Annual Meetings to meet with a number of country officials, IMF Executive Di-rectors, IMF staff, and visitors. Our preparatory work included, in addition to theseinterviews, reviewing the Article IV staff reports for these countries covering the pe-riod from 1990 to the present, as well as a considerable amount of relevant internalFund documentation, including the mission briefings, back-to-office reports, and in-terdepartmental memoranda. We also met with a group of representatives of smallstates. In February 1999, we met in Washington to discuss our approach to draftingthe report, and to conduct a number of further meetings. Between February and June,we met again in Washington a number of times to prepare and finalize the report, andto take account of comments on the draft from Fund staff.

While the main focus of our research was our sample countries, the subject matterof the report is Fund surveillance as a whole, and the conclusions and recommenda-tions are intended to be of general applicability. The report is therefore organized the-matically, rather than in the form of country case studies. In Chapter I, we give somehistorical and institutional background. Chapters II–IVdescribe the results of our in-terviews and research, grouped under the three headings: conduct and method, sub-stance, and impact. We should note that this material, while descriptive of the viewswe heard, does not necessarily represent our own views; in particular, specific quotesare included only when they are of use in elucidating a particular point. We then setout our own analysis, conclusions, and recommendations in Chapter V.

We would like to express our gratitude to all those who assisted us in various ways.Most of all, we would like to thank all those who met with us to discuss our work—country officials, Fund staff and former staff, Executive Directors, academics, andprivate sector and nongovernmental organization (NGO) representatives. Withouttheir time, trouble, and willingness to speak frankly, our task would have been impos-sible. A list of those we met is in Appendix III. We would also like to thank all thosewho assisted us in setting up our country visits, and the Director and staff of the Of-fice of Internal Audit and Inspection, in particular, Elena Frolia and Cathy Song.Soren Baunsgaard provided valuable research assistance. We are also grateful to Marina Primorac for editorial assistance.

Preface

11

This evaluation of Fund surveillance over mem-bers’policies under Article IV of the Articles of

Agreement was commissioned by the ExecutiveBoard in June 1998. Our terms of reference are inAppendix I.

Much of our research centered on the Fund’s sur-veillance over the past decade of a sample of 12 coun-tries: Brazil, Chile, China, the Czech Republic, India,Japan, Korea, Saudi Arabia, South Africa, Sweden,Thailand, and the United States. We reviewed docu-mentation prepared by the Fund, we discussed the is-sues with the relevant Fund staff, and we visited thecountries, where we met present and former govern-ment officials and others who had come into contactwith the Fund. We also met numerous other presentand former Fund staff, a number of Executive Direc-tors, officials of other national governments, other in-ternational institutions where we could, academics,private sector, and NGO representatives.

The report comprises five chapters. The first de-scribes the historical and institutional background tosurveillance. The next three describe the results ofour interviews and research, grouped into the threeheadings of conduct and method, substance, and im-pact. The concluding chapter sets out our own con-clusions and recommendations.

Our main findings in our discussions were thefollowing.

• In bilateral surveillance, the quality of Fundanalysis and advice in the Fund’s traditionalareas of core expertise—exchange rate policiesand the associated macroeconomic frame-work—is generally rated highly. The quality ofmultilateral surveillance is also generally seenas good. The most important reservation wasin the area of the Fund’s analysis of capital ac-count issues.

• The scope and coverage of bilateral surveil-lance has expanded significantly, especially inthe past few years, into structural issues of anonfinancial nature. This expansion is ques-tioned. Doubts were raised as to the Fund’s ca-pacity to deal adequately with such issues.Furthermore, there were concerns that this ex-

pansion of coverage in fact detracted from theeffectiveness of surveillance overall.

• There was considerable critical comment onthe fact that bilateral surveillance concentratesalmost exclusively on the country’s own situa-tion, and rarely focuses on the various interna-tional dimensions of a country’s macroeco-nomic and financial situation.

• While in their own cases member country au-thorities did not generally regard the Fund’s bi-lateral surveillance advice as insufficientlyfrank and direct, this concern persists amongother observers.

• Member country authorities were generallycontent with the broad process of bilateral con-sultation. However, there was also an interestin shortening Article IV consultation missionsand in making more use of telecommunica-tions as a way of keeping on top of develop-ments in addition to visits.

• Among many close observers of the institu-tion, knowledge transfer across departments isseen as deficient. In the surveillance context,this manifests itself in an inadequate cross-fertilization between multilateral and bilateralsurveillance, and in insufficient use of what islearned in different countries.

• The interdepartmental review process for sur-veillance documents, while necessary, has become excessively formalized and time-consuming, adding to an already significantproblem of overwork for many staff involvedin surveillance. Beyond this, the Fund faces anumber of other internal organizational, man-agement, and staffing challenges that have animpact on surveillance.

• The Executive Board spends a great deal oftime on surveillance matters, especially in theconcluding discussions of Article IV consulta-tions. However, all agree that these discussionsare not well focused. And despite the importantpeer review role that the Board has in principle,it has little ownership of surveillance priorities.

Executive Summary

13

EXECUTIVE SUMMARY

• While surveillance is generally seen as con-structive by officials, its impact on countrypolicies is bound to be secondary to domesticforces, and less than many outside observersbelieve. In the case of four countries that wentinto crisis, we judged that the single most im-portant factor limiting Fund impact was a diffi -cult internal political situation.

• Internal reviews have identified reasonablywell the key areas in which surveillance shouldbe improved. However, progress has beenslow in translating the general principles setout in the reviews into actual change.

• The introduction of Public Information Notices(PINs) is generally regarded as a success. How-ever, views on the desirability of publishing Ar-ticle IV staff reports remain divided. At the sametime, we noted that in practice these reports al-ready receive a relatively wide, if unofficial, cir-culation, and, furthermore, what is put in thosereports is filtered in response to this reality.

Based on these findings, our key recommendationis that bilateral surveillance should focus as much aspossible on the core issues of exchange rate policyand directly associated macroeconomic policies (in-cluding financial sector and capital account issues).Furthermore, the international implications of suchpolicies should be given significantly greater atten-tion. In regard to the latter point, there are three dis-tinct areas where the Fund has a clear comparative,and as yet underexploited, advantage:

• in relating the country’s position to the interna-tional economic situation and prospects;

• in analyzing the experiences of other countriesconfronting similar policy problems; and

• in discussing the likelihood of, and possible re-sponses to, significant negative externalshocks, whether originating from direct effectsthrough trade flows or interest rates or frommore general contagion.

Bringing the Fund’s expertise to bear on surveil-lance more effectively in this way will require somereallocation of resources. Accordingly,

• We recommend a significant shift in surveil-lance work toward the areas described above.We cannot rule out the possibility that this maymean some increase in resources for functionaldepartments, but principally it involves reallo-cation and refocusing within area departmentsas a group.

• We recommend that the Fund curtail the expan-sion of the scope of surveillance into nonfinan-cial structural areas, with consequent savings in

resources. Analysis outside the areas of core ex-pertise—exchange rate policies, the associatedmacroeconomic framework, and financial sectorand capital account issues—should only be un-dertaken if directly relevant to a particular case.

• We recommend that there should be a greateremphasis on more continuous surveillance,through shorter, leaner, more focused visits,and more regular long-distance communica-tion and exchange.

• We recommend a reduction in the resourcesdevoted to the surveillance of small andmedium-sized industrial countries (and, moregenerally, participants in the euro area). Thisscaling back would essentially be achievedthrough the prioritization described above, andthrough longer intervals between Article IVconsultations, in part replaced by more contin-uous surveillance.

• We recommend that surveillance of the largestindustrial countries—the United States, Japan,and the euro area (which should be viewed as awhole)—focus more on the international as-pects of policy.

• We recommend that surveillance devote sub-stantially more attention to identifying vulner-abilities. A point should be made of presentingthis analysis to national authorities for theirconsideration.

• We recommend quarterly publication of theWorld Economic Outlook forecasts, with ulti-mate responsibility given to the Research De-partment for forecasts and analysis.

We also make a number of suggestions of a moreorganizational nature:

• We recommend some internal measures to bet-ter attune the skills mix of staff involved insurveillance to new demands, and to improvethe accountability and the incentives for staff.We also make a suggestion for a broader,deeper, review of such matters.

• We are concerned about the extent and conse-quences of overwork among staff engaged insurveillance and have aimed at making recom-mendations that, besides enabling a reallocationof resources, also make possible some reductionin the overall volume of surveillance work rela-tive to the number of staff engaged in it.

• We recommend significant changes in the wayin which the Executive Board discharges itsstatutory role in the surveillance process byproposing extensive use of committee work;all Article IV reports should initially be dis-cussed in a Board committee.

14

Executive Summary

On some issues—transparency, internationalstandards, and contingent credit facilities—thepractice of surveillance was changing as we pre-pared this report.

• The introduction of PINs and the recent deci-sion to experiment with the voluntary publica-tion of some Article IV reports are welcome.However, we do not believe this goes farenough. We recommend full publication of allsuch reports, while defining also an area forconfidential exchange between staff and na-tional authorities.

• We recommend that—outside the Fund’s coreareas—monitoring international standards

should to the maximum extent possible be del-egated to other international institutions or as-sociations with the necessary expertise, withthe Fund, because of its existing surveillancerole, acting as a clearinghouse for information.

• We believe that the recent introduction of theContingent Credit Line could increase the im-pact of surveillance by raising the demand forhigh-quality assessments and by adding finan-cial clout. But we underline that this will re-inforce the need to implement a number of thechanges that we recommend.

For reference purposes, a complete list of recom-mendations is attached in Appendix IV.

15

Objectives of Surveillance

Objectives

1. How is Fund surveillance defined? Our terms ofreference requested us to examine “all channels andinstruments of Fund surveillance, including bilateralsurveillance, regional surveillance, multilateral sur-veillance and the content and format of the WorldEconomic Outlook, and surveillance of internationalcapital markets and financial systems and of the pro-vision by member countries of economic and finan-cial data to the IMF and the public.” We take surveil-lance as referring to all aspects of the Fund’sanalysis of, scrutiny over, and advice concerning,member countries’economic situations, policies,and prospects. Surveillance is conducted for the ben-efit both of the international community as a wholeand of individual member states. 2. While the Fund regards the surveillance it under-takes on behalf of the international community asone function and our remit treats it as such, in prac-tice it is clear that Fund surveillance is far frombeing one unified concept. In fact, as currently prac-ticed, surveillance reflects a number of overlappingbut conceptually distinct purposes. We have identi-fied six such purposes.

• Policy advice. The Fund offers advice and pro-posals on the main economic policy dilemmasfacing a member country. It also provides asounding board for discussion.

• Policy coordination and cooperation.1 TheFund cannot, at least through the surveillancefunction, force cooperation. However, by pro-viding what might be seen as neutral and reli-able data, forecasts, and analysis, it providesmachinery through which policy coordinationcan take place if countries so wish.

• Information gathering and dissemination. TheFund gathers both statistics and other informa-

tion about the direction of policy. In this, itprovides a service to member countries, whocan rely on the Fund’s reports rather than hav-ing to maintain their own economic informa-tion-gathering networks. To the extent that theFund then disseminates this information pub-licly, this service also benefits private marketsand the general public.2

• Technical assistance and aid. In many coun-tries, surveillance is essentially technical assis-tance; that is, it supplies expertise in macro-economic policymaking that is scarce in thecountry. Not surprisingly, this is especially trueof smaller and poorer countries, but it is also arole in some medium-sized and larger ones.

• Identification of vulnerabilities. This is an ex-tension of the informational role and policy ad-vice role that is particularly relevant when acountry’s policies are likely to be unsustain-able. It is also a role that has become muchmore prominent in the past five years. If thegovernment is warned of such problems by anoutside, objective source, it may be able totake the necessary policy measures. It is alsosuggested—although the Fund has yet to takesuch action—that by providing warnings to fi-nancial markets and the public, the incentivesfor policymakers to take measures in a timelyfashion would be substantially increased.

• “Delivering the message.” This is an extensionof the advice role. The Fund provides coun-tries with policy prescriptions on numeroustopics, from the advantages of moving towarda system of indirect instruments of monetarycontrol to the need to liberalize labor markets.By doing so, the Fund provides a way bywhich the prevailing consensus of the econo-

1 Introduction

17

1In what follows, we do not make the distinction sometimesmade in the academic literature between “coordination” and “cooperation.”

2It is notable that the informational function of the Fund hasonly recently come to the fore; many authorities do not mention itat all. For example, see Manuel Guitián, The Unique Nature ofthe Responsibilities of the International Monetary Fund, IMFPamphlet Series No. 46 (Washington: International MonetaryFund, 1992), which does not mention the provision of informa-tion as one of the Fund’s “key institutional functions” (p. 8).

I INTRODUCTION

ics profession is disseminated to governmentand policymakers throughout the world.

Underpinnings

3. Why might Fund surveillance be necessary?When government, or another public authority likethe Fund, produces goods or services, it is standardpractice for economists to demand a justification.What is the market failure that prevents the good orservice from being produced by the private sector?In the case of Fund surveillance, it is necessary toask not only why the private sector does not do it,but also why national governments cannot. As thediscussion of objectives above makes clear, Fundsurveillance may be held to provide three types ofservices to members and the general public: infor-mation, policy advice, and policy coordination.4. Information, including economic data and infor-mation about government economic policies, clearlyhas many of the characteristics of a public good. It isvery difficult for a private supplier of information tocapture all the returns to that information; in otherwords, the social return to the provision of informa-tion is likely to exceed the private return. As a result,less information may be supplied than is socially effi -cient. This is perhaps less true than it used to be; theprivate return to high-quality information in financialmarkets is quite high, and as a result there are a greatmany private sector providers of such information,especially in developed financial markets. However,even in industrialized countries, the private sectorprovides little information about economic data andpolicies that is not directly relevant to financial mar-kets, and even less in developing countries. More-over, private sector providers are likely to make theinformation they collect available primarily to payingclients; and in the case of some (e.g., investmentbanks), conflicts of interest may arise. 5. But all this does not in itself explain why theFund, rather than national governments, should pro-vide the information. Why have an international in-stitution writing reports on each country rather thaneach country simply submitting an annual economicreport? Surely the latter would be more efficient.The answer here is that such an arrangement wouldclearly create a conflict of interest; countries wouldhave an incentive to be too optimistic, for both polit-ical and economic reasons. This would be true evenif the analysis were merely for the use of other gov-ernments; it is even more so if the analysis is to formthe basis for decision making in the private sector. 6. It should be noted here that the provision of in-formation does not always unequivocally improvematters. In particular, while complete information isfirst-best, more information is not always better thanless. This is a point that has recently been made in

the context of the disclosure of countries’reservepositions. Some finance ministries and central bankshave argued that as long as supposedly importantprivate sector players—such as hedge funds—facelittle or no obligation to disclose their positions, it isnot clear that forcing countries to disclose theirs in-creases either economic efficiency or social welfare. 7. It is also far from clear that the Fund is any betterpositioned than the private sector to play the role ofidentifying vulnerabilities to private markets. Itsopinion on the sustainability of a country’s policiesis presumably valuable to private markets becausethe Fund has access to information that enables it toproduce a better analysis than private sector agentscan; that is, it has a comparative advantage. How-ever, if countries know that the Fund will “blow thewhistle” in certain circumstances, they are unlikelyto provide it with such information, so the Fund willlose its comparative advantage. These are clearlycomplex issues.8. Turning to policy advice, why should an interna-tional organization provide it? As noted above, insome cases governments may simply not have suffi -cient capacity to do as much or as good policy analy-sis as they would like. In this case, surveillance, likeother Fund technical assistance, is essentially a formof development aid, provided and financed by the in-ternational community. More problematic are the cir-cumstances under which the country itself has amplecapacity within government to formulate economicpolicy. In this case, the justification for Fund surveil-lance must be that the Fund can provide some formsof advice more efficiently than the government cando itself; in other words, that it has a comparative ad-vantage. In particular, it could be argued that theFund might have a comparative advantage in the pro-vision of advice on the international macroeconomicenvironment, on the policies adopted by other coun-tries, or in specific areas where it has developed aparticular expertise (hypothetically, for example, theidentification of financial sector vulnerabilities). 9. The presumed objectivity of the Fund may alsobe an advantage in the provision of advice, even ifits basic analysis is no better than that of the govern-ment or of private sector agencies. This is particu-larly relevant when there are internal disagreementswithin the government: the Fund may be able to saythings that certain branches of government cannot,or to strengthen the position of those within govern-ment arguing for appropriate policies; or even toserve as the scapegoat for the implementation of un-popular policies (although this is probably more rel-evant in the context of an IMF-supported program). 10.Finally, the economic literature on internationalpolicy coordination describes the circumstances inwhich countries might benefit from coordination.For example, in the classic “Prisoners’Dilemma”

18

Origins and Development of Fund Surveillance

example, two agents, or countries, that set policynoncooperatively both end up with outcomes thatare inferior to those attainable under coordination. Inthe international macroeconomic policy arena, thismight be thought to apply to fiscal expansion, com-petitive devaluation, trade barriers, etc. It might bethought that the Fund would have a role in enablingcountries to coordinate to achieve the best outcomefor all. In the macroeconomic policy area, the Fundis clearly a plausible candidate for such a role. 11. That role could in principle range from the mini-malist—providing reliable information and data, andperhaps some analysis—to a much more powerfulone of setting and enforcing international coordina-tion. In the extreme, the Fund could identify the setof policies that would produce the optimal outcomefor the world as a whole, and even provide some sortof enforcement mechanism to ensure that countriesadopted them. Of course, even if such an arrange-ment were desirable—and there are a number of ar-guments, both theoretical and practical, why it mightnot be—it is clearly highly implausible at present.The Fund’s current role is much closer to the mini-malist end of the spectrum. 12. It should also be noted that the nature of this rolevaries along a second dimension, depending onwhether international policy coordination is basedon discretionary action or on rules and standards.3

Under the Bretton Woods regime, there were majorelements of a rules-based system; countries weresupposed to conduct domestic policies in a mannerconsistent with their chosen exchange rate. Since1973, policy cooperation has been largely discre-tionary in nature. But recent initiatives to place moreemphasis on rules and standards—not through fixedexchange rates, but through codes of conduct and thelike—may lead to the pendulum swinging backagain, with important implications for surveillance,as discussed in Chapter V.13.This discussion suggests that Fund surveillancewill be most useful and efficient when the serviceprovided is a public good, that is, undersupplied bythe private sector; and where the Fund, as a particu-lar type of international organization, has a compara-tive advantage in providing such a service, either byvirtue of its resources or its “objectivity.” In all threeareas—information, policy advice, and policy coor-dination—there are tasks the Fund may undertakethat meet these criteria. However, that does not meanthat it is sensible for the Fund to undertake everyconceivable task in these areas. There are many

areas of policy advice, for example, where it maynot command the best expertise available.

Origins and Development of Fund Surveillance4

Origins and Legal Basis

14. Scrutiny of members’economic policies—evenoutside the context of Fund programs or of a prospec-tive change in the exchange rate regime—has alwaysbeen one of the functions of the Fund. Indeed, the con-cept of a degree of oversight of the international mon-etary system, undertaken by an international organiza-tion and based on some common framework of normsfor the economic policies of individual countries, orig-inated with the League of Nations, although theLeague was a political rather than economic organiza-tion.5 Many authorities see this as the underlying prin-ciple on which the Fund is based: “There is a well-defined thread that binds together all the activities ofthe IMF: the promotion and safeguarding of an inter-national code of economic conduct.”6

15. In the immediate postwar period, the Fund’smain task, in addition to the oversight of the BrettonWoods par value system, was to encourage membersto move toward current account convertibility. Inthis context, the Fund examined the economic cir-cumstances of individual members under ArticleXIV of the Fund’s Articles of Agreement.7 This au-thorized members to maintain certain exchange re-strictions, provided that they held regular consulta-tions with the Fund “as to their further retention.”16.Between 1958 and 1961, most European indus-trialized countries moved to convertibility; that is,acceptance of Article VIII of the Fund’s Articles of

19

3See, for example, Ralph C. Bryant, International Coordina-tion of National Stabilization Policies(Washington: BrookingsInstitution, 1995), for a helpful exposition of the two dimensionsof international policy coordination (minimalist to maximalist,and rules versus discretion).

4This section draws heavily on James M. Boughton, Silent Rev-olution: The International Monetary Fund, 1979–1989(Washing-ton: International Monetary Fund, forthcoming). See also LouisW. Pauly, Who Elected the Bankers?(Ithaca, New York: CornellUniversity Press, 1997), for a description of the origins of eco-nomic surveillance.

5See Pauly (1997), chapters 3–5. Pauly views the surveillancefunction as inherent in the design of the Fund, which in turn was aproduct of the experience of the interwar years. “The surveillancefunction of the Fund has existed in embryonic form ever since theorganization was established. . . . In a very basic sense, the accep-tance of these obligations, implicitly in 1945 and explicitly threedecades later, was a response to the depression studies of theLeague of Nations.” However, Jacques Polak suggests Pauly“paints a picture of more continuity—and that means lessprogress—than the facts warrant.” (Jacques Polak, “Commentson Louis Pauly,” unpublished mimeo, 1998, p. 1.)

6Guitián (1992).7In the early postwar period, surveillance of Western European

economies took place primarily through other institutions, no-tably the Organization for European Economic Cooperation andthe European Payments Union.

I INTRODUCTION

Agreement. As a result, they would no longer havebeen subject to regular consultations. The UnitedStates, however, did not want the Fund to lose its ca-pacity to provide information and analysis of Euro-pean economies. The Europeans, for their part,wanted symmetry, so that the Fund should also beable to examine the United States (and Canada).8 Itwas therefore agreed that the Fund should introducevoluntary consultations for all members, beginningin 1961 with the United Kingdom. 17.These consultations dealt with the general eco-nomic circumstances of the country, thus rangingwider than simply issues relating to exchange re-strictions. However, they had no formal status. As aresult, there was a limit to how meaningful theycould be, particularly in the par value system estab-lished at Bretton Woods. In such a system, the realissue for a country with serious problems is whetherto adjust the exchange rate. There was little prospectthat countries would discuss such an option seri-ously in voluntary consultations.9

18. Multilateral surveillance would have been an ob-vious task for the Fund; under the Bretton Woods sys-tem there was a need, when imbalance became appar-ent, to determine who should adjust to whom.However, the Fund’s work in the 1960s was almostexclusively based on country-by-country analysis, ex-cept for the evaluation of the need for internationalliquidity. It was left to the Organization for EconomicCooperation and Development (OECD) to initiatemultilateral surveillance, through its Working Party 3and through work on systematic comparisons of na-tional policy experience in various policy areas.10

19. In August 1971, the United States closed the“gold window,” effectively ending the par value sys-tem. However, it took several years before policy-makers recognized that the system of fixed exchangerates could not be revived. So it was not until January1976, at the Interim Committee meeting in Jamaica,that a framework for the new international monetarysystem was created. This new system accepted float-ing exchange rates, but foresaw a key role for theFund in avoiding “excessive” fluctuations. In partic-ular, under the new Article IV of the Articles ofAgreement, the Fund: “shall exercise firm surveil-lance over the exchange rate policies of members andshall adopt specific principles for the guidance of allmembers with respect to those policies.”11

20. The introduction of surveillance as an explicit partof the Fund’s mandate therefore came as part of theFund’s adaptation to the post–Bretton Woods mone-tary system. In particular, the use of the phrase “firmsurveillance over the exchange rate policies of mem-bers” was an attempt to ensure that the internationalcommunity still exerted some discipline over ex-change rates, even in a world of floating rates.12 Assuch, it represented a compromise between those whobelieved that a substantial degree of international co-ordination was necessary to maintain exchange ratestability, pending—it was hoped—a return to a moremanaged system of exchange rates, and those who be-lieved that such an objective was unrealistic (cor-rectly, as it turned out) and probably undesirable.13

21.Surveillance also had an important role in theFund’s lending activities. The policy advice given bythe Fund in the surveillance exercise was likely toprovide the framework for the conditionality at-tached to any future Fund program.14 It has oftenbeen argued that, as a consequence, surveillance isthe one essential core of the Fund’s activities.15

22.For operational purposes, the Fund’s ExecutiveBoard set out five ways in which surveillance wouldoccur.16

• Periodic (usually annual) consultations withindividual members would take place;

• The Board was periodically to review “broaddevelopments in exchange rates”;

• The Managing Director was to maintain closecontact with members regarding exchangerates;

• Members were to be required to notify theFund of any changes in exchange rate policies;and

20

8See Pauly (1997).9As illustrated by, for example, the U.K. experience in the

1960s.10See OECD, The Balance of Payments Adjustments Process

(Paris: OECD, 1966), and subsequent OECD studies of fiscal andmonetary policy.

11The text of the relevant sections of Article IV and of Article I(which describes the general purposes of the Fund) is reproducedin Appendix II.

12Boughton (forthcoming) describes the phrase as a compro-mise between advocates of flexible and stable exchange rates:“Those who sought a flexible system in which exchange ratescould adjust freely to market forces saw surveillance as a meansof discouraging countries from manipulating exchange rates inopposition to market pressure. Those who sought greater stabilityin exchange rates saw it as a means of encouraging countries toadopt economic policies that would ensure such stability.”

13See Harold James, “The Historical Development of the Prin-ciple of Surveillance,” Staff Papers, International MonetaryFund, Vol. 42 (December 1995), pp. 771–72.

14The Contingent Credit Line, whereby the Fund agrees in ad-vance to make funds available to a country if it were to requirethem, clearly blurs the line between surveillance and conditional-ity even further.

15See, for example, James (1995), p. 775: “The consultationexercise made the IMF aware of problems that might potentiallyrequire financial assistance. As a result, the IMF’s financial pro-grams, and the conditionality attached to them, could be regardedas nothing more than an extension of the surveillance procedure.”This point was also made to us in conversation by the ManagingDirector and Jacques Polak.

16Executive Board Decision No. 5392, April 29, 1977.


• The Managing Director could initiate “specialconsultations” with members.

22.The third and fourth of these points had littlesubstantive content, while the fifth (as describedbelow in paragraph 30) has been invoked only veryrarely. However, the first and second formed thebasis for the Fund’s bilateral and multilateral sur-veillance operations, respectively.

Scope

23. It was always recognized that “exchange rate poli-cies” could not be viewed in a vacuum, and that thesustainability or otherwise of a country’s exchangerate policy was likely to be determined by its compat-ibility with domestic macroeconomic policies. As theoriginal Decision put it, the Fund’s appraisal of ex-change rate policies: “shall be made within the frame-work of a comprehensive analysis of the general eco-nomic situation and economic policy strategy of themember, and shall recognize that domestic as well asexternal policies can contribute to timely adjustmentof the balance of payments.”17 Of course, virtuallyany domestic policy could be argued to affect, insome way, the external position. Therefore, as onepractitioner put it: “the challenge is to identify the do-mestic policy areas that influence primarily the econ-omy’s external position, so as to provide the basis fora general consensus among the membership that theyare of legitimate international concern and properlybelong within the scope of surveillance.”18

24. So Article IV consultations always focused onmonetary and fiscal policy. However, there is clearly adifference between examining a country’s macroeco-nomic policies solely with a view to determiningwhether they are consistent with its exchange rate pol-icy and pronouncing more generally on the merit ofthose policies.19 Even early on, the Fund generallytook the broad approach. And over time, the scope ofsurveillance has steadily expanded into “structural”policies (e.g., market policy, privatization, industrialpolicy, and competition policy), into the financial sec-tor (e.g., capital account issues, banking supervision,deposit insurance, and other financial sector regula-tion), and into a number of other areas (e.g., the envi-ronment, military spending, the “millennium bug”).20

26.This expansion, which has been particularlyrapid in recent years, has been driven by a numberof factors.

• The need for the Fund to address the problemsof formerly centrally planned economies. Inthese countries, structural questions—likepublic enterprise reform and privatization—clearly had major macroeconomic implica-tions. While most of these countries had Fund-supported programs, the effects of this shifttended to permeate throughout the organiza-tion and hence into surveillance.

• Political pressures. As the Fund became moreopen, and as nongovernmental organizationsfocused more of their attention on internationalorganizations like the Fund and the WorldBank, there were pressures for the Fund tolook wider than macroeconomic policy, con-ventionally defined, to the implications of is-sues like the environment and poverty reduc-tion. In recent years, the U.S. Congress hasexerted considerable indirect pressure on theFund in this regard.

• A growing consensus in the economics profes-sion that a number of important economic vari-ables, including unemployment and growthrates, can in the short term be affected by de-mand management policies, but are primarilydetermined in the long term by supply-sidefactors, combined with a desire on the part ofthe Fund to be relevant to what was regardedas important.

• The Mexican crisis (see below) and other fi-nancial crises in developed and developingcountries (from Sweden to Indonesia, but par-ticularly the recent Asian crisis) focused atten-tion on the potential for external crises to beprecipitated not only by traditional macroeco-nomic policy failures (government fiscal ormonetary laxity) but also by structural weak-nesses in the financial sector.

27.Of course, these factors do not necessarily ex-plain why the Fund is involved in all these issues.From an institutional perspective, the origin of thisexpansion is the fact that the Fund is the only organi-zation that has a mandate to examine on a regularbasis the economic circumstances of virtually everycountry in the world. So even though this mandatewas originally intended to apply only to a strictly

21

17Executive Board Decision No. 5392, April 29, 1977.18Guitián (1992), p. 12. It is notable, reading Guitián’s descrip-

tion (pp. 12–14) of the proper coverage of surveillance, how muchthe scope of Fund surveillance has expanded in the last decade.

19One of the academics we spoke to described this as the differ-ence between simply advising a country how to avoid a crisis, onthe one hand, and trying to optimize the use of all policy instru-ments on the other.

20An interesting illustration of this expansion can be found inthe terms of reference for this evaluation, which ask us to assess

“the effectiveness of surveillance in identifying those macroeco-nomic, structural, and financial weaknesses and imbalances inmember countries and the world economy that are an obstacle toachieving sustainable noninflationary economic growth and ex-ternal viability.” This clearly goes well beyond exchange rates.

I INTRODUCTION

limited set of macroeconomic questions, when theinternational community thinks that surveillance isrequired in other areas—even if those areas gosomewhat beyond the Fund’s original mandate—theFund is the obvious institution to call upon. Policy-makers frequently say that it is the only institutionthat actually has “troops on the ground.” This hasbeen particularly noticeable in the recent move to-ward giving the Fund the responsibility of monitor-ing international standards in areas like accounting,auditing, and corporate governance.21

28.Within the Fund, both the Executive Board andmanagement have supported this expansion of theFund’s mandate. In the general context, the most re-cent guidance promulgated by the Board in the dis-cussion of lessons from the Asian crisis in March1998 stated that “the focus of surveillance needs toextend beyond the core short-term macroeconomicissues, while remaining selective.” In relation to in-dividual countries, when discussing staff reports, anExecutive Director frequently asks why a particularissue, which he or she considers relevant to the eco-nomic circumstances of the country in question, hasnot been covered: this generates pressure for thestaff to cover that issue in future reports on thatcountry and others in similar circumstances. And,especially in recent years, Fund management hassignaled a readiness to expand the mandate of the or-ganization to issues like social policy.22

Problem Cases

29.A perennial issue in the operation of Fund sur-veillance has been the tension between “equal treat-ment” and “focus.” On the one hand, on the princi-ple that all Fund members, all sovereign states, areequal, many of them have argued for surveillance tobe comprehensive and comparable among all mem-bers. Moreover, there is a natural tendency for alarge bureaucracy with many functions to seek to or-ganize its work around a predetermined and rela-tively fixed schedule. As the description of the cur-

rent Article IV consultation process below shows,this is precisely what happens. On the other hand,some members are clearly more important to theworld economy and their neighbors than others;some members are more vulnerable to the externalenvironment and to external shocks; and some mem-bers have a less well developed capacity to formu-late and implement sound economic policies. Giventhat Fund staff resources are, and should be, limited,some have argued that the Fund’s attention shouldfocus more closely on members falling into one ormore of these categories. As indicated in the Execu-tive Board discussion of the 1997 biennial review ofsurveillance:

Directors emphasized that the principle of annual consul-tations represented a cornerstone in ensuring the continu-ity of Fund surveillance. At the same time, Directors rec-ognized the need for flexibility in Fund procedures toensure an effective focus of Fund surveillance, particu-larly in the context of the Fund’s strained resources.

30. In general, the Fund has resolved this dilemmain the direction of equal treatment rather than selec-tivity. Furthermore, although the Fund has thepower to initiate “special” or “supplemental” con-sultations with countries whose exchange rate poli-cies are, in the view of the Managing Director, ofparticular concern, this procedure has only been in-voked extremely rarely because of the perceivedpolitical implications.23 When a member’s eco-nomic situation gives rise to serious concern, theFund’s response has usually been informal andconfidential, taking for instance the form of a letteror visit from management.24

31.One respect in which the focus of surveillancehas changed is in the balance between the industrial-ized and developing countries. The surveillancemechanism, as described above, was part of a frame-work primarily aimed at securing exchange rate sta-bility among the major industrialized countries. Inthe 1970s and early 1980s, much of the Fund’s ef-forts and resources were concentrated in that area,albeit with mixed results. However, from 1982, theFund became heavily involved in efforts to resolvethe debt crisis. And, in the 1990s, with the growth ofprivate capital flows to emerging market countries,surveillance has increasingly come to be seen—bothinside and outside the Fund—as a way of helpingthese countries to attain or preserve domestic macro-

22

21As the Managing Director put it: “Countries need new laws,new institutions, and strong professionals to adopt and apply thenew standards. And the international community needs mecha-nisms to make the standards operational and to monitor progress.The IMF, which has been given a universal mandate for surveil-lance, will have here a critical role—a daunting task indeed—forwhich it will need to avail itself of the support of the variety ofother bodies with more practical experience in each of these spe-cific areas.” Michel Camdessus, Speech to the Foreign Policy As-sociation, February 24, 1999.

22In particular, management has sought to define the ultimategoal of the Fund as “high-quality growth,” meaning growth thatis sustainable, equitable, and environmentally friendly. See, forexample, Michel Camdessus, “Addressing Concerns for the Poorand Social Justice in Debt Relief and Adjustment Programs,”speech by the Managing Director, October 22, 1998.

23Supplemental consultations were held with Sweden in 1982and Korea in 1987. Neither appears to have had much effect onpolicy in the countries concerned, but the Swedish exercise mayhave helped to reduce the political and economic tensions arisingbetween Sweden and its trading partners as a result of a devalua-tion viewed by many as excessive. See Boughton (forthcoming).

24As in the case of Thailand. See the discussion in Chapter IV.


economic stability, so as to allow them to preserve orexpand their access to such capital flows. Mean-while, the interest of the Group of Seven industrialcountries (G-7) in international policy coordinationhas waxed and waned; it certainly has not shownmuch inclination to give the Fund any major role inexchange rate management. As a consequence, thefocus of Fund surveillance—at least in the publicmind—is increasingly on developing and emergingmarket countries, particularly those considered to beof systemic significance.

Multilateral Surveillance

32. Not surprisingly, the central problem of multilat-eral surveillance since the breakdown of BrettonWoods has been the exchange rate system.25 Thedilemma has always been that while a floating rateregime has frequently allowed the exchange rates ofthe major economic powers to diverge significantlyfrom fundamentals, with consequent destabilizing ef-fects on international trade and capital flows, no bettersystem has been on offer. In particular, estimates of“fundamental equilibrium” exchange rates are fraughtwith uncertainty. Moreover, even if such consensusestimates existed, there would still be ample scope fordispute—both political and economic—over the ap-propriate policy instruments to be adopted to bringmarket rates into line. So, while the Fund has soughtto make estimates of equilibrium exchange rates, ifonly for internal use, it has recently been rare for it tomake significant policy recommendations to themajor countries on this basis.33. Multilateral surveillance has therefore tended toconcentrate on forecasting and analysis rather thanpolicy prescription. It has centered on the World Eco-nomic Outlook(WEO), produced at first annually andnow semiannually. The WEO, with the forecastingand analytic apparatus that its production requires,provides a comprehensive set of economic forecastsfor the world economy. This in turn provides a basisfor discussion by countries in multilateral forums likethe G-7. However, unlike the Article IV surveillanceexercise, there is no attempt by the Fund itself to de-velop a comprehensive set of policy recommenda-tions, although the published WEO contains somediscussion of individual countries’policies.

Publicity

34. Initially, it was clear that surveillance was forthe benefit of the Fund’s member governments, not

for private markets or the public at large. However,over time, the Fund has moved toward greateropenness. To a large extent, this was the result ofpressure from certain governments to make publicsome of the material produced in the surveillanceprocess. In the 1980s, some countries began to re-lease the Fund mission’s “concluding statement.”In the context of the 1980s debt crisis, the Fundagreed to allow a few countries making use of the“enhanced surveillance” procedure to make avail-able the staff report itself to private sector credi-tors, on a confidential basis. 35.The first formal publication of Article IV mater-ial by the Fund came in 1989 when the “Recent Eco-nomic Developments” (RED) for Germany was pub-lished. Over time, the principle of routinepublication of REDs (now “Selected Issues”) wasestablished. However, the RED/Selected Issues pa-pers are usually either factual and statistical, or morelike academic working papers, focusing on amedium- or long-term policy issue. As such, they areless sensitive than the staff report itself, which in-cludes the more politically and market-sensitive dis-cussion and recommendations on current macroeco-nomic policies and exchange rate issues. 36. In May 1997, the Fund introduced Press Infor-mation Notices (PINs, later Public Information No-tices), described in more detail below. Althoughpublication has remained voluntary, this initiativerepresented the first real recognition that the Arti -cle IV analysis of current macroeconomic develop-ments and policies was of concern and interest, not only to the governments of members, but also to the general public and to financial marketparticipants.

Resources

37. In terms of resources, surveillance is the Fund’smost important activity. In 1999 surveillance isbudgeted to account directly for 617 staff-years, 22percent of the total. However, this is misleading,since considerable surveillance resources are attrib-uted to central functions such as administration andexternal relations. Of the three principal “outputs”of the Fund—surveillance, “use of Fund resources”(programs), and technical assistance—surveillanceaccounts for about 42 percent. This proportion isforecast by the Fund to remain roughly constant inthe next few years. Within the total resources de-voted to surveillance, well over one-half was attrib-uted to bilateral surveillance (Article IVconsulta-tions), with the remainder divided amongmultilateral surveillance, policy development, re-search, and evaluation.

23

25Peter B. Kenen, “What Role for IMF Surveillance?” WorldDevelopment, Vol. 15 (December 1987), pp. 1445–56.

1. Under this heading the Executive Board asked usto address the following general question: how help-ful are the procedures of surveillance, the resourcesand the staff skills employed, the means of interac-tion with member country authorities, and the dis-semination methods of Executive Board surveillance conclusions? This general question wasbroken down into a number of more specific ones.26

(iii) Has Fund surveillance paid sufficient at-tention to regional surveillance, to interac-tion among countries, and to the externaleffects of policies in major countries?

(ix) Did the Fund balance openness and sen-sitivity appropriately from the authori-ties’ perspective? How helpful were thedocuments that emerged from bilateralsurveillance?

(xi) How do you assess the role of the Execu-tive Board in surveillance?

2. This chapter begins with discussions of how bi-lateral (or Article IV), regional, and multilateral sur-veillance is conducted. We then discuss the internalprocedures by which the Fund undertakes surveil-lance, and conclude with a discussion of the role ofthe Executive Board.

Bilateral Surveillance

3. We start with a description of the mechanics ofthe Article IV consultation.27 (See Box 2.1 for thechronology of an Article IV consultation.) This isfollowed by a discussion of the views we receivedon the consultation process and on other elementsof bilateral surveillance.

Mechanics

The Internal Process

4. Within the Fund, primary responsibility for theArticle IV process lies with the relevant area depart-ment. The process begins with the drafting of a“mission brief” by that department. The brief is bothdescriptive and prescriptive: that is, it describes theeconomic situation of the country and sets out thestaff ’s view of current policies and desirablechanges. It therefore serves both as an agenda for themission itself and as an outline of the major issues tobe discussed in the staff report.5. Before the mission brief is finalized, it is re-viewed by other departments (and sometimes theWorld Bank). The Policy Development and ReviewDepartment (PDR) plays a particularly importantrole in this process; while other departments are in-vited to comment on the brief—comments that thearea department may or may not choose to incorpo-rate or address—PDR must “sign off,” or approveit, before it is finalized. PDR’s role is to ensure thatthe brief is consistent with general Fund policies,that it addresses all the topics that Article IV con-sultations are supposed to cover, and that it takesinto account properly other departments’com-ments. The brief is then sent to management (theManaging Director or one of the Deputy ManagingDirectors) for final approval. 6. If there are significant disagreements betweendepartments on an important issue in the brief, acovering memo will call this to the attention ofmanagement. If the issue is felt to be sufficientlyimportant, or if management takes a particularlystrong interest in a forthcoming consultation, therewill be a meeting of the “Surveillance Commit-tee”—chaired by management, with representa-tives of the interested departments—to review whatthe Fund’s approach should be.28 These internalmechanics are of interest because it is largely atthis stage—the drafting of the mission brief, not the

II Conduct and Methods

24

26Numbers in parentheses indicate the number assigned to thequestion in the original Executive Board request. See the Termsof Reference in Appendix I.

27See Richard Harper, Inside the IMF: An Ethnography of Doc-uments, Technology, and Organizational Action(San Diego: Aca-demic Press, 1998), for an interesting discussion, from a ratherdifferent perspective, of the “career” of an Article IV staff reportand of an Article IV mission (to “Arcadia”).

28This committee is ad hoc (that is, it meets as required) anddoes not keep minutes.


preparation of the staff report—that the Fund’s“line” on the country’s policies is thrashed out in-ternally. It is also worth emphasizing that the mis-sion brief is purely a staff document, not shown tothe authorities of the country concerned or to theExecutive Board.

The Mission

7. The next stage is the mission itself. A missionusually consists of a visit of about two weeks to thecountry by at least four or five staff members. Themission is headed by a relatively senior staff memberfrom the area department—generally either the chiefof the relevant division or someone more senior—and consists of other staff from the area departmentand, if required in a particular area of interest, possi-bly one or two staff from functional departments. Themission consists primarily of meetings with seniorand midlevel officials from the finance ministry, cen-tral bank, and other government ministries involvedin the formulation of economic policy.

These meetings have two distinct purposes:

• Information gathering. This varies from sim-ply collecting and checking data and ensuringits consistency to ascertaining how govern-ment officials view the current state of theeconomy, and to getting more informationabout current and future policies.

• Policy discussion and advice. Here the Fundofficials discuss with government officialswhat they consider to be the main policy prob-lems or decisions facing the country, and maketheir own suggestions for how the governmentshould proceed.

8. There may also be meetings with the private sec-tor and representatives of “civil society”—in roughorder of frequency: bankers, financial market partic-ipants, business associations, academics, tradeunions, and NGOs.

9. At the end of the mission, there is often a formalconcluding meeting with senior officials (including,in many countries, the finance minister and centralbank governor). At this meeting the mission leaderwill sometimes read out a “concluding statement,”which summarizes the staff ’s view of the economy,the current policy stance, and any recommendationsfor changes. As noted in Chapter I, in recent years,some countries have made the concluding statementavailable to the media. 10. Immediately after returning to headquarters, themission leader will prepare a “back-to-office” re-port to management, summarizing the most impor-tant information about the visit. The mission staffwill then prepare a staff report for the ExecutiveBoard. The staff report includes a comprehensiveand reasonably detailed description of the macro-economic environment, the main economic policyissues facing the country, the staff ’s forecast, andan account of the discussions between the staff andthe authorities. However, the most important sec-tion is the “staff appraisal,” which sets out the staffview of the soundness of the authorities’policiesand what, if any, policy changes the staff recom-mend. The staff report will go through the same re-view process as described above, with commentsfrom functional departments, and approval by firstPDR and then management. It represents the for-mal, final view of Fund staff and management onthe country.

The Executive Board

11. After approval by management, the staff reportis submitted to the Fund’s Executive Board. TheExecutive Board discusses most staff reports insome detail, with the discussion of even a middle-sized country occupying most of a morning.29 Often

25

Box 2.1. Chronology of an Article IV Consultation (Approximate)

Day 1 Mission brief draft circulated among departments

Day 8 Mission brief submitted to management

Days 15–28 Mission visits country

Day 32 Back-to-office report circulated to management and departments

Day 51 Draft staff report circulated to departments

Day 65 Staff report circulated to Board

Day 90 Board discussion

29The discussion of a very small country, however, can be rela-tively brief.

II CONDUCT AND METHODS

most or all Executive Directors (or their representa-tives) will speak. Directors may also use the discus-sion of the staff report as an opportunity to bring upother issues, either specific to the country or moregeneral, that they wish to discuss.30 At the end ofthe meeting, the Chair (the Managing Director or aDeputy Managing Director) will “sum up” theviews of the Board. Usually, the summing up willbegin with the phrase “The Board broadly agreedwith the thrust of the staff appraisal,” or similar. Itwill then summarize the views expressed by Execu-tive Directors, including dissenting views (usingphrases like “Most Directors felt . . . However,some Directors felt . . .”).12.Since it represents the official collective viewof the Fund—not just the staff—and since it formsthe basis for the PIN (which also contains a factualbackground section), it is worth examining in somedetail how the summing up is prepared. A first draftis prepared before the Board meeting takes place. IfDirectors do not specifically dissent from the staffappraisal in a particular area, it is assumed that theyendorse it. The draft is then modified during themeeting to take account of Directors’views, partic-ularly in areas where they do not agree with thestaff appraisal. The new draft summing up is thenread out at the end of the meeting; this is the onlychance most Directors have to challenge it. Thesumming up is then finalized by the Secretary’sDepartment, in consultation with the area depart-ment, PDR, and the Director representing the coun-try concerned.13.The summing up is then transmitted to the coun-try concerned. In cases where that country agrees,the summing up, together with a brief summary ofthe factual portion of the staff report, also forms thebasis for the PIN. The draft PIN is submitted to theauthorities by the Director, and only exclusion ofhighly market-sensitive material (and corrections offactual errors) is in principle allowed.

Views on Article IV Missions

Frequency and Intensity

14.For most countries, Article IV consultations aresupposed to take place annually, although theprocess sometimes slips by a few months. Therehave been occasional attempts to reduce the fre-quency of consultations for smaller, “nonsystemic”countries, on a voluntary basis. However, in prac-tice, countries have been reluctant to agree to moveto biennial consultations; currently, only 23 of theFund’s members are on a biennial consultation

cycle. In financial year 1998/99, the Fund completed124 Article IV consultations, out of its total member-ship of 182. 15.Member states seem generally content with anannual consultation, although there are differencesin how countries approach them. Small countriesoften regard the annual Article IV consultation as avaluable opportunity for bringing together the eco-nomic policy establishment of the country. Forsome, this may be the sole such occasion. Largeand medium-sized member states felt that one an-nual visit (though not necessarily an Article IVmission) is about the minimum for maintaining thepersonal contact between Fund staff and nationalofficials, and the “feel” of the former for the coun-try’s problems, which are essential ingredients in aconstructive professional relationship. 16.However, differences of view were apparentover the desirable length and intensity of the mainannual mission, its coordination with regional ormultilateral efforts by the Fund, and the need for fol-low-up staff visits in between Article IVs.17.A normal Article IV mission spans at least twoworking weeks. While recognizing the value to do-mestic officials of discussions with a well-preparedFund team, most officials thought that the length ofmissions was excessive and could be shortened inthe initial fact-exploring phases. Essentially, ques-tions of fact (as distinct from interpretation) couldand should in their view be settled in advance.18. It is not uncommon for countries to receive fol-low-up visits from the staff once between the Arti -cle IV consultations. Some officials thought thiswas unnecessary, particularly since they also re-ceived visits from Fund staff responsible for theWEO and the International Capital Markets report(ICMR), as well as World Bank staff. Indeed, sev-eral country officials complained that they endedup answering the same questions repeatedly. TheAnnual Meetings (and in some cases the InterimCommittee meetings) provide another opportunityfor informal discussions. Quite a few thought thatthe Fund surveillance process was somewhat old-fashioned, with too much emphasis on large mis-sions and not enough on more continuous surveil-lance based on watching the country data fromFund headquarters, as investment banks or ratingagencies, for example, would tend to do.19. In terms of resources, it is useful to separate theFund’s membership into three categories.

• Small states. Country surveillance seems par-ticularly useful to, and appreciated by, smallstates; they strongly oppose any diminution inthe resources devoted to them.

• Industrial countries. There was general agree-ment among officials of all countries, staff,

26

30See Harper (1998), p. 254.


and outsiders that because of the systemic im-plications the Fund needed to monitor closelythe economies of the largest industrial coun-tries. This view was perhaps strongest amongofficials of those countries that relied on theFund for information about the largest coun-tries. However, some felt that the resourcesdedicated to industrialized countries, espe-cially the medium-sized and smaller ones,were too great.

• Emerging market countries. Many officials inboth industrialized and emerging market coun-tries thought that the Fund should give greaterpriority to emerging market countries, particu-larly those of systemic importance.

Contacts

20.The central bank is often the main contact, pro-viding logistical support, and sometimes providingthe venue for official meetings. Relations with cen-tral banks were generally good. Indeed, a concernvoiced by some national officials was that the Fundended up being too close to the central bank; in sev-eral of the countries we visited the Fund was seen bynonmonetary officials as insufficiently critical of thecentral bank’s well-known views in the domesticpolicy debate.21.The ministry of finance may also be the basiccontact point, but in any case it is always a mainstop for a mission. Although the line of the Fund isalmost invariably that the budgetary positionshould be strengthened, hence tending to reinforcethe position of the finance minister in conflicts withspending colleagues, the Fund appeared to get anapproval rating in finance ministries that was not ashigh as that in central banks. Sometimes this atti-tude was based on the perception that the missionwas insufficiently sensitive to domestic politicalproblems in obtaining support for budgetary con-solidation. Another contention was that the Fundwas too dismissive of Keynesian views of the im-pact of changes in budgetary policies. 22.Missions do aim to see important economic ac-tors in the government outside the central bank andthe ministry of finance—for example, ministries ofeconomic planning or of labor. However, coveragebeyond this list varies. In countries where the leg-islature has a central role in economic policymak-ing, some officials suggested that it might be use-ful for an Article IV mission to have theopportunity to exchange views with selected mem-bers of the legislature. This occurs in some coun-tries, but the practice does not appear to be wide-spread or systematic. In the United States, forexample, the mission does not meet with legisla-

tors although Congress clearly exercises a majorinfluence over U.S. budgetary policy. In Japan,even some officials from the central bank and theministry of finance suggested that the shift of in-fluence over economic policy and financial regula-tion from the Bank of Japan and the Ministry of Fi-nance toward the Diet would justify a meeting withthe mission. However, staff said that in some coun-tries the Fund’s traditional interlocutors prefer tokeep missions more to themselves. Similar consid-erations apply to meetings with representatives ofopposition parties; some missions have done this,but it is not standard practice. 23. In federal countries such as Brazil and India,where state or regional governments have substan-tial power over tax and expenditure policies, it maybe useful for missions to meet when they can withregional or local representatives. Fund practice hasmoved in this direction. Here, central governmentshave generally been supportive. We were told in acouple of instances that this was because surveil-lance had in such cases a strong element of technicalassistance to improve budgetary practice and/or debtmanagement in a way that was also helpful to thecentral government.24. Independent research institutions, financial sec-tor representatives, other private sector organiza-tions, and labor unions (in rough order of frequency)are also often included in the list of visits. However,the coverage of meetings with outside organizationsappeared to vary substantially from country to coun-try; it largely seemed to be determined at the discre-tion of the mission chief and the main governmentcontact. We also noted that in a number of countriesthe central bank or finance ministry made it a rule toaccompany Fund missions to all meetings, includingthose with nongovernmental organizations. Someoutsiders said that this inhibited them from talkingfrankly to the Fund. 25.We found that Fund interlocutors in the privatesector and in independent research institutions—andsometimes even within the government—were inone respect somewhat disappointed in the nature oftheir exchange of views with members of an ArticleIV mission. The Fund team asks good and searchingquestions to which those visited seek to reply to thebest of their ability, but there is reportedly little give-and-take in the discussion and the team leaves fewclues as to how the information provided is to beused in the staff report. Nor are the interlocutorsgiven an opportunity to see the final product or sim-ply the part of it to which they have contributed.Some said that missions would be better received ifthere could be more in the way of reciprocity. How-ever, the scope for such reciprocity is of course lim-ited when the staff consultations are supposed to beconfidential.

27


Dissemination of Surveillance Results

26.We discussed the communication of Article IVconsultations—be it through the PIN, the concludingstatement, or the full report—in most of our inter-views. We similarly reviewed dissemination and fol-low-up within government with officials. Since thechoices made in these areas are up to the individualcountry, rather than to decisions in the Fund on theconduct and methods of surveillance, we postponethe discussion of these important questions to Chap-ter IV, on the impact of surveillance.

Other Bilateral Surveillance

27.Country surveillance through the Fund makesuse of a number of other instruments in addition tothe Article IV missions and the whole structure ofintensive, but somewhat discontinuous, attention toa country to which the missions give rise.

Communications from Management

28. Management occasionally communicates in writ-ing with the central bank governor, the finance minis-ter, or even the prime minister, of a country where anArticle IV report or some follow-up to it gives causefor concern. Such messages, obviously confidential,are typically much clearer and more pointed thanstaff reports, let alone the Fund’s public statements.They no doubt serve to raise the attention level andthe eyebrows of national authorities; as a conse-quence, they are regarded as a tool to be used spar-ingly. The Managing Director also visits countries forconfidential discussions and/or public, national or re-gional, meetings; so does his First Deputy and, onrarer occasions, the two other Deputy Managing Di-rectors. Furthermore, they consult with high nationalofficials when the latter visit Washington to attendthe Interim Committee or for other purposes. Theseare valuable occasions for focusing on one or twocritical issues in relations with a country.

Informal Country Matters Sessions

29. Informal Country Matters sessions in the Execu-tive Board have taken place for a number of years.They are designed to allow the Board to discuss thesituation in countries whose economic situationgives cause for concern, in particular where a pro-gram may be in prospect. These sessions are held ir-regularly, on average monthly, with the list of casesmostly determined by the staff—most will be coun-tries already in programs—but Executive Directorsalso have an opportunity to add to the list. The lattersometimes gets so long—15 or so countries—thatthere is not sufficient time for each case to be dis-

cussed in depth. Executive Directors said that theyfound these sessions helpful, but would like to beable to examine key cases in greater detail. Severalalso observed that the staff made no attempt to pre-sent a regional view at these sessions, although thathad originally been stated as one of the objectives.One senior staff member also made the point that,from the point of view of the staff, it would be usefulif Executive Directors would use these sessions topass on any information they had from their capitalson the countries discussed. Apparently, this does nothappen.

Resident Representatives

30.Although resident representatives are not for-mally part of the Article IV surveillance process, anumber of officials expressed the view that theymade a significant and distinct contribution. Coun-tries generally welcomed proposals to appoint a resi-dent representative, and regretted their withdrawal.

Technical Assistance

31.Technical assistance was generally consideredhelpful. This was particularly the case when it waswell integrated with surveillance. For example, theBrazilian authorities were very complimentary aboutthe technical assistance they had received from theFund, especially on the fiscal side, and the way thatit had complemented bilateral surveillance moregenerally. However, even though opinions on techni-cal assistance were broadly favorable, successful in-tegration with surveillance does not appear to havebeen very common elsewhere.

Regional Surveillance

32.Efforts at regional integration, particularly in Eu-rope and to a lesser extent in Asia and in LatinAmerica, have raised the issue of whether the Fundshould engage itself more in surveillance of a re-gional nature. This question has become more topi-cal following the experience of the past two years,when aspects of the international financial crisishave taken on a regional dimension.

The European Union and the Euro Area

33.When regional integration advances as far as hasrecently happened in the European Union (EU), theFund has to make significant adjustments in itsmethods of surveillance. For a few years during thetransition to Economic and Monetary Union (EMU),Fund missions regularly visited the precursor of theEuropean Central Bank (ECB) and the European

28


Commission. We therefore visited the ECB and theCommission to discuss with leading officials howthey saw the future evolution of surveillance of theeuro area and its individual participants. We alsosought the views of some national officials from EUcountries inside or outside the euro area. 34.There are now plans to have a meeting at theECB with a Fund mission twice a year to review thecommon monetary policy and its exchange rate im-plications, and the ECB has been given observer sta-tus at the Executive Board. However, country sur-veillance of the 11 participants in the euro area willcontinue on an annual basis (except for Luxem-bourg, which is on a two-year cycle). Officials of EUcountries pointed to both formal and substantive ar-guments for continuing country surveillance of non-monetary policies. From a legal point of view, theparticipants continue to be members of the Fund andare hence obligated to receive the Fund’s Article IVmissions. For euro area countries, only monetarypolicy has been centralized (trade policy, of course,has already been centralized for all members of theEU). Budgetary and structural policies continue tobe a national responsibility, though the Pact for Sta-bility and Growth subjects such policies to intensivemonitoring by the Commission and by the Councilof Finance Ministers (ECOFIN).35. It is clearly in the interest of participants in theeuro area, officials from those countries told us, tohave the Fund apply the wide range of experience inits global membership to these policies. It may alsobe of interest to other Fund members to be kept in-formed about national policy developments in theeuro area with respect to tax, expenditure, and labormarket policies, not least because they are still farfrom uniform.36. However, a number of interviewees (official andother) suggested that the resources devoted to suchsurveillance could now be considerably reduced. Itwas pointed out that with a centralized monetary pol-icy there was no need to consult national centralbanks on this topic. Moreover, in areas other thanmonetary and exchange rate policy, EU states are al-ready subject to monitoring both by the EuropeanCommission and by the OECD. In the view of theseobservers, it should be possible for Fund surveillanceto rely largely on this work, albeit not in an uncriticalway. Accordingly, it was also argued, the Fund shouldfocus surveillance more on aggregate policies notonly in the monetary but also in the fiscal area, since itis the policy mix of the euro area as a whole that is ofconcern to the world as a whole. Here the main inputwould be the Stability Programs submitted to the Eu-ropean institutions. They also suggested that just asFund analysis of the United States and Japan has tra-ditionally devoted considerable attention to the policymix in these countries and its impact on other coun-

tries, this needs to be paralleled by a similar reviewfor the euro area. We develop in Box 2.2 some ideasconcerning the future subject material and organiza-tion of such surveillance.

Other Areas

37. In other areas of the world, regional integrationis far less advanced than in Europe and will remainso for a long time. The focus on regional surveil-lance in other areas is therefore not as strong. How-ever, member countries have taken a number of ini-tiatives in recent years, and the Fund has beeninvolved in several. We examined the Fund’s role inregional surveillance in Southeast Asia and LatinAmerica.31 These revealed both the potential andlimitations in such an approach.38. In Southeast Asia, the Fund has been designatedas the technical secretariat of the Manila FrameworkGroup. The Fund is already engaged in preparingbackground documentation for meetings of the Fi-nance Ministers of ASEAN, and the Managing Di-rector has occasionally been invited to participate inand address the meeting. These contributions arerecognized by country officials—and by the Fundstaff—to be valuable, though the Managing Direc-tor’s comments on impending crisis symptoms in thespring of 1997 were seen to have come too late tohave any major impact. Similar remarks were madeabout Fund input into the deliberations of the Asia-Pacific Economic Cooperation (APEC) officials,meetings that now include participants from some20 Fund members circling the Pacific.39.The Fund’s main contribution to such meetingsis the preparation of background documentation.However, we were informed that this documentationfocused not so much on regional interrelations andissues as on the individual country experiences—inother words, largely an updated replication of con-sultation material already available. In the view ofseveral Asian officials, notably in Japan and Thai-land, the Fund has been too ready to see difficultiesin stronger regional cooperation, particularly inAsia. The Fund has traditionally been wary of mu-tual support facilities among central banks and/orgovernments at the regional level—on the legitimategrounds that such initiatives must not be allowed toerode the role and the discipline of the global systemthat the Fund is designed to monitor and reinforce.This concern also marked the Fund’s attitude to a re-gional support mechanism in the earlier stages ofEurope’s monetary integration. But there might, in

29

31The Fund also has regular discussions with some other re-gional groupings, including regional monetary unions in Africaand the Caribbean. We did not examine these in detail.


the view of several Asian officials, have been netbenefits from encouraging a number of Asian coun-tries to develop some degree of monetary coopera-tion. At a minimum, better information flows in theregion might have reduced the risk of contagion,hence ultimately reducing the need for Fund resources.

40. In September 1998 the Fund convened a meetingin Washington of the finance ministers and centralbank governors of the Americas to review the short-term prospects of the hemisphere—not least in thelight of an impending crisis in Brazil. This meetinggot mixed notices; the staff prepared relatively elab-orate country notes for discussion, but felt that the

30

Box 2.2. Surveillance of the Euro Area

European monetary unification poses special chal-lenges for Fund surveillance, but also new opportunities.The Fund staff is recognized as objective and indepen-dent, and tougher in its analysis than European institu-tions. Unlike those institutions, it is not constrained tofocus on precise numerical criteria. So we believe thatFund surveillance will continue to have an important roleto play, provided it focuses on the area as a whole.

However, our suggestion that future surveillance ofthe euro area should focus on aggregate performanceand policies meets with practical difficulties. Withinthe European Union the monitoring by ECOFIN, pre-pared by the Commission, of policies outside the areaof monetary policy remains largely country based. Thisreflects the reality that budgetary and structural poli-cies remain primarily a national responsibility. Moni-toring of the aggregate performance of the euro area isnot yet well developed outside the European CentralBank (ECB), which has to review its monetary stanceon the basis of aggregate indicators. This asymmetry inthe policy framework implies that it will initially becomplex to focus Fund surveillance of the euro area onthe aggregate performance and the policy mix, the coresubjects for the Fund in conducting regional and multi-lateral surveillance. Furthermore, there is as yet no ob-vious authority to consult with outside the area of mon-etary policy.

This asymmetry in the policy framework is no acci-dent. It reflects the view, still broadly shared in the euroarea, that there is a much stronger economic case forcentralizing monetary and exchange rate policies in aregional institution than for centralizing other macro-economic or structural policies. The economic argu-ments are clear: (1) the introduction of the single cur-rency implies that participants in the euro area requiresome fiscal flexibility, to allow their budgetary policiesto absorb some of the residual divergence in nationalbusiness cycles; and (2) it was hoped that problems ofan inappropriate policy mix would not arise, becausethe combination of a monetary policy aiming at low andstable inflation and constrained national budgetary poli-cies would prevent serious policy conflicts. There werealso political arguments; the transfer of decision makingon fiscal policy is a more sensitive issue than for mone-tary policy, where governments and parliaments havealready delegated authority to national central banks;and concerns that the ECB would see its autonomythreatened by an ECOFIN with reinforced clout.

Fund surveillance may, however, gradually be facili-tated by an evolution in the policy framework, which isalready under way. Aggregate indicators of the perfor-mance of the euro area, notably the current account andthe exchange rate of the euro, increasingly find a role inpolicy reviews by the Commission and in the debate inECOFIN. Policy coordination—obviously so far on avoluntary basis, since the Treaty does not contemplatemore mandatory forms—is moving onto the agenda.The purpose of coordination will be both to review thepolicy mix in the light of aggregate performance and toconsider, in the light of that, the desirable speed of bud-getary consolidation within the 3 percent limit todeficits. The recent debate over the level of the Italiandeficit is a good example; in our view, illustrating thatthe broad stance of individual countries’fiscal policywill increasingly be influenced by what happens at theeuro area level.

Fund regional surveillance should not only anticipate,but take advantage of, this evolution. As noted else-where, surveillance has relatively little impact on indi-vidual euro area countries. For the reasons stated above,the Fund is well placed to give advice on the overallstance of euro area policies, both monetary and fiscal.To do so, visits to the ECB, the Commission, and na-tional authorities will not suffice. Fund consultationsshould be extended to ECOFIN and to the Economicand Financial Committee, as has already begun tooccur.1 The Managing Director might also usefully beinvited to attend meetings of ECOFIN once or twice ayear. In our view, given the increasing role of Europeaninstitutions not only in monetary but in fiscal policy,Fund surveillance is more likely to have an impact atthe euro area level than at the level of individual coun-tries.2 Consequently, Article IV missions to participantsin the euro area should become less frequent, more fo-cused, and leaner.

1We understand that such meetings took place in the June1999 consultations with the authorities of the euro area, andthe mission’s concluding statement therefore referred, appro-priately, to the “economic policies of the euro area.”

2This point is underlined by the fact that, in recent discus-sions of staff reports on individual euro area countries, euroarea Executive Directors have maintained positions collec-tively agreed at ECOFIN.

Fund Internal Organization and Procedures

effort might not have been justified since the processof peer pressure did not visibly help to bring aboutany adjustments in Brazil’s policies. Reluctance ofneighboring countries to engage in frank exchangeand criticism in such a forum was, we were told,characteristic of Asia also, and is apparently not un-known in European regional economic forums aswell. Apparently, the Fund’s bilateral “frankness”can stand out by comparison, and becomes rathersalutary. This being said, some Latin American offi -cials who had been present in Washington weremore positive, pointing to the usefulness of at leastsensitizing high U.S. officials to the problems oftheir southern neighbors.


41.The principal vehicle for the Fund’s multilateralsurveillance is the WEO, produced twice a year (andsometimes more frequently). In addition, the ICMRhas gained increasing attention in recent years. Bothare produced by the Fund’s Research Department.Conceptually, WEO has two main components: a de-scription of the current state of the world economy,including forecasts of the main economic variablesfor the world as a whole and for large and medium-sized countries, and analytic chapters on a variety oftopics. The forecast is prepared by a bottom-uprather than top-down procedure; that is, it is not pro-duced by a single model of the world economy, butcompiled from forecasts for individual countries byarea departments. These individual country forecastsare scrutinized by the Research Department to en-sure that they are both plausible and consistent, andif necessary changes are made after discussion be-tween the Research Department and the area depart-ment. The ICMR is also produced by the ResearchDepartment, but with relatively little input from areadepartments or the Monetary and Exchange AffairsDepartment (MAE). Again, it has both descriptiveand analytic sections. 42.As described in more detail in Chapter III, onsubstance, WEO and the ICMR are generally heldin high regard among officials and others who mon-itor major international economic developments.We did not examine the process by which they areproduced in as much detail as for bilateral surveil-lance. However, the following points did emerge inour discussions:

• Some of our interviewees, government officialsas well as academics, suggested that quarterlypublication of an updated and leaner version ofWEO would be desirable, with maybe one ofthem as a larger annual overview. The Fund hasmoved in this direction by publishing toward

the end of 1997 and 1998 an update of the as-sessment in the October WEO (and ICMR).

• A number of senior staff and some governmentofficials raised the question of the interactionbetween multilateral and bilateral surveillance.It was suggested that the two processes tendedto proceed on parallel, but largely separate,tracks, with area departments managing bilateralsurveillance and the Research Department re-sponsible for WEO and ICMR, and that agreater degree of integration would be desirable.

43.We heard favorable comments on the OccasionalPaper series, which has been a useful vehicle for dis-seminating, for example, cross-country experiencewith different exchange rate regimes and capital ac-count liberalization, two essential elements in thesubstance of Fund advice in country surveillance(see below). 44.Besides the publicly available information, mul-tilateral surveillance also takes place inside theFund; the chief vehicle here is the World Economicand Market Developments (WEMD) meetings, heldapproximately every six weeks, at which the Eco-nomic Counsellor presents his view of the worldeconomy and of certain key countries (usually themajor industrial countries and countries where thesituation gives rise to particular concern). Finally,the Fund also has a role in the G-7 process. It pro-vides a background note on the economic situationin the world and in each of the G-7 countries. At themeetings themselves, the Managing Director (andthe Economic Counsellor at the G-7 Deputies meet-ing) gives a summary of the Fund’s views on the sit-uation. However, they are not present for all of thesubsequent discussion and play no part in draftingthe communiqué.

Fund Internal Organization and Procedures45.A number of our interviewees, particularly—butby no means exclusively—present and former staffmembers, raised internal and organizational issuesthat affect surveillance. Most of these relate primar-ily to bilateral surveillance, although some raisewider issues that affect the effectiveness of the Fundoverall.

Staff Continuity

46.Many, perhaps most, government officials ob-served that there was a lack of continuity in the per-sonnel involved in bilateral surveillance. Staff, espe-cially junior and midlevel staff, rotate frequently.While staff believe that this has important advan-

31


tages in career terms, it causes irritation amongcountry officials, who sometimes feel they are an-swering the same questions on each mission. Thereis usually (but not always) more continuity at thehead of mission level. The head typically partici-pates in a minimum of two or three visits. It was alsopointed out that sometimes the head of mission isnot the division chief for the country concerned, butsomeone from the front office in the area departmentwho will be more senior, but sometimes lacks famil-iarity with the country. However, staff pointed outthat often such problems are at least in part the re-sponsibility of the country, since some authoritiesconsider it a slight if the mission head is not, in theirview, sufficiently senior.32

The Skills Mix

47.The majority of Fund economists are recruitedeither more or less immediately after completing aPh.D. (usually in macroeconomics with a specializa-tion considered relevant by the Fund) or relativelyearly in their careers, with some previous experi-ence, often in finance ministries, central banks, orresearch institutions. The quality of the intake is un-doubtedly high, though a couple of leading academ-ics we interviewed thought that the Fund had beenless successful than a decade or two ago in attractingthe very best graduates from the top universities. Inany case, this recruitment policy has certainly pro-duced a relatively homogeneous staff with highcompetence in the core areas of the Fund. 48.However, this situation also gave rise to someconcerns among people we interviewed. These basi-cally revolved around the question of whether coun-try policy experience was particularly valuable insurveillance. The point was made to us by quite afew staff that because outside experience is, by defi-nition, from the outside, and because the Fund has astrong culture regarding the way things should bedone and the need to build up relevant Fund experi-ence to do them properly, the value of significantoutside policy experience (be it from a central bank,economics ministry, or a private financial institu-tion) is undervalued. This was particularly the casefor those recruited rather later in their career than thecore group described above. On the other hand, inmany of the countries we visited, officials were atpains to point out that they valued dealing with staffwho had themselves been in a country policy posi-

tion. They thought it added something to the qualityof the policy dialogue.49.Another, somewhat related, issue was the needto ensure adequate career opportunities for staff withdifferent skills. Here it was felt that a large premiumwas placed on macroeconomic expertise relative toother areas, with the result that other skills were un-derappreciated, making it difficult to recruit and re-tain staff who possessed them. This gained addedimportance with the Fund’s increased emphasis onfinancial sector matters, where macroeconomics isless central.

Interdepartmental Relations and the Review Process

50.There is a tradition of strong departmental iden-tity and autonomy in the Fund. As a consequence,there is, as there should be, vigorous debate amongdepartments on important issues. However, a num-ber of staff members (present and former) expressedthe view that interdepartmental relations were not asconstructive as they could be and that this had an ad-verse impact on the effectiveness of surveillance. Insome cases, the relationship was adversarial; in oth-ers, distant. This situation obviously does not prevailfor all departments or at all times. However, a num-ber of comments we received, including from verysenior staff, indicated that the issue was of concern. 51. In terms of its direct effect on surveillance, themost striking example was the generally poor rela-tionship (in the period we were studying, notably therun-up to the Asian financial crisis) between the Re-search Department, responsible for capital marketssurveillance, and the area departments responsiblefor country surveillance, especially the Asia and Pa-cific Department. All sides must share responsibilityfor this. Disagreements between the two departments,and the consequent breakdown in communication,were at least in part responsible for the fact that con-cerns about the health of Korea’s financial systemwere not properly reflected in surveillance nor com-municated to the Executive Board (see also the dis-cussion of Korea in Box 3.2). This situation may alsohave reduced the effectiveness of the Fund’s surveil-lance of other Asian countries, such as Japan. Internalprocedures in this area appear, if anything, to havedeteriorated since; the confidential “Financial SectorNotes” circulated to management and other depart-ments by the Capital Markets group of the ResearchDepartment were discontinued in late 1997.52.More specifically, we also encountered a numberof problems in the functioning of the interdepart-mental review process. While, we should emphasize,there is no suggestion on our part or on that of othersthat the review process should be discarded, therewas a general concern that the process was becom-

32

32We learned from the staff that the mission head’s seniorityhad been an issue for one of our sample countries (Korea) in theperiod leading up to a crisis. This had delayed the consultationand had, in the staff ’s opinion, reduced the effectiveness of sur-veillance in this instance.

Role of the Executive Board

ing very time consuming and inadequately focused.Concerns were expressed from the sides of both areadepartments (for example, too many comments thatwere more form or detail than substance) and func-tional departments, where there was an impressionthat area departments did not take their commentsseriously enough.33 Some further stated that the situ-ation had worsened in recent years, as the Fund, andthe coverage of staff reports, had expanded and moredepartments had therefore been drawn into theprocess. However, departments did not feel con-strained to limit their comments to their areas of pro-fessional expertise—again adding to the workloadall around for no particularly beneficial result. 53.While the review process does of course seek toensure that the Fund’s views or “line” are properlyreflected in the brief,34 concerns were also expressedto us that the process hinders innovation and flexi-bility; departments are inhibited from trying to dothings differently. Some interviewees suggested thatthis tendency may also lead to insularity—one of themost common external criticisms of the Fund, bothamong our interviewees and in general—becausestaff are preoccupied with getting their analysisagreed and accepted internally rather than listeningor learning from outside.

Other Organizational Issues

54.Here, brief mention is made of a number of otherorganizational challenges for the Fund. Their effecton surveillance is indirect, but they were raised fre-quently in our interviews within the institution.

• Hierarchy and accountability. A number ofconcerns were raised here—notably a ten-dency to centralize decision making and a re-luctance to delegate. This tendency, we werealso told, leads to a general lack of individualaccountability and responsibility. Several ofour interviewees, inside and outside the Fund,felt that it was notable that no one had beenheld accountable for what they saw as the fail-ures of Fund surveillance in Asia.

• The number of senior staff in area departmentfront offices. While the numbers of front officestaff look large by outside standards, and quitea few staff raised doubts about the necessity ofoffices of this size, we were told by senior staffthat the workload, given the sheer number ofcountries in each department and the needs of

Fund-wide coordination, justified these num-bers. And in any event, we were told, somenominally front office members were in effectoperating as division chiefs. In any case, it wasapparent that not all front offices were able toprovide what one senior staff member, rightly,believed were their central functions: first, toprovide an overview of the department’s workacross countries; and second, to bring a per-spective on what was going on in other depart-ments and outside the Fund that would be use-ful to the department.

• Finally, the team was struck by the burden ofoverwork that prevails among staff involved inthe surveillance process: this has been a long-standing issue, but it seems to have worsenedin recent years. Above and beyond the conse-quences for personnel, it surely detracts fromthe quality of surveillance overall.

Role of the Executive Board

55.The role of the Executive Board in the conductof surveillance is, in principle, paramount. It is theBoard’s discussion, and the summing up thereof,that completes the consultation process with a coun-try or that authorizes the release of publishable workby the staff. We sought comments from a number ofExecutive Directors (EDs), from senior staff andmanagement, and from national officials with per-sonal experience of participation in the Board.56. Views in general, both among current EDs andothers, were negative both about the process and theresults. The following points in particular were raised.

• The size of the Board, although understandablefrom the viewpoint of membership representa-tion, makes a free-flowing or well-focused dis-cussion very difficult. On important countries orissues, most or all EDs will speak, often fromprepared statements. (Indeed, silence is offi -cially consent: if a Board member does notspecifically dissent from the staff view, he orshe is normally recorded as endorsing thatview.) This can lead to sterile discussions,where one Director after another opines on eachelement in the staff report: monetary policy, fis-cal policy, exchange rate policy, etc. As can beimagined, this is neither enjoyable for the par-ticipants nor useful for the country.35

33

33This latter point does not apply to PDR, which has to sign offon every brief.

34At the same time, it was also suggested to us that in terms ofthe PDR involvement in the review process, there was now lessof a “line” to be adhered to than earlier.

35While no one suggested that the problem did not still exist,some thought that it had lessened somewhat recently, especiallywith the increased use of preliminary statements (“greys”) dis-tributed in advance of the meeting.


• The workload is huge. The Fund has 182members, most of them on an annual consulta-tions cycle. So bilateral surveillance alonegenerates more than 150 substantive agendaitems annually, and there are also numerousmultilateral surveillance items. Program coun-tries, of course, generate even more work thansurveillance-only countries, not to mentiondiscussion of general issues, administration,and the like.

• Staff members observed that EDs tended to bedefensive about the countries they represent,and that other EDs deferred to this, partly be-cause they expected the same deference in re-turn in due course. As the internal review ofthe Mexican experience put it, peer pressurecan become peer protection.

• Some staff and EDs felt that EDs rotated toorapidly, so that they could not develop the in-stitutional knowledge to contribute to theBoard. In part, this is a consequence of themulticountry constituency system, when EDsfrom different countries rotate every two yearsor so.

• The Board is chaired by management. SomeEDs regarded this as anomalous, potentiallyputting management in a difficult position if astaff paper is strongly criticized.

• The Fund staff and management attach consid-erable importance to presenting a unified viewin Board discussions. Internal disagreementsare generally not divulged to the Board. SomeDirectors thought it could be healthy for this tohappen more frequently.

Directors noted that efforts to address the first twoproblems by introducing a committee structure hadnot made much progress. Even when committeeswere set up—as for some administrative matters—they were open, that is, any ED could attend (or senda representative) and speak. In the event, little realstreamlining was accomplished. Some EDs thoughtthat their colleagues would insist on making use ofsuch an open access provision to a committee thatdiscussed Article IV reports.

Some EDs complained that they had relatively lit-tle influence on the policy positions taken by staff.Others, however, pointed to one or two occasionswhen advice had been modified by the Board. (Inter-estingly, these were cases in which the staff had rec-ommended devaluation to countries with pegged ex-change rates.) We did examine the mechanism bywhich the views of the Board were transmitted backto staff at the working level. In the surveillance con-text, the primary formal mechanism by which thishappens is the summing up; the staff report always

begins by reviewing what the Board said in the pre-vious year. So, to the extent that the Board views arestill relevant, the staff will try to incorporate theminto the report. Staff we spoke to said that they didtry to do this; they did not want to be embarrassed byhaving the Board ask the same question, or repeatthe same criticism, two years in a row. However, theprocess was less systematic for cross-country issues.Some Board members suggested that if they couldsee the mission briefs in advance of Article IV mis-sions, this would enable the Board to have moreinput into the conduct of surveillance than it doesnow, when it only sees the final staff report. (Indeed,as noted earlier, the mission brief—which sets theagenda—is arguably more important than the finalreport.) However, some staff and national officialssaid that such access would, among other things,raise confidentiality concerns.

This brings us, finally, to the delicate issue ofequality among the EDs. The resources they can de-vote to the monitoring and anticipation of papers andinitiatives by staff and management clearly vary.Some EDs also have substantial resources to backthem up with analytical and policy efforts in their re-spective capitals. It is no surprise, therefore, that theperception of many EDs is that information on im-portant aspects of surveillance, on policy initiativesdeveloping in the Fund, and particularly, on thepreparation of programs, is not equally accessible toall shareholders. In particular, there is a firm impres-sion among developing countries that the G-7 coun-tries have disproportionate access, information, andinfluence; and a perception among many coun-tries—including some industrialized countries—thatthis is doubly the case for the United States.

There is no doubt that the U.S. Executive Direc-tor’s office and to a lesser extent those of some otherG-7 countries are better informed about and haveprovided more input into Fund policy than their col-leagues. U.S. officials, indeed, seem more capable of“working the system”36 to advance their preoccupa-tions in selected cases through contacts to staff earlyin a consultation process or in the preparation of apolicy paper. It is less clear, according to our infor-mation, to what extent this asymmetry is due togreater resources and capacity to take initiatives orrather to differential treatment by management andstaff. To some extent this is inevitable; all Fundmembers are not of equal weight, as the Board vot-ing system recognizes. However, the tensions causedby this asymmetry do apparently detract from colle-giality within the Board.

34

36The expression was used by the Deputy Secretary of the U.S.Treasury. Lawrence Summers, Remarks to the Senate ForeignRelations Subcommittee on International Economic Policy andExport/Trade Promotion, January 27, 1999.

1. This section deals with the quality, relevance,depth, and scope of Fund advice. The questionsposed by the Board that are most pertinent here arethe following.37

(i) How did the Fund’s advice correspond tothe short-term objectives and medium-term strategies of existing policies?

(ii) How did this advice correspond to theanalysis and advice of other domestic andinternational institutions? Did the Fund’sadvice add value?

(iii) Has Fund surveillance paid sufficient at-tention to regional surveillance, to interac-tion among countries, and to the externaleffects of policies in major countries?

(v) Have the frequency and general focus ofthe Fund’s surveillance been appropriate,with hindsight? Has advice been consis-tent? Has advice helped foster noninfla-tionary economic growth?

(vii) Did the advice take into appropriate ac-count the institutional, political, and socialframework? Did it pay adequate attentionto the uncertainties and political con-straints that lead to “small” deviationsfrom first-best policies?

2. In this light, we first discuss bilateral surveil-lance, beginning with what we were told about itsgeneral quality and consistency and following witha more detailed review of individual policy areas.Then we look at the scope and coverage of Fundsurveillance. This is followed by a brief review ofspecial considerations that might relate to surveil-lance of small states and a discussion of multilat-eral surveillance. We conclude with a brief compar-ative survey of other institutions’surveillance-typeactivities.


Quality

General View

3. Fund surveillance was generally regarded asbeing of high quality and as a process of exchange ofviews that many officials found stimulating. To cite,just as one example, the views of a couple of experi-enced European central bank officials, “the Fundconsultation discussions are stimulating and alwaysto the issues,” and “the staff talk to many people andfinish up with a view that is comprehensive and in-dependent.” However, the high praise tended to belimited to the Fund’s work in the macroeconomicarea. In more microeconomic, less demand-oriented,areas such as financial sector surveillance and struc-tural questions, views were more critical.4. Perhaps not surprisingly, therefore, the most favor-able appraisals came from those whose lines of workbore close similarities to the Fund’s—central banksand, to a lesser extent, finance ministries. To the extentthat one can generalize about less positive reviews, itis fair to say that these tended to come from policy-makers in areas other than demand management, andparticularly from those who, for example, were notparticularly taken with the prescription of cautious fis-cal policies and monetary policies aiming at low infla-tion, combined with liberalization of product and fac-tor markets, including removal of all capital controls,sometimes labeled the “Washington consensus.”5. Many favorable comments stressed that Fund staffwere high quality, hard working, and well prepared. Ingeneral, staff visiting a country were seen to have aclear and reasonably comprehensive understanding ofthe principal economic problems. However, two criti-cisms recurred in a large number of our interviews, es-pecially in nonindustrial countries:

• lack of flexibility in the Fund’s analysis; and

• failure to appreciate adequately the politicalenvironment in which decisions are taken,and/or to allow for it in policy advice.

6. On the first point, while most of these criticsagreed that it was useful—and intellectually stimu-

III Substance

35

37Numbers in parentheses indicate the number assigned to thequestion in the original Executive Board request. See the Termsof Reference in Appendix 1.

III SUBSTANCE

lating—to exchange views with the Fund, they wereleft with the general impression that missions cameinto countries with a preconceived template of ideas,based on a theoretical or textbook model, housed inWashington, into which they fed country informa-tion. This was perhaps particularly true in areas out-side the Fund’s traditional expertise, such as labormarket or tariff policy.38 In their opinion, the result-ing policy recommendations did not adequatelyallow for, or perhaps were not permitted to allow ad-equately for, differences among countries. Thistended to lead to a “one-size-fits-all” approach thatthey thought weakened the effectiveness of the exer-cise. As one Latin American official put it, “theycome with a preconceived image of what the issuesare, and this makes it difficult to look behind the sur-face to the real problems.” 7. It should also be noted, however, that this apparentrigidity may still have a positive impact, especially inthe core macroeconomic areas, as the following re-vealing quotation from an Asian official shows:“Strangely, it is worthwhile to go through the ArticleIV consultations that dispense this very standard, un-differentiated advice. It is like a confessional—thegovernment sits on one side of the booth and the IMFis on the other side. Each knows what the other isgoing to say, but by repeating the lessons of an intro-ductory course in macroeconomics to the sinning gov-ernment, the IMF priest does help limit temptations todeviate too far from the orthodoxy.”39 This view of theFund as a useful reality check was not uncommon.8. The second criticism just noted related to thepressures applied by the Fund to achieve speedy im-plementation of “first-best” recommendations, with-out giving sufficient weight to political or institu-tional constraints. The authorities in one countrywere particularly upset when the mission’s conclud-ing statement suggested that legislation, recentlypassed with considerable difficulty and after muchdebate, “should be repealed immediately.” Althoughuse of such patronizing language seems to have beenexceptional, there was, nonetheless, a perception—although not a universal one we should point out—that Fund staff did not sufficiently see it as their func-tion to come up with policies that, while less thanfirst-best, moved the country in the right directionand were politically and institutionally feasible.40 At

the same time, it was also pointed out to us that thisargument should not be used to justify insufficientmovement in policy to ward off a looming economicor financial problem.9. On forecasting, the Fund’s work was generallyconsidered to be reasonably accurate—not excep-tionally good. If there was any bias, our interlocu-tors thought it tended toward optimism. Privatesector observers in particular thought that this re-flected the fact that the Fund was not good at look-ing for, or incorporating, information and data fromsources other than the authorities. At the same time,it should be noted that in one of our sample coun-tries the staff resisted considerable pressure to pro-duce a more favorable forecast, one that would bemore in line with the domestic authorities’publictarget. 10.Another concern that should be ventilated hereis one that featured strongly in the Fund’s internalreview of surveillance following the Mexican cri-sis—namely, the tendency of the Fund, particularlystaff in area departments, to be insufficiently frankand direct in its assessment of a country’s policiesor economic situation (a culture of “clientism”).Not surprisingly, opinion among our intervieweeswas sharply divided on this topic. Only a few gov-ernment officials whom we quizzed directly on thisthought that the Fund was insufficiently frank intheir own country’s case (although some officialsin Thailand conceded that in retrospect the Fundshould, if anything, have been stronger in its lan-guage earlier). However, most if not all staff andformer staff—including those in area depart-ments—agreed that on the evidence this remained aserious problem for effective surveillance, as did anumber of outsiders. In summary, it was far fromclear that matters had improved much since theMexican crisis. 11. Some blamed the incentive structure for thisproblem. For example, a former Fund official saidthat staff got ahead not by challenging or criticizingcountry policies, or by asking hard questions, butrather by “not rocking the boat.” A senior staffmember said that staff were still far too reluctant torisk forecasting a crisis that did not in fact happen.Many staff, especially in area departments, whilerecognizing the criticism, put the blame on officialsin member countries. They said that a missionleader who was regarded as highly critical of gov-ernment policy and insufficiently diplomatic inpresenting such views could expect that next timehe or she would not be granted meetings with moresenior officials. This would prevent them fromdoing their job properly. In more extreme cases, ithad been known for government officials to com-plain to the mission leader’s superiors. Either way,this was bad for mission leaders’careers.

36

38See also the discussion of capital controls in Chile later inthis chapter.

39This comment was actually made to one of the external eval-uators of Fund research activities. Coincidentally, the official wasfrom one of our sample countries.

40By contrast, in one country the Fund, while critical of a par-ticular political commitment made by the government policy,provided detailed advice on its implementation that also softenedthe rougher edges of the measure. This was regarded as construc-tive and helpful.


Changes Over Time

12.Since opinions on the quality of Fund adviceover time differed significantly, it is difficult to gen-eralize. However, there appeared to be a fairly broadsense among interviewees that surveillance had be-come more difficult to do well, largely because ofthe capital account and financial sector issues dis-cussed below. 13.Some current staff members and some govern-ment officials believed that the quality of surveil-lance had improved significantly in the last fewyears. They judged that it had managed to adapt itsfocus to ensure that it kept up on the right issues, andthat the main emphasis of discussions had shifted toinclude more discussion of those issues. However,other officials and some former staff members wespoke to thought that the arguments of the Fund hadbecome more routine, relying on textbook-like posi-tions in the face of situations that were not coveredin the textbooks. A view we heard was that discus-sions of possible different techniques in coping withpolicy challenges tended to end too early, with mis-sions being less open to new approaches to dealingwith problems. One interpretation was that a lower-ing of quality in advice had coincided with the rapidincrease in Fund membership since the early 1990s.The resulting staff expansion meant that latterly staffwere on average less experienced.

Consistency

14. If anything, the concern in countries we visitedwas that there was too much consistency in Fund ad-vice rather than too little. As mentioned, a frequentcriticism related to lack of recognition of country-specific characteristics. A central banker from a non-industrial country considered that “it would be betterif the Fund took more account of the differencesamong regions and economies. Sometimes the pol-icy advice is weakened because what it provides ismerely a copy of advice elsewhere.” Some officialscomplained that the Fund was tougher on small andmedium-sized countries than on, say, the UnitedStates or Germany. What we can add here is thatwhile it may be debatable whether what the Fundsays is tougher, what is less open to debate is that ad-monitions from the Fund probably matter more insmaller countries. 15.There was also a general feeling (particularlyamong staff members, but also among officials andother outsiders familiar with the institution) that theFund devoted a disproportionate share of surveil-lance resources to precisely those countries thatneeded them least and where the impact was least,that is, the G-7 and medium-sized industrializedcountries, especially in Europe. This applied to both

the quantity and quality of resources—a number ofstaff members observed that the most prestigiousarea department, and in their view the best staffed atmiddle and junior levels, was European I (dealingprimarily with Western Europe).

Specific Policy Areas

Monetary Policy

16.On monetary policy we can be very brief.While there were occasional criticisms (fromwithin the Fund as well as outside) of Fund tech-niques of analysis (such as an excessive relianceon “net domestic assets” as a tool of analysis; inad-equate recognition as to how the demand formoney can shift as a result of financial policy re-form; and an insufficiently broad appreciation ofwhat might constitute a domestic “nominal an-chor”), and while of course there was not alwaysfull agreement with the particular advice that wasgiven, most felt that the advice and the technicalassistance given on monetary issues were usefuland of high quality.

Fiscal Policy

17. Fiscal policy is front and center in most Fund doc-uments, and there was general agreement that this wasappropriate. The advice and, we should emphasize,the technical assistance provided are generally highlyvalued and highly rated. Furthermore, the Fund isseen as more sophisticated—and diplomatic—in deal-ing with fiscal issues now than in past decades.

Particular favorable mention was made of effortsdirected at:

• making public finances more transparent;

• getting more complete public sector accounts;

• developing and clarifying the concept of a“quasi-fiscal deficit” stemming from centralbank operating losses; and

• sorting out the analysis of structural and cycli-cal factors in government finance.

Some less positive comments were made, to theeffect that

• the staff did not have a good understanding ofthe redistributive effects of some of the fiscalpolicy measures it proposed—or of the need totake such effects into account, and that

• missions tended to recommend fiscal tighten-ing almost as a matter of principle, not distin-guishing sufficiently between situationswhere tightening was urgent and essential,and others where it would merely be helpful.

37

III SUBSTANCE

This made national authorities discount ad-vice in the former situations.

18.Finally, mention should be made of the sharp in-ternal Fund debate in recent years over its fiscal ad-vice to Japan. Without venturing into the details ofthis debate, what it makes clear is that there is stillample room for disagreement over the potency offiscal policy and the desirable mix of policies, evenamong economists who share the same broad ap-proaches and institutional objectives. Furthermore,the Fund’s “line” on Japanese fiscal policy in recentyears has shown a tendency to fluctuate. For a longperiod up to 1997, the main recommendation wasthat the Japanese authorities should not delay bud-getary consolidation—though the line wavered asthe Japanese economy ceased to grow for severalyears. When growth appeared to pick up in 1996 andearly 1997, the Fund strongly supported raising theVAT rate from 3 to 5 percent. After the economyturned down later in the year, the Fund recognizedthis step as a mistake and subsequently advocated—in increasingly strong language—a budgetary stimu-lus despite the by then very large deficit.

Exchange Rate Policy

19.The analysis of exchange rate policies and ap-propriate exchange rate levels is bound to occupy acentral role in surveillance. At the same time, whatconstitutes good, or even appropriate, exchange ratepolicy has been a question dividing the academiccommunity and a prominent question in debateswithin the Fund. The institution’s present position is,by all reports, quite pragmatic. Indeed, from differ-ent sources we heard criticism that the Fund was ex-cessively keen on exchange rate flexibility; that itwas too much in favor of fixed rate systems; and thatit had no clear line on the topic. 20. Most observers were happy with the pragmatismshown, since it gave countries an opportunity to selecta system that they thought was attuned to their historyand characteristics. Accordingly, the Fund has tended,at least in nonprogram countries, to go along with na-tional preferences in terms of the exchange rate sys-tem adopted.41 However, it is also worth noting that ithas seemed to emphasize pretty consistently, and ap-propriately, the need for other policies—basically fis-

cal, monetary, and wage policies—to be consistentwith the exchange rate system chosen. 21.That being said, some of those interviewed werecritical of what they saw as a lack of a consistent ap-proach to exchange rate policy. The contrast be-tween the high degree of flexibility recommended inthe case of some Asian countries, while Russia wasbeing encouraged to defend its nominal exchangerate peg came in for much comment. An academicobserver noted that the Fund had bitterly opposed acurrency board in Indonesia while effectively impos-ing one in Bulgaria, but in his view had not ex-plained adequately the reasons for these different approaches.42

22.Many of the countries selected as part of our“representative sample” (but not selected for thisreason) had traumatically abandoned their peggedexchange rate systems after prolonged experienceswith them (Brazil, Chile, the Czech Republic,Korea, South Africa, Sweden, and Thailand). Manyof those interviewed in these countries—even thosewho had earlier supported the pegged rate, and stillthought that pegging was constructive in the rightcircumstances—criticized the Fund for not pressingthem enough to abandon the peg at an earlier stage.The need to discuss exit strategies and alternativeregimes was frequently mentioned. Several referredto a recent Fund Occasional Paper on this topic43

that was felt to be a very useful reference point. 23. Some, both in governments and academia, notedthat the current conventional wisdom on exchangerate systems seemed to be shifting toward the viewthat floating or firmly fixed exchange rates were dis-tinctly preferable to anything in between. This wouldappear to be the emerging view in the Fund as well.44

Capital Mobility and Capital Account Convertibility

24.The issue of capital mobility and capital accountconvertibility remains a lively topic within and out-side the Fund. It has persistently divided both theacademic community and government officials, be-coming even more contentious since the Asian crisis.

38

41One notable exception was the run-up to EMU, and in partic-ular in the aftermath of the ERM crises of 1992–93, when Fundstaff and management appeared to take a rather skeptical viewboth of EMU’s feasibility and its desirability, suggesting insteadthat more exchange rate flexibility would be beneficial. Whateverthe analytical merits of this view, it clearly tended to reduce theimpact of Fund surveillance. However, with the introduction ofthe euro, this is water under the bridge.

42Note, however, that these commentators were not making thedistinction between program and nonprogram countries that theevaluation team was asked to adhere to.

43Barry Eichengreen and Paul Masson, Exit Strategies: PolicyOptions for Countries Seeking Greater Exchange Rate Flexibility,IMF Occasional Paper No. 168 (Washington: IMF, 1998).

44See, for example, Stanley Fischer, First Deputy ManagingDirector, “The Financial Crisis in Emerging Markets: SomeLessons,” speech to the Economic Strategy Institute, April 28,1999. “We are thus likely in coming years to see more countriesadopting flexible exchange rate systems or, if they choose to fix,to do so in a definitive way, for example by adopting a currencyboard arrangement.”


25.Many of those we spoke to were critical of theFund’s advice in this area. There were two main crit-icisms. The first was that the Fund, being more wed-ded in general to analysis of flows rather than stocks,simply did not take capital account issues seriouslyenough, and as a consequence its surveillance in thisregard had been too sanguine. For example, one par-ticularly outspoken individual, in the financial sectorin Europe, thought that the principal weakness of Ar-ticle IV consultations was “the lack of depth andrigor in the treatment of capital account issues.” An-other comment, from a senior Asian finance official,was that “the Fund has not yet adjusted to an erawhen the capital account has become prominent inthe diagnosis of, and remedies to, foreign exchangecrises.” Some observers thought that this lack of at-tention to capital account developments helped toexplain the Fund’s failure to foresee the Asian crisis.(We should in fairness note here that no institution,public or private, can be reckoned to have genuinely“foreseen” the Asian crisis.)26.A second criticism was that while the Fund hadbeen very opposed to the use of controls to limitshort-term inflows, its opposition was based more onideology than on a careful consideration of the evi-dence and the policy alternatives. While the Fund’sline on this has changed recently, the general percep-tion is that the Fund lagged rather than led the gen-eral consensus on this topic.27. In this context, the case of Chile is particularlyinteresting. The Chilean officials we spoke to werefamiliar with the general arguments in favor of free-dom of capital movements, and therefore believedthey understood why the Fund favored the removalof controls. However, they were frustrated by whatthey saw as a failure of Fund staff to appreciate whytheir controls were a reasonable second-best responseto the problems they faced—pointing out that theFund did not produce any empirical evidence aboutthe costs and benefits of Chile’s control regime.Given this, they thought that the advice on the topicwas neither convincing nor useful. More recently theFund’s interpretation of the Chilean experience hasbecome distinctly more positive; the Fund now as-sesses the main impact to have been a modification inthe maturity of foreign debt rather than in its totalvolume, resulting in a helpful reduction in the vulner-abilities in the Chilean financial sector.45

Financial Systems

28.Everyone who spoke to this topic agreed thathaving the assessment of financial systems and theirvulnerabilities become an integral part of Article IVconsultations was highly appropriate. 29. In our interviews, the Fund was often criticizedfor not having alerted countries to financial weak-nesses and the imminence of a financial crisis. Forexample, it was marked down for not having warnedSweden about the weakness of its financial systemtoward the end of the 1980s and the possibility of afinancial crisis, notwithstanding the fact that similarcrises had already occurred at the time in Argentina,Chile, Finland, and Norway. It was also taxed withhaving missed the same kind of problems in Koreaand Thailand, notwithstanding the experience ofMexico. As regards Korea, while the Capital Mar-kets group had identified the weakness of the do-mestic corporate sector, and the potential implica-tions for the banking system, the Asia and PacificDepartment—and consequently its Article IV staffreports—had not focused on this area. 30.Clearly, the Fund has not performed well inspotting mounting weaknesses in financial systemsbefore they trigger crises.46 But it is also fair tonote, although this cannot be a full excuse given theFund’s acknowledged responsibilities and store ofinternational knowledge, that many officials in therelevant countries also allowed that they had notbeen aware of the importance of this issue. Somehad thought that their financial systems were ingood shape.31.Looking ahead, a number of those interviewedthought that it was appropriate for the Fund to becharged with monitoring compliance with the codeof good practices on monetary and financial policiesthat is now under discussion. But there was less con-sensus that the Fund should be in charge of its de-sign. This task was best left in the hands of the Bankfor International Settlements (BIS) or a committeeof international financial regulatory authorities,some suggested. For one thing, rule making (regula-tion) is a conceptually different exercise from ruleenforcement (supervision). For another, such a codewould not be useful unless there was a feeling of“ownership” on the part of those who would actually

39

45See, for example, Stanley Fischer, First Deputy ManagingDirector, “Reforming the International Monetary System,” DavidFinch Lecture, Melbourne, November 9, 1998: “We see no casefor controlling long-term inflows, particularly of foreign directinvestment, but can see the disadvantages of surges of short-termcapital, both inflows and outflows, and therefore can supportmarket-based controls, along Chilean lines, that are intended todiscourage short-term inflows.”

46However, we should also draw attention to the Fund’s analy-sis of the framework for financial supervision in the euro area(see International Capital Marketsreport, 1997 and 1998), whichwarned of the potential risks of a lack of coordination of nationalauthorities and of the absence of a clearly defined lender of lastresort. While ECB officials believed this concern to be over-stated, many in academic and financial circles shared the Fund’sconcern. At a minimum, the Fund’s timely probing prompted theECB to clarify its position.

III SUBSTANCE

have to live by it. The Fund was, in their view, un-likely to supply that “ownership” feeling.32. In the same vein, there was also widespreadagreement that while the Fund could well have a re-sponsibility for monitoring the vulnerabilities of in-dividual financial systems, this task could not betaken to imply a permanent in-depth analysis of allthe microeconomic aspects of various financial sys-tems. The Fund currently is short of this capability,and even though it has been trying hard to improveits skill set by recruiting new and experienced peo-ple, there is a general shortage of expertise in thisarea that cannot be remedied in the near term.33.Providing technical assistance on financial is-sues and on bank supervision is considered vital ifthe new code of good practices is to be implementedeffectively. Institutions such as the BIS, the WorldBank, the OECD, and the Fund have supplied suchassistance in the recent past, together with the helpprovided directly by individual central banks. How-ever, there was a general feeling that this activityshould not form part of the core activities of theFund, since these needs could be better satisfied byother institutions. There was an underlying concernthat the resources available in this area would bespread too thinly as different international institu-tions competed for them.

Interrelations, External Shocks, Spillover,and Contagion

34.A consistent theme in our discussions was thatthe Fund had failed to emphasize in its surveillanceactivities what many thought should be its maincomparative advantages: its knowledge of the in-ternational macroeconomic environment; and itsknowledge of the experiences of other countries indealing with similar policy problems. It was veryunusual for missions to ask questions such as:“Have you thought about the impact a measure im-plemented in another country (for example, a de-valuation or an increase in interest rates) will haveupon your country?” or “Have you evaluated theimpact that the measures you are planning to im-plement will have upon other countries?” Simi-larly, it was unusual for an Article IV mission—un-less specifically requested—to provide an analysisof how other countries in roughly similar situationshad dealt with a specific policy problem, and of thepros and cons of different approaches. But, wewere told, when such analyses were prepared, theywere found to be helpful and of high quality. 35.Some observers considered that bilateral surveil-lance needed to be looked at more broadly and morepreemptively in the light of changed global circum-stances. There was a specific suggestion that the

Fund be given a mandate to assess a country’s vul-nerabilities to specific shocks, and that it should dis-cuss with the authorities what, if any, contingencyplans the authorities had if the worst did indeedoccur. While country officials saw the conceptualmerit of such an approach, it was not clear how will-ing they were themselves to engage in such open-ended discussions. 36.Nonetheless, many of those interviewed empha-sized that one of the most useful aspects of surveil-lance is the information that they got about othercountries through the consultation material that theFund distributes to its members. In this regard, theconsistent approach and format of the Fund was seenas a plus.37.While many thought that a greater role for re-gional surveillance might contribute to improvingthe effectiveness of surveillance, as noted in the pre-vious chapter neither in Asia nor in Latin Americadid we find anyone who thought that recent Fundparticipation in (or organization of) meetings inthose regions had constituted any kind of break-through. It was also noted that contagion is not just aregional problem. As a senior central banker empha-sized, “Russia affected Brazil, which in turn affectedHong Kong.”

Early Warning Indicators

38.There is, naturally enough, basic agreement thatthe development of macroeconomic and financial in-dicators, as part of an early warning system, wouldbe useful for surveillance. It was considered thatsuch indicators would be useful not only in currentsurveillance, but even more so if the proposed Con-tingent Credit Line were to be introduced. Our con-versations with Fund staff, however, indicated thatthey were yet to be convinced that such indicatorswere sufficiently reliable to bear much policyweight. The Board has recently reviewed work on anearly warning system (EWS) and found that “itcould constitute an additional tool . . . [of surveil-lance]” and that it might, in particular “usefully sup-plement the current discussions on WEMD.” TheBoard cautioned, however, against any use of EWSin publicly available documents because of the mar-ket-sensitive information contained in the analysisand, particularly, the risk of generating self-fulfillingexpectations of crises. These issues are discussedfurther in Box 3.1.

Standards

39. One important strand in the current debate on theinternational financial architecture has been whetherand how to apply internationally agreed standards invarious areas. The most important of these so far have

40


been the Basle capital adequacy standards, but therehas been discussion of standards in quite a few otherareas, such as accounting systems, bankruptcy codes,corporate governance, as well as the code of goodpractices in the financial and monetary area that wasreferred to earlier. It was put to us that this general ap-proach constituted a potentially major expansion ofsurveillance. Thus, the current system is based on giv-ing advice on each nation’s particular macroeconomicpolicies and, increasingly, on other policies as well,with the hope that consistent policy advice acrosscountries would lead to a generally better outcome.The “new philosophy” is based more upon assuringthe observance of internationally agreed standards ina range of areas, with the Fund playing a—perhapsthe—leading role in monitoring them. It would marka shift toward greater emphasis on rules and sound in-stitutional designs to accompany the more traditionalmacroeconomic surveillance.40.While this development has yet to have much ofan impact on surveillance, we did ask a number ofour interlocutors for their views. Some argued that itwas impossible to apply universal standards to coun-tries without taking into consideration their culture,history, and structural characteristics. Others—in-cluding many staff members—were not opposed inprinciple, but pointed out that the monitoring ofstandards in a wide range of specialized areas calledfor a very different set of talents from that neededfor macroeconomic policy analysis, and that it wasfar from clear that the Fund was well equipped atpresent for this type of surveillance, or when itwould be. We return to this challenge in Chapter V.

Scope and Coverage

41.There was a consensus that Article IV consulta-tions should continue to focus on asking basic hardquestions about key macroeconomic policies—mon-etary and fiscal policies and their implications forthe balance of payments and the exchange rate. Andas the discussion above shows, it was felt that theFund does this quite well. At the same time, theFund has gradually incorporated new issues to beexamined in surveillance. In addition to the tradi-tional macroeconomic demand-side topics, andabove and beyond its more recent addition of finan-cial sector and capital account issues, it has gradu-ally become involved in more microeconomic andsupply-side matters such as trade liberalization,labor markets, offshore banking supervision, tax re-forms, expenditure streamlining, income distribu-tion, poverty, land reform, environment, and soforth.42. It was generally acknowledged that the expan-sion of Fund surveillance into financial sector andcapital account issues was inevitable. These issues

were seen as increasingly intertwined with theFund’s traditional analysis of macroeconomic andexternal sector issues. However, most observersdrew a sharp distinction between such matters andthe varied structural or microeconomic issues intowhich the Fund has recently been expanding. Therewas a fairly strong consensus—extending over gov-ernment officials, present and former Fund staff, andacademics—that this expansion detracted from theeffectiveness of surveillance, for three reasons.

• It diluted the focus on basic macroeconomicissues, to the detriment of those issues, al-though they were still central to the Fund’smandate.

• Staff did not really have much to contributeon many of the issues that they were now ex-pected to discuss, because they had neitherthe training nor the experience. Views variedabout the quality of the advice provided inthese new areas, but in general its quality wasperceived to compare unfavorably with thatoffered in the more traditional macroeco-nomic areas. It also did not measure up wellagainst the advice given by other interna-tional institutions that specialized more onthe microeconomic side, such as the WorldBank or—for most industrial countries—theOECD. Even apart from the question of qual-ity, there was a general feeling that the Fundshould rely more on the microeconomic workdone by institutions with more specializedexpertise.

• Some interviewees simply thought that as amatter of propriety or national sovereignty, theFund was getting into issues—for example,military outlays, income distribution—thatwere “none of its business,” taking into ac-count its stated purposes and legal standing.47

43.Many interviewees felt that surveillance, insofaras it was useful, was useful because it was relativelynarrow and focused. They felt that it should be kepta “limited-purpose vehicle,” with the basic functionof “bouncing off ideas” in a frank exchange, and thatthe Fund should function as a “clearinghouse” ofideas and experiences. As put by a highly experi-enced former minister of finance, “the essence ofsurveillance is not to try to verify the quality of gov-ernment, but to have positive exchanges of viewswith the authorities.”

41

47At the same time, it should be recognized that the reasons forFund commentary in these areas can be quite subtle on occasion.We heard, for example, that commentary on military spending inone country was more or less invited as a way of strengtheningthe hand of government in its desire to curb military outlays.

III SUBSTANCE

44.One partial exception to the above range ofviews came from our meetings with NGO and tradeunion representatives, who considered it essentialthat the Fund understand fully the likely social andenvironmental ramifications of its advice, especiallyin the area of fiscal policy. For example, the NGOofficials we spoke to were particularly concernedabout the short time horizons that prevailed in thedecision-making processes of governments, suggest-ing that good economics often called for a longer-term view. For example, they noted, environmentalsustainability needed to be judged over a particularlyextended horizon. They thought that such considera-tions were a very appropriate part of surveillance,given that under Article I of the Articles of Agree-ment the Fund is committed in their view to “theproper use of natural resources.”48

45. However, these representatives too recognizedthat the Fund had little expertise in such areas, andthey preferred to see it make more use of the workdone by other institutions such as the InternationalLabor Organization (ILO), the OECD, and the WorldBank, rather than producing work itself that was un-likely to be as thorough. That did not mean the Fundshould not have ideas of its own—and indeed theFund should be asking “awkward questions” in theseareas—but that it should be more prepared to incorpo-rate the analysis of others in its work.

Small States

46. In general, the views of small states were notradically different from those of others. Their repre-sentatives generally thought that surveillance hadbeen reasonably good and that policy debates weregenerally useful. However, they emphasized particu-larly that the Fund did tend to take a “one-size-fits-all” approach to policy. Among the particular con-cerns raised were the following.

42

Much academic interest has recently focused on theproduction of early warning systems (EWS) that mightgive early indications of looming balance-of-paymentstrouble based on selected economic indicators.1

In the current literature on early warning systems, twodifferent approaches have been taken. The first is basedon standard econometric estimations of qualitative re-sponse models (e.g., probit models) explaining the proba-bility of the binary occurrence of a crisis with a set of ex-planatory variables. The second is a “signals approach,”in which those indicators that typically show exceptionalbehavior preceding a crisis are singled out. Optimalthresholds for these variables are then estimated, so thatindicators issue crisis signals when they surpass thethreshold, and these signals are then combined into acomposite crisis index. Finally, Berg and Pattillo (1998)suggest a combination of the two approaches by embed-ding the step-function of the signals approach in a probitmodel using several variables as regressors.

Variables that have typically been included in theEWS investigations include the real exchange rate,credit growth, and the ratio of M2 to reserves. Somestudies have also included export growth, the govern-

ment budget deficit, and the ratio of the stock of foreigndirect investment to total external debt.

The essential test for the EWS, of course, is theirability to predict out-of-sample crises. That is, given theinformation available prior to a particular crisis, wouldan early warning system estimated prior to the crisis infact have predicted it? In that connection, Berg and Pat-tillo (1998) test the out-of-sample properties of some ofthe recently proposed models during the 1997 currencycrisis by excluding the recent observations from thecomplete data sample and then calculating the predictedcrisis probabilities based on the in-sample estimations.

Here, the signals approach suggested by Kaminsky,Lizondo, and Reinhart (1998) seems to produce betterout-of-sample predictions than the proposed standardprobit models, but the combination of the two models aswell as revised probit models, both suggested by Bergand Pattillo, also seem successful. The general perfor-mance of these models, however, still leaves much to bedesired—even though some crises are indeed correctlypredicted, the number of false alarm signals is high andmost often outnumbers the true warnings. This currentstate, of course, stresses the fundamental dilemma be-tween overlooking actual crises (“Type I errors”) andwrongly predicting crises (“Type II errors”). The mixedresults were also confirmed in EWS investigations car-ried out for this report based on the probit approach.The out-of-sample properties of the model were testedusing periods of crisis for 7 of the 12 countries chosenas case studies.

One reason for the relatively limited success of theEWS is that the implicit assumption in these estima-tions of identical behavioral relationships across coun-

1A comprehensive review of the EWS literature, which hasbeen highly influenced by research done at the IMF, is providedin Graciela Kaminsky, Saul Lizondo, and Carmen Reinhart,“Leading Indicators of Currency Crises,” Staff Papers, Interna-tional Monetary Fund, Vol. 45 (March 1998), pp. 1–48. An as-sessment of the capabilities of early warning systems to predictout-of-sample observations is contained in Andrew Berg andCatherine Pattillo, “Are Currency Crises Predictable? A Test,”Working Paper 98/154 (Washington: IMF, November 1998).

Box 3.1. Early Warning Systems

48The relevant part of Article I is the following: “to facilitate . . .the development of the productive resources of all members as pri-mary objectives of economic policy.” See Appendix II.

• There was some concern that too much timewas devoted to data issues rather than policyadvice. (On the other hand, some staff mem-bers felt that this data work was, simply put, anessential prerequisite for any sensible discus-sions of policy issues.)

• Often it was necessary to educate the staff inthe particular needs of small states—such is-sues as why social services tend to be expen-sive in archipelagos, or the difficulties in intro-ducing a VAT in countries that do not have therequisite administrative capacities.

• Some representatives of small states thoughtthat the Fund was excessively enthusiasticabout exchange rate depreciation, failing inparticular to appreciate sufficiently all the im-plications of a devaluation in a very small openeconomy.

• On fiscal issues, representatives expressed theview that general statements like “strong fiscalaction is needed” were not very useful, since

they required specific suggestions on tax is-sues and where cuts could be made. (Suchstates can of course receive technical assis-tance in the fiscal area. How well this assis-tance links up with the regular consultationswe are not in a position to judge.)


47. The Fund’s published work on multilateral sur-veillance is widely recognized as being of high qual-ity. We learned of many instances where the WEOwas a basic source document and building block forofficials engaged in monitoring and forecasting in-ternational developments. The ICMR was alsohighly rated, though clearly less widely known andused. While it appealed to a narrower audience thanthe WEO, it was particularly appreciated by thoseinterested and/or involved in assessing internationalfinancial developments as bringing more analyticalsubstance to the review of issues than is typically


43

tries and across time periods does not always hold. Ifthis is not the case, the resulting estimations and pre-dicted crisis probabilities may of course be misleading.For example, macroeconomic external and internal im-balances may previously have been more important,whereas variables capturing a country’s vulnerabilitytoward sudden capital outflows have become more im-portant in recent years.

Another problem with the EWS is that potentially im-portant variables are missing, for example, variablescapturing financial sector weaknesses and more politi-cal and institutional variables such as poor corporategovernance and the general level of administrativecompetence. In an attempt to address the potential im-portance of financial sector distress, a variable captur-ing the occurrence of banking crises was included in theEWS analysis carried out for this report. This inclusiondid indeed correctly increase the predicted crisis proba-bility in some cases (most noteworthy, Chile in 1982and Sweden in 1992/93), but at the cost of increasing itwrongly in other cases. More analytical work on thisissue, including attempts to improve the quality andavailability of data, would clearly be desirable in futureFund work on EWS.

A possible explanation for the quite low ability topredict the timing of a crisis may also be the presence ofself-fulfilling attacks as a consequence of the existenceof multiple equilibria for certain ranges of fundamen-tals. It is clear that if this is the case, EWS can only givean indication of a country’s potential vulnerability to anattack. Related to this issue, the likely presence of inter-national contagion has so far largely been ignored in theEWS literature, which has primarily focused on coun-

try-specific indicators. It would be useful to focus moreon this issue in future work.

EWS would obviously be helpful in surveillance ifthey could improve crisis projections as compared withthe informed predictions by staff members. However,even if the EWS were neither systematically better norworse than normal staff predictions, their objective na-ture could still make them a valuable contribution tosurveillance. Our recommendations on this score arediscussed in the text of Chapter V.

Some obvious objections to the likely success of im-plementing EWS, however, seem worth mentioning.First, the existence of a successful early warning systemis almost a contradiction in terms, since corrective policychanges following an early warning signal might preventthe crisis and thus prove the original crisis predictionwrong. On the other hand, the publication of results froman early warning system might prove to be self-fulfillingif international investors were to use them as focal pointsof speculative attacks. Finally, if it is assumed that finan-cial markets are efficient, a natural question seems to bewhy the private sector has not already acquired theknowledge assumed present in the EWS.

There is understandable concern, not least in theBoard, that the current high number of both Type I andType II errors could damage both countries’ and theFund’s credibility if they were to be published. How-ever, it is clear that the private sector will—and alreadydoes—provide this type of analysis. We see no reasonwhy the Fund should not present the data it uses inter-nally to evaluate vulnerabilities in WEO and ICMR, inthe form of tables and graphs, without necessarily offer-ing written analysis or conclusions.

III SUBSTANCE

found in other coverage available. The same seemstrue for WEO as well, although it has more directcompetition from other public and private publica-tions than does the ICMR.48.At the same time, there is no doubt that themajor errors in forecasts that marked, in particular,the Fund’s assessment of output trends in 1998 havediminished the reputation of the multilateral publica-tions. As shown in recent Fund documents evaluat-ing the Fund’s programs in the Asian countries moststrongly affected by the financial crisis, the forecastswent from a sharp slowdown in the growth rate ofoutput in five Asian countries to about 2 percent(forecast in late 1997) to more recent estimates, sug-gesting a decline in their aggregate 1998 output of 8percent. The prospects for recovery in Japan in 1997and 1998 were also evaluated far too optimistically.The magnitude of the reversal of short-term capitalflows, obviously the most difficult element to assessin the macroeconomic outlook, was severely under-estimated in a number of cases. Admittedly, these er-rors were only slightly larger than those of the so-called consensus. Nevertheless, it was put to us thatthe Fund, with its close contact with policymakers,privileged access to information, and substantial re-sources, must expect to live up to a higher standardthan other forecasters.49.We are aware that it may be unfair to attributeblame for the somewhat unsatisfactory state of theforecasting record of WEO—already documented inArtis (1996)49—only to the Research Department(RES), which is ostensibly responsible for its publi-cation.50 Global forecasting in the Fund is built upfrom the country analysis supplied by the area de-partments—that is to say, it is more “bottom up”than “top down.” We were told of major disagree-ments in a number of cases in recent years, as wellas earlier in the 1990s (when recessions in Japan andGermany were seriously underestimated) betweenarea departments and the Research Department,which management seems to have resolved largelyin favor of the former—with the consequence ofworsening the forecast record further.51

50.Moving beyond forecasts, we also examined theFund’s multilateral surveillance, both published andunpublished, in the run-up to the Asian crisis. We

found that the Fund—in both bilateral and multilat-eral surveillance—largely failed to identify the vul-nerabilities of the countries that subsequently foundthemselves at the center of the Asian financial crisis,except in the case of Thailand. In particular, it faileduntil rather late in the day to address a number ofsystemic issues. Moreover, to the extent that surveil-lance did identify these vulnerabilities, the tone ofpublished Fund documents—notably WEO—wasexcessively bland prior to the December 1997 up-date of WEO/ICMR, after the crisis had erupted.51.WEO eventually did express concern over finan-cial vulnerabilities in emerging market countries.But this was done in language that was not suffi -ciently clear to raise the attention level in the coun-tries concerned or in the Executive Board. Morepointed analysis was contained in unpublished mate-rial, especially on Korea, where the analysis of theResearch Department was more hard-hitting thanthat of the area department. But this analysis, whichwas unpublished (and indeed not communicated tothe Board) did not lead to a sharpening of countrysurveillance. We offer a more detailed assessment inBox 3.252.Mention must also be made here of the conceptof equilibrium exchange rates. The Fund is chargedabove all with “exercising firm surveillance over theexchange-rate policies of members.”52 This puts apremium on using the best possible methods for as-sessing departures from sustainable exchange rates.Within the Fund, work in this area takes place pri-marily through the Coordinating Group on Ex-change Rates (CGER), a working group establishedin 1995 by PDR and RES. Recent work by theCGER has been presented in an Occasional Paper53

and some use is made of the analysis in bilateral sur-veillance of the major industrial countries, in WEO,and in the Economic Counsellor’s presentations tothe WEMD sessions (for a discussion of these ses-sions, see below).54

53. It is notable that while this work is both centralto the Fund’s mandate, and of high analytical qual-ity, relatively little is published. Some observerssuggested that regular publication of Fund estimatesof equilibrium exchange rates would be useful toboth policymakers and market participants. 54.Although much less known than WEO and the ICMR, the WEMD sessions also constitute an

44

49Michael T. Artis, “How Accurate Are the IMF’s Short-TermForecasts? Another Examination of the World Economic Out-look,” IMF Working Paper 96/89 (Washington: IMF, 1996).

50It may be worth bearing in mind also that WEO and theICMR are not, formally, documents of the Board, but rather of thestaff. The Board reviews them but does not release them in itsname; the May 1999 WEO includes for the first time a summingup of the two Board discussions of a draft.

51Although, equally, it was also pointed out to us that on anumber of occasions the judgment of area departments had beenmore accurate than that of the Research Department.

52Article IV, section 3, reproduced in Appendix II.53Peter Isard and Hamid Faruqee, eds., Exchange Rate Assess-

ment, IMF Occasional Paper No. 167 (IMF: Washington, 1998).54A different method for the modeling of real exchange rates has

also recently been developed by Fund economists, with promisingresults for some developing countries. Ronald MacDonald, “WhatDetermines Real Exchange Rates? The Long and the Short of It,”IMF Working Paper 97/21 (IMF: Washington, 1997).


45

Box 3.2. Surveillance, Capital Flows, and Financial Crises

We looked at how Fund multilateral surveillance, es-pecially as expressed through the Fund’s two mainpublications, the World Economic Outlook(WEO) andthe International Capital Markets report (ICMR), dealtwith the issue of capital flows and financial crises since1993; and at bilateral surveillance of Korea in the sameperiod. Three main impressions emerge: the first twosuggest an underestimation of the risks involved, whilethe third makes it difficult to avoid the conclusion thatthe risk of contagion was barely perceived prior to thecrisis.

First, there was generally a positive evaluation of thesustainability of sizable inflows to emerging markets.Net capital inflows were projected to remain at highlevels or even to increase. This evaluation was linked tothe general euphoria surrounding growth prospects inemerging markets. International capital markets wereperceived as well-functioning with the risk of reversalslargely linked to the occurrence of excessively expan-sionary policies resulting in domestic overheating andexternal imbalances. Despite the Mexican experience,an optimistic perspective on the beneficial effects ofcapital inflows soon reemerged.

Second, when doubts about the sustainability of capi-tal flows were voiced—more frequently in the ICMRthan in the WEO—they did not put much emphasis onweakening financial systems in the capital-importingcountries during the process of simultaneous liberaliza-tion of the domestic financial sector and of capital flows(both with the strong encouragement of the Fund).1 As acorollary of this relative lack of priority to financialfragility, the interrelationships of the latter and balanceof payments crises did not figure prominently in policyadvice, particularly in the early period. This is surpris-ing, since economists at the Fund have contributed sig-nificantly to the development of analytical insight intothe area of the “twin crises.”

Third, the importance of potential regional and in-ternational contagion of currency crises was givenvery little attention prior to the crisis, even in internalanalyses. For example, after a mission from the Capi-tal Markets group visited Asia in April 1997, an inci-sive memo to management warned accurately of thedangers of crises in Thailand and Korea, resulting, re-spectively, from an attack on the currency and fromdeepening problems in the banking sector. However,the same memo explicitly dismissed the risk of re-gional contagion. Indeed, even in late August 1997,well after the floating of the baht, a(broadlyupbeat(memorandum to the Board on the risks of con-tagion in Asia did not even mention Korea.

The overall impression of Fund multilateral surveil-lance as expressed through the WEO and the ICMR isthat while these documents did make a number of perti-nent observations on capital flows and financial crisesthat are helpful in understanding subsequent develop-ments in Asia and elsewhere, the risks were not force-fully presented. Nor did the documents draw the appro-priate conclusions and present them in pointed, country-specific policy advice.2

Inadequate as the attention to risks appears to havebeen in multilateral surveillance, the evaluation of cri-sis signals was—with the benefit of hindsight—furtherdivorced from reality in some cases of bilateral sur-veillance. We have looked at Korea as a prime casestudy.

In bilateral surveillance of Korea, the potential dan-gers of a rapid process of capital account and domesticfinancial liberalization were not properly assessed. Inparticular, little attention was paid to the vulnerabilityof Korean banks to a potential weakening of the won.Furthermore, it was only in early 1997 that criticalviews were first expressed as to the credibility of offi -cial estimates of nonperforming loans. More generally,overoptimistic forecasts of output growth in Korea—even in September 1997, the staff ’s judgment was thatthe authorities would weather the gathering storm andthat a worst-case scenario was a deceleration of outputgrowth to 4 percent for the year—delayed a realisticevaluation of financial weakness.

As regards exchange rate policies, Fund recommen-dations advocated a more flexible regime, clearly onthe assumption that the currency would appreciateunder the pressure of continuing sizable net inflows.This analysis was based strictly on traditional macro-economic fundamentals without reference to micro-economic weaknesses in the financial and corporatesectors—even though such weaknesses had been iden-tified by the capital markets mission referred to above.

Accordingly, almost up to the outbreak of crisis inNovember 1997, a low probability was attached to theoccurrence of an external crisis. The real effective ex-change rate measures did not show signs of over-valuation, the current account deficit had narrowed,available external debt indicators were not alarming,and remaining capital controls prevented foreignersfrom shorting the won. The recognition that domesticevents, notably a banking crisis, could trigger pressureson the exchange rate, was not totally absent, but it wascertainly not given much weight.

1While the May 1997 WEO—by which time the Thai bahtwas already under severe exchange market pressure—did identify banking system fragility as a general concern,the discussion of Asian countries was generally quite sanguine.

2In our interviews, staff suggested that such warnings werein fact present and indeed couched in language that was quitestrong, in Fund terms. However, one well-informed and dis-interested observer was of the view that the basic draftingstrategy was to say as little about risks as possible, while atthe same time still being able to claim, if the risks did becomereality, that they had been addressed.

III SUBSTANCE

important element within the process of multilateralsurveillance. In these sessions, Executive Boardmembers, selected staff members, and managementengage in relatively open and informal discussionsof issues. These can range from the most recent de-velopments in the international monetary system toan assessment of vulnerabilities in different coun-tries around the world. Many participants have ratedthese informal meetings among the most interestingand important of Board meetings, and those eligibleto attend are keen to do so.55.Unfortunately, since the main goal of these ses-sions is to generate open and free discussions amongparticipants, neither the background documents northe thrust of the discussions are known much beyondthe meeting room—except to the extent that they arerelayed back to the home countries. We assume thatthese discussions do exert some influence upon whathappens on other occasions in the institution, at leastby ensuring that Executive Directors are well in-formed about the latest relevant developments.56.The Fund is also charged with preparing back-ground material for the Ministerial and Deputies’G-7 meetings. These notes are not openly distributedbut they do receive attention at a high level prior toG-7 meetings. The Fund devotes considerable atten-tion to this material, and some participants considerit one of the more important Fund instruments ofmultilateral surveillance. However, the materialserves as useful background to the G-7 meetingsrather than as a policy agenda.

Other Sources of Surveillance

57.There are a number of other organizations whosefunctions overlap with the Fund’s surveillance role,although not all of them describe their functions assurveillance (in addition to regional surveillance in-stitutions, which were discussed in an earlier chap-ter). We should first note that with just a few excep-tions, there was a consensus among our interlocutorsthat, in the macroeconomic sphere, Fund surveil-lance taken as a whole, both published and unpub-lished (essentially Article IV consultation reports),compared favorably in terms of quality, focus, andcoverage. 58.The “country examinations” of the OECD, rep-resenting presently 29 countries (overwhelminglyadvanced economies), is perhaps the closest ana-logue to the Article IV process, and this was the or-ganization to which the Fund was most often com-pared by interviewees. The OECD staff produces“country reports” (on a roughly biennial cycle) oneach of its members, which are then discussed bycountry representatives. The key differences fromthe Article IV process are the following.

• The reports focus more on structural issuesand less on macroeconomic conditions. As aconsequence, they are rather longer and moredetailed.

• The examination takes the form of an discus-sion, in which high officials from one or twoother countries, on the basis of the country re-port, examine (candidly) the country’s poli-cies; the country representatives then present adefense.

• The country report is published, with the con-sent and after negotiation in a drafting sessionwith the authorities of the country concerned.However, the record of the discussion is notpublished.

59.The OECD also has a multilateral surveillancerole: it produces a semiannual Economic Outlook,which covers some similar ground to the WEO, andit hosts Working Party 3, at which senior officialsfrom member countries discuss current global eco-nomic developments in quarterly sessions. Given thelevel of representation, this forum, in the view of ex-perienced officials, tends to generate more interest-ing and lively exchanges than are possible in the Ex-ecutive Board. No report on those discussions isissued.60. In some of the larger countries, the OECD(which, as just noted, publishes its regular countryreports and also issues a press release alongside) hasbeen generally more in the public eye, while the IMFhas had a lower profile. Most of those who spoke tothese matters thought that both the Fund and theOECD did a good job in their respective surveillanceprocesses, and that these processes were at least inpart complementary. While most considered that theFund’s output was more useful overall, as noted ear-lier there was general agreement that the OECD dida better job in analyzing structural issues.61.The World Bank, in countries where it is a majorlender, occasionally produces a “Country EconomicMemorandum,” which gives a comprehensiveoverview—both macroeconomic and structural—ofthe country’s economic situation. The World Bankalso produces “Country Assistance Strategies” forborrowing countries. While these are not specificallydesigned as instruments of macroeconomic surveil-lance, lending operations clearly need to be seen inthe context of the general macroeconomic situation.This is particularly the case for “structural adjust-ment” lending, which has become an increasinglyimportant part of the Bank’s operations in recentyears. However, the general feeling was that it didnot really engage in surveillance as such, at least notin the way practiced by the Fund, or by the OECDfor that matter. Its focus appeared to our interlocu-tors to be much more on sectoral issues.

46

Other Sources of Surveillance

62. The BIS has a largely informal, but increasinglyimportant, role in financial sector surveillance. Aswell as keeping an eye on industrial country situa-tions, it monitors emerging markets and coordinates anetwork in developed countries of central bank econ-omists who share information about developing coun-tries. This process provides input to meetings of cen-tral bank governors hosted by the BIS—in particularwhen BIS management feels that a country presents apotential problem. The BIS will also provide the sec-retariat for the new Financial Stability Forum, whichwill effectively formalize these arrangements and alsoinvolve nonbank regulators (securities markets, ac-counting standards), as well as other international fi-nancial institutions, including the Fund. It also pro-vides technical assistance, and is expanding itsactivities through a new Financial Stability Instituteand a Hong Kong office. 63. In the private sector, rating agencies (and to alesser extent investment banks) perform a surveil-lance function for the benefit of private investors.The methodology used by rating agencies is not dis-similar to that of the Fund, involving analysis ofmuch the same data and a country visit to talk to theauthorities. As a result, reports produced by ratingagencies often cover similar ground to that of an Ar-ticle IV consultation, albeit usually in less depth.The Institute of International Finance (IIF), an asso-ciation of financial institutions, produces “CountryReports” that are closer both in form and function toArticle IV staff reports. 64.The main difference is that private sector insti-tutions generally take a much narrower view,

guided by the changing thrust of market interest.Rating agencies, for instance, are concerned aboveall with the issue of whether a country would beable to service its debts, not with the general qual-ity of its overall macroeconomic policies. Thispoint was also emphasized to us by the rating agen-cies themselves, who made it very clear that theydid not regard themselves in any sense as substi-tutes for the Fund. 65.However, it is also apparent that private sectorinstitutions do a more explicit job of the assess-ment of risks, including political risks. While it isquite rare for the Fund explicitly to examine vul-nerabilities, or to look at adverse-case scenarios,private agencies regard this as fundamental. Morethan one observer expressed the belief that the pri-vate sector was able to produce more up-to-dateand precise information than the Fund for countriesthat have the market’s attention, with the added ad-vantage of not having to be as polite to govern-ments. At the same time, it was also noted by somethat information put out by private financial insti-tutions was in principle less trustworthy, inasmuchas it could reflect the financial interest of the insti-tution. This would not, of course, be true of the IIF,which does surveillance work for private financialinstitutions in general. Its surveys were thought tobe more narrowly focused than those of the Fund,and in any event, they do not have the same coun-try coverage. Those who saw both the IIF countrystudies and the Fund staff reports considered thatthe IIF material was far from being a substitute forthat of the Fund.

47

1. This chapter deals with the “bottom line,” namely,did Fund advice make a difference? The questionsposed by the Board that are most directly relevanthere are:55

(iv) Did surveillance have different impacts indifferent groups of countries?

(viii) Was the Fund’s advice implemented? If not,why?

(vi) How successful have been the specific ef-forts made since early 1995 to strengthensurveillance? What effect have these effortshad in the context of the difficulties emerg-ing in some Asian countries in 1997?

2. Other questions also have an “impact” aspect—inparticular, those that focus on communication andpublicity, namely:

(vi) . . . How has the provision of information bythe authorities affected surveillance?

(ix) . . . How useful were the methods to makesurveillance available to the public? Shouldthe Fund go further in publishing country reports?

(x) How did governments disseminate surveil-lance conclusions within and among govern-ment institutions? Was the circle of partici-pants limited to economic agencies?

3. First we review the extent to which the impactdoes in fact vary by country groupings (question iv).We then move on to a discussion of the general con-siderations that stand out for the broad range of“middle” countries. This is followed by more de-tailed discussion of four instances in which financialand economic vulnerabilities became particularlyapparent. There we focus on the advice given andthe impact that the Fund had as developments un-folded. We then describe the results of the Fund’sown internal efforts to improve surveillance since

1995. Finally, since they are so closely bound upwith impact questions, we discuss matters of com-munication and transparency. This latter discussioncovers the experience with Public Information No-tices (PINs) since their start in 1997, as well as thebroader questions related to dissemination of Fundsurveillance material in general, and in particular theArticle IV consultation reports. In the final part ofthis chapter, we discuss practices and views relatedto another aspect of communication of Fundviews—that within government itself.

Impact by Country Groupings

4. Our review of country impact did not turn upsurprising or anomalous results in terms of differ-ences. Most members fell “in the middle.” That isto say, for most members the impact of surveillanceappeared to be sensitive not so much to the type ofcountry but rather to more general factors.5. However, size and sophistication always matter.And as was to be expected in the sample of coun-tries on which we focused our interviews, Japanand the United States stood out in terms of the ex-tent to which surveillance was seen as a low-impactexercise. Our nonsample discussions on countryexperiences also confirmed this. That being said,we were assured that in the largest industrial coun-tries surveillance through Fund missions, etc., wastaken seriously. Although a great deal was not ex-pected from the consultations, U.S. officials in par-ticular considered it especially important to “set agood example” in the way they entered into them—something also attested by the staff. However, forthe large industrial countries, in no case would it beright to claim that the Fund had more than a mar-ginal or occasional impact on national policy deci-sion making, even in a case such as that of Japan,where economic and financial difficulties havebeen particularly pronounced. At the same time,some Japanese observers also made the interestingpoint that their bilateral consultations did “fill agap.” The Fund approach provided a blend of theo-retical and practical analysis of issues that was dif-

IV Impact

48

55Numbers in parentheses indicate the number assigned to thequestion in the original Executive Board request. See the Termsof Reference in Appendix I.

Other Elements Shaping Impact

ficult to replicate from Japanese sources such asacademics (very theoretical) and private thinktanks (focusing on business-cycle analysis). Thatbeing said, the general view in the public and pri-vate sectors in both Japan and the United Stateswas that the issues on which the Fund was capableof pronouncing would receive such wide and in-tense discussion within the country that it would besurprising if the staff were able to add very much towhat was already on the table, beyond an interna-tional voice—however hard they tried. At the sametime, our official informants were virtually unani-mous that it was always interesting to hear what thestaff had to say and that it was worthwhile in termsof their own thinking to have to respond to thestaff ’s informed probing.6. Another aspect was that even in relation to inter-national surveillance, let alone compared with thecrucial and ever-present forces of domesticdebate, theFund was far from paramount. Some bilateral rela-tionships were extremely important, and probablymore so than the peer group influences, or even pres-sures, coming out of the G-7 process for example.Furthermore, the impact of the G-7 process was alsogreater than that of Fund surveillance. In this latter re-gard, however, mention was made of the usefulness ofthe Fund’s input into the meetings of G-7 financeministers and governors as providing background fordiscussion of macroeconomic issues, and even for“clearing some issues off the table.” In Europe, the in-fluence of the convergence process toward EMU hadapparently dominated other external surveillance in-fluences for all relevant Fund members.7. The other countries in our sample were econom-ically varied. However, the Fund’s impact, althoughon balance apparently greater than in the very largecountries, did not seem to vary in any very system-atic way. The general tendency was to see Fund ad-vice as an input that could on occasion be signifi-cant, depending upon the stage of the domesticpolicy debate. 8. There could well be from time to time an impactfrom new ideas or ways of looking at an issuebrought by the Fund. One example is the advice thatthe Czech (then Czechoslovak) authorities receivedfrom the Fund in the early 1990s in setting up mone-tary and exchange rate policy. Another, already re-ferred to in Chapter III and which has applied to anumber of countries, is the Fund’s work in sortingout the proper analysis of central bank operatinglosses—“quasi-fiscal deficits.”9. More mundanely, it was quite common for theFund’s views to be absorbed not as a revelation, as anew way of looking at an issue, but rather as supportfor a particular approach to policy that was alreadybeing advocated internally. In cases like this, the im-pact tended to be gradual, resulting from the contin-

ued reiteration of the same basic message over alonger period, and the building, perhaps, of a policyconsensus. Furthermore, it was often the case thatsome policymakers were inclined to find the Fund’srecipes more attractive than other policymakerswithin the same country—depending on which cor-ner of the government policymaking apparatus theywere located. As suggested earlier, central bankstended to find the Fund advice particularly congenial,given the emphasis on financial stability and theavoidance of fiscal excess. More broadly than this, apoint made in South America was that as views ondesirable policy on that continent shifted from earlierdirigiste, autarkic, approaches to more market-based,open-economy principles—an intellectual shift towhich the Fund may have contributed—this helpedto promote the acceptability of Fund advice. Domes-tic policymakers and the staff had more commonground on which to base their dialogue. At the sametime, it was also noted that the level of economic pol-icy sophistication had probably risen more rapidlyover time in South American countries than it had inthe Fund. This “catching up” would, naturallyenough, tend to lessen the impact of Fund advice.Such a tendency, which in itself is by no means a badthing, was also apparent elsewhere, although not asclearly identifiable as in South America.10.Beyond the sample countries, and as already re-ferred to in the previous chapter, we held special dis-cussions with a group of small states. There, the sur-veillance impact, including a heavy dose of technicalassistance (something also apparent, if to a lesser de-gree, in middle-ranking countries), was obviouslylarge. Fund visits were a major event that involvedvirtually the entire economic policymaking appara-tus. However, questions remained as to whether sur-veillance was being delivered as well as it could be,particularly with the turnover of staff. This was par-ticularly apparent given the lengthier stretches be-tween consultations and the fact that such surveil-lance is probably viewed as less high profile and lesschallenging by Fund staff.

Other Elements Shaping Impact

11. The methods and substance of advice have beendealt with specifically in Chapters II and III. Here,the focus is on what can be added to that discussionfrom the viewpoint of impact.12.But before that, one general point. In quite a fewinstances it was volunteered, with emphasis, thatFund advice carried more weight when it was at-tached to a financial program. This itself is virtuallya truism. However, we note that the implicit corol-lary is that Fund advice (as opposed to that of otherinstitutions that do not lend) may carry some addi-

49

IV IMPACT

tional impact because it is an institution that mightbe called upon to extend financing. Furthermore, thefocus in our remit on advice free of financing andconditionality56 may well become less clear-cut stillwhen the Contingent Credit Line becomes opera-tional. Given our terms of reference, in our ex-changes we did not dwell directly on these some-what hypothetical considerations, and neither for themost part did our interviewees. However, we do re-view the implications of the Contingent Credit Linein Chapter V.13. Reverting to current Fund practice, we presentbelow our listing of the general features that appear toenhance or detract from the impact of Fund advice,other than any matters relating to communication andpublicity. These are dealt with separately below. 14.Since these main features will already have beenlargely dealt with in the previous chapters on con-duct and method and on substance, we aim to be rel-atively brief.

• In terms of general Fund approach, the mostsalient point was the value of conveying theability to see the issues from the country’sviewpoint. Put another way, quite often mis-sions were seen as coming with a preconceivedframework. However, the impact of advicewas clearly enhanced when missions wereable, without compromising the general princi-ples of financial prudence on which Fund ad-vice is based, to adapt the advice to the partic-ular situation. In this regard, members placedgreat stock on a mission’s ability not only togive good and/or creative policy advice butalso to come up with concrete suggestions asto how it might best be implemented, recog-nizing the particular challenges the countryfaced. This was particularly true of smallstates. Here, the Fund’s surveillance role over-laps with its technical assistance function. Thefunctional departments (in particular Monetaryand Exchange Af fairs and Fiscal Af fairs)clearly have particularly relevant roles to playin this regard.

• Obviously, the capacity to accomplish thiscomes in part with experience, as well as fromhaving specialized resources. In that regardsome concern was expressed that many Fundeconomists had limited experience outside theFund. The feedback we received was that theywere good macroeconomists, especially goodat theory, but often lacked the additional credi-bility from experience in practical policy-

making—at least from the country side—or inimplementation.

• Staff turnover from one mission to another, al-ready mentioned as a clear problem for smallstates, was seen more generally as somethingof a difficulty. Correspondingly, it seemed to anumber of our interviewees that the institu-tional memory was quite limited. This meantthat more country time was spent getting mem-bers of a mission up to speed than would bedesirable, and consequently that there wouldbe less of an impact on policy.

• In terms of attitude, certainly there were caseswhere difficulties had arisen because of whatwas seen as peremptoriness on the part of indi-vidual staff. However, these were, as notedearlier, very much the exception. On the otherhand, as discussed in Chapter III, concern wasalso raised, both within the Fund, particularlyfrom functional departments, and within coun-tries on occasion, that missions were quiteoften more accommodating than was desirableif the real policy issues and vulnerabilitieswere to be adequately addressed.57

• While the Fund had begun, especially followingthe Mexican experience, to look at the quality ofthe domestic financial sector in most of thecountries we examined, it did not appear thatthe analysis was particularly deep or sophisti-cated, or that it had yet had any significant im-pact. This may well reflect the Fund’s tradi-tional macroeconomic focus and the training ofits economists as well as the fact that improvingthe financial sector is often a long and difficulttask. At the same time, it is also evident thatcountries have tended to accept closer examina-tion of their financial systems at best cautiously,sometimes probably out of concern over whatan outside look might turn up.

• Questions relating to the microeconomic andstructural scope of the Fund’s advice havebeen dealt with extensively earlier. Here, how-ever, it may be noted that the very breadth ofthat advice tends to dilute the impact. In partthis can be because the Fund is almost invari-ably seen as having less competence in micro-economic and structural areas generally(where it tends to rank below the OECD andthe World Bank, for example) than in macro-economic areas. Further to this, the impact ofthe main macroeconomic policy messages canbe directly lessened by signals emanating from

50

56”Countries [that] received no, or at most sporadic, disburse-ment of Fund resources.”

57This would of course be one argument, but a second-best one,in favor of rotating Fund staff.

Impact in Four Countries

other areas where advice was given. In one in-stance, for example, while the fact that theFund was keen on curbing the deficit was welltaken, the fact that the mission stressed“poverty reduction” as well, reportedly pro-vided a confusing message at the political levelas to whether it was reallyadvocating fiscal re-straint. Of course, the two objectives are notcontradictory, but it is easy to see how the clar-ity of the message can be reduced.

• Besides the potentially dilutive or contradictoryeffect of advice given over a broad range of mat-ters, an important question mark arose over theextent to which country advice adequately re-flects the Fund’s comparative expertise. That isto say, it was clear from our interviews that mis-sions were not at all active in bringing to theconsultations an explicit view of the interna-tional (or regional) economic and financialbackground into which specific country policyadvice might usefully be situated, thereby im-proving its impact. Nor did they often bring in-ternational comparative experience directly tobear, although they would seek it out if asked.However, when such advice was provided, oftenby functional departments, it was generally re-garded as being of high quality (for example, in-formation provided to Sweden on other coun-tries’experience with inflation targets).

Impact in Four Countries

15. In one of the countries looked at in this section, afinancial statesman made the comment that “If theobjective of the Fund surveillance is to preventcrises, then it clearly failed in our case. The questionis why?” Warding off crises is not the only elementin surveillance, but given its emergence as a particu-larly important component, it merits special atten-tion. In this light, we look here at the impact of sur-veillance in Brazil, the Czech Republic, Korea, andThailand—which have all experienced a financialand economic crisis in their recent history. 16.Looking at the advice actually given in the pe-riod leading up to the crisis, three policy areas are ofparticular relevance: fiscal policy, exchange rate pol-icy, and capital account liberalization.

• For all these countries, the Fund was for sometime counseling fiscal consolidation to agreater or lesser degree. Even if, as some con-tend, it is generally too quick to advocate fiscaltightening, and even if such tightening was atbest only part of the right recipe (in particularin Korea and Thailand), it is hardly conceiv-able in these four cases that more fiscal action

along the lines of Fund recommendationswould have brought on a crisis. Rather, itwould have lessened the probabilities.

• On exchange rate policy, the advice was lessclear-cut and more problematic. In the case ofBrazil and the Czech Republic, the authorities’strong desire to maintain a pegged exchangerate did get mild Fund support over an ex-tended period, even as the financial pressureswere accumulating. In regard to Brazil, the ex-change rate (the “real plan”) was the confi-dence anchor, and the Fund was unwilling toargue strongly for a regime change, eventhough staff believed the real to be signifi-cantly overvalued.58 For the Czech Republic,an important domestic consideration in hold-ing the exchange rate, which the Fund wentalong with, was convergence with the EU.Some will argue that in these instances theFund acquiesced too readily to exchange ratepegging. Perhaps, but where it did expressconsistent concern was whether domestic fi-nancial policies, again in particular fiscal pol-icy and in the Czech case wage policy as well,were adequate to support the peg. For Koreaand Thailand, the advice was clearly tilted to-ward increasing exchange rate flexibility. In nocase (except, arguably, Brazil, where in the lat-ter part of the period the advice came togetherwith massive conditional financing) did theFund positively encourage a member to holdon to the peg when it became broadly apparentthat some kind of exchange rate change or ex-change system change could become neces-sary in response to the emerging crisis.

• On the issue of capital account liberalization,however, the Fund’s advice certainly did nothelp prevent the crisis. In particular, in Korea,it encouraged capital account liberalizationwith little attention to problems of appropriatesequencing implied by liberalizing short-termforeign borrowing before foreign direct invest-ment. While the Fund would have preferredliberalization of foreign direct investment first,the tone of its advice was that the approachtaken by the Koreans was better than noth-ing.59 In fact, in retrospect, it was probably

51

58Indeed, these concerns were expressed—albeit in veryguarded terms—in the published Recent Economic Develop-ments sections of the 1998 staff report (published as IMF StaffCountry Report 98/24, April 1998). See in particular Chapter VII,“The Post-Real-Plan Developments of Tradables and Nontrad-able Prices and the External Current Account.”

59For example, the 1996 Article IV staff report states: “The recentacceleration in the schedule for capital account liberalization and thesomewhat greater precision on prospective measures are welcome.”

IV IMPACT

worse than nothing, as the Fund has now rec-ognized.60 In Thailand, Fund advice sufferedfrom a similar defect, albeit to a lesser ex-tent.61 By contrast, advice to India, which didnot suffer a financial crisis, was more nuancedand gave considerably more weight to the se-quencing issue—as did the Indian authoritiesthemselves.62

17.However, what is more directly relevant for thepurposes of this discussion is that in all these casesthe authorities’determination of the policies actuallypursued was apparently not affected very much byFund advice.18.Why did the Fund have little impact? No singlefactor stands out, but a number of considerations, re-lating to both the Fund and the recipient of advice,can be noted.

• It is contended by some in Thailand that whilethere was forceful advice that trouble wascoming, including through visits from seniormanagement, it did not come early63 enough asregards the mounting difficulties in and via thebanking system, that is, by 1995 or early 1996.By the time forceful advice did come, “it wastoo late” (although policy mistakes continuedto be made until, and during, the onset of thecrisis). In other cases, the advice, early or not,was not nearly as strong.

• In two instances, Korea and Thailand, the Fundcollectively64 was not focusing on the problemsof financial structure—banks and bankdebtors—that proved to be at the heart of theeventual confidence crisis. Nor was the Fundable to obtain from the authorities all the statis-tical information—on international reserves andforeign debt—that would have enabled it tomake a more forceful case. This was especiallyimportant in the case of Korea, where the fiscaland current account deficits, the usual vulnera-bility signals, were both relatively modest.

• In some of these instances (something that alsocan probably be generalized across Fund mem-bership), the authorities were apparently notvery sure themselves what the actual bankingcum external indebtedness situation actuallywas. They also likely lacked the mechanismsto find out, to the extent that they in factwanted to. And they did not accede readily tothe Fund taking a look—even late in the day.

• In no case does the Fund appear to have comeclose to going public with its advice, althoughto do so would surely have increased its im-pact substantially. Indeed, in three of the fourcases, Brazil being the exception, communica-tion of Fund advice, even within government,was not good. However, few of those we spoketo—either in the countries concerned or in theFund—suggested that the Fund should havegone public (although in the case of Thailandinterviewees did allow that going public couldhardly have made things worse).

• In all these cases there was an exchange ratepeg, and that peg had political importance. Soeven if the Fund had advocated a change moreforcefully, it might not have had an effect.

• Finally and emphatically, the political situationwas either difficult or delicate throughout therelevant period in all four cases. And exchangerate policy, in particular, was a highly politicalissue. Political difficulties are of course likelyto lead to policy inertia, whatever the advicereceived, and however skillfully or forcefullyit is presented.

19.To summarize, what was common to these situa-tions, and what might therefore be reasonably

52

60See, for example, Stanley Fischer, First Deputy ManagingDirector, “Economic Crises and the Financial Sector,” speech tothe Conference on Deposit Insurance, Washington, September 10,1998: “Although country circumstances differ, the general adviceon international financial sector liberalization is first to open tolonger-term investment, particularly foreign direct investment,and only to open at the short end when the necessary precondi-tions, in the form of macroeconomic stability and a strong bank-ing and financial system, are in place. This was not the path cho-sen in Korea and Thailand.”

61The 1996 staff report argued that “particular progress [instructural reform] has been made in the financial sector; the mis-sion believes that consideration could now be given to relaxingrestrictions on foreign equity participation in the domestic capitalmarket.” While this might indeed have been helpful, the reportfailed to caution on the dangers associated with the then prevail-ing system, which tended to encourage short-term foreign bor-rowing.

62It is interesting to note in this regard that while the important“Report of the Committee on Capital Account Convertibility” is-sued by the Reserve Bank of India in May 1997 put major em-phasis on the importance of the proper preconditions and se-quencing for capital account liberalization, its message has not,according to one of the authors of the report, been properly inter-preted. Rather than, as the report proposed, getting on with estab-lishing the necessary preconditions and lining up the appropriatesequencing so as to be able to progress with capital account liber-alization, the policy lesson many have drawn has been rather thatliberalization itself is a poor idea.

63Of course, whether advice given earlier would have beentaken is quite another question. Quite possibly it would not havebeen, but that does not contradict the basic point that early warn-ing is better than late.

64By “collectively” we mean in terms of advice actually given bythe staff to the member. Within the Fund there were different levelsof awareness of the precariousness of the domestic financial situa-tion—especially in regard to Korea. Another consideration is thatwhile the Mexican crisis of 1994–95 had been in part a conse-quence of a weak banking system, this aspect of financial vulnera-bility had not been fully absorbed across Fund area departments.

Transparency, Publicity, Communication

viewed as the “explanation” of the lack of impact,was a politically important exchange rate peg; do-mestic political difficulties or uncertainty; and Fundadvice that was not, in the event, forceful or early.

Internal Reviews

20.The preceding section puts into relief one of thespecific questions posed to us, namely, “how suc-cessful have been the specific efforts since early1995 to strengthen surveillance?” With that in mind,we examined the Fund’s main internal reviews sincethen, and looked at the extent to which their recom-mendations had been implemented. 21.The review in 1995 was dominated by the expe-rience of the 1994–95 Mexican crisis and itsspillovers. It highlighted the need for members toprovide timely and accurate official data, and spe-cific initiatives were agreed upon to improve statisti-cal reporting. It was also agreed that the Fund shouldbe more aggressive in pursuing surveillance—inparticular with more continuity and follow-up,cross-checking official data against market informa-tion, and generally being more pointed in its analy-sis. The need to take a much more explicit accountof capital flows was recognized, as was also, al-though less emphatically at that point, the need forgreater attention to domestic banking soundness.22.From early 1995 to mid-1997 the world econ-omy experienced a relatively crisis-free period onthe financial front, and the biennial review con-ducted in early 1997 reflected this. The main con-crete result of that review was the introduction ofPINs. The staff document discussing surveillancethat provided background for the 1997 Board reviewdid note, rather circumspectly but not inappropri-ately given what transpired later in the year, “thatprogress has been made; that some pitfalls remain;65

and that further efforts are needed.” However, thesumming up of the Board discussion was much more

positive, in that it was recorded that “Directors ob-served that steady progress had been made, not leastin our ability to detect emerging financial tensions atan early stage.” The final report on the reviewprocess, in the spring of 1997 to the Interim Com-mittee, noted that “current surveillance procedureshave generally worked well.”23.Clearly, the events of the past two years showedthat this equanimity was not warranted. 24.On the evidence to date, the answer to theBoard’s specific question has to be negative. As il-lustrated in earlier chapters, and particularly in thesection immediately preceding this one, there has re-mained a significant gap between the Fund’s generalefforts to strengthen its surveillance procedures, es-pecially in regard to vulnerabilities, and satisfactoryoperating results.25.This gap was underlined by the Fund’s special,post-Asian crisis, review of surveillance that was un-dertaken in the spring of 1998.66 From this review,less extensive but more country-intensive than ear-lier ones, the Board drew five lessons: (1) the impor-tance of timely availability of accurate information;(2) more attention to banking system and capital ac-count issues; (3) greater attention in bilateral surveil-lance to policy interdependence and the risks of con-tagion; (4) the importance of transparency inimproving policy credibility and restoring marketconfidence; and (5) the fact that Fund surveillancewill only be effective if members take the advice. 26.The first two items on this list are reiterations of,and to that extent serve to reemphasize, themes inprevious reviews. The third, highlighting policy in-terdependence and contagion, is a new emphasis,and something we address in our conclusions. Thefourth, focusing on transparency, also remains anissue, and one that we will also be taking up belowand in our conclusions. The fifth, noting that mem-bers needed to be willing to take the advice, is anissue that we have confirmed, particularly in our ex-amination of four countries that went into crisis. Aswe pointed out there, the reasons why the Fund hadless impact than might have been hoped depend asmuch on the country as they do on the Fund, withpolitical factors being prominent.

Transparency, Publicity,Communication

27.This set of issues has attracted great attentionwithin and outside the Fund. In particular, the teamwas aware that the Group of 22 (G-22) Report on

53

65In this regard, particular note should be taken, especiallygiven the eruption of the Asian crisis a few months later, of thecautionary observation in paragraph 71 of that document: “Whilesubstantial attention has been paid to financial sector issues in sur-veillance, the coverage of these issues in Article IV consultationsappeared in most cases to have been backward looking. Moreover,it is difficult to infer from the language the extent of the risks seenby the staff. In part, this hesitation may reflect the fact that the ex-tent of banking problems often emerges only with a delay togetherwith practical limitations on the ability of the Fund to attempt toidentify banking problems in advance.” More generally, the staffreview also cautioned in the same piece that “the absence of majorcrises with systemic effects does not provide evidence that Fundsurveillance has become more effective. Such welcome develop-ments may reflect good management by countries unrelated toFund activities, good luck, or other factors.”

66The conclusions of the review are summarized in the 1998IMF Annual Report, pp. 34–38.

IV IMPACT

Transparency and Accountability, including a chap-ter on transparency and accountability of interna-tional financial institutions, had been released inearly October, soon after it started work. That reportmade the case for having “international financial in-stitutions adopt a presumption in favor of the releaseof information, except where release might compro-mise a well-defined need for confidentiality.”Against this background, the team raised trans-parency issues on many occasions, with particularfocus on the pros and cons of publication of ArticleIV consultation reports and, linked to this, on experi-ence with PINs.28.By way of further background, the team didmeet with individuals who had participated in the G-22 exercise. We were particularly interested in learn-ing of any progress in the transparency workinggroup in specifying what constituted “a well-definedneed for confidentiality.” We were informed that,unfortunately, no real progress had been made. Thiswas essentially because no agreement in principlecould be reached on whether to publish the ArticleIV consultation reports themselves. In those circum-stances, discussion of what might constitute appro-priate confidentiality would have been academic.29.The team’s interviews covered both the matter ofreport publication and how one might go aboutspelling out “appropriate confidentiality.” These willbe reported on below. However, before that, it is use-ful to discuss the experience with PINs.

Experience with PINs

30.Most countries in our sample have authorizedPINs. Among officials, the consensus was that theintroduction had gone quite smoothly, at least in thesense that the PIN release had not made a splash buthad made news in a more sober fashion. Indeed, in anumber of instances the news had apparently beenso low key that a number of the people interviewedin the private sector, while quite knowledgeableabout the Fund, could not recall PIN-based reports inthe media. In any event, the consensus was that re-lease had been constructive in helping to inform anddemystify, and ripples were probably better thanwaves. As one official observer in Asia put it, it wasjust as well for the PIN not to be “sensational,” evenif this meant that the impact was not as great as itmight have been.31.This being said, there were reservations on lan-guage and substance, although it must be empha-sized that these were far from being from the samedirection.32.A worry that went to the very heart of the con-cept of the PIN as a vehicle for communication wasover the extent to which the PIN might be abused bythe media rather than merely used. This was ex-

pressed by a minority, vehemently, and seemed torelate essentially to the media’s continual search formaterial that could embarrass a government or atleast make people sit up. PINs have indeed occasion-ally given rise to embarrassment. In one case the au-thorities suggested that they might discontinuePINs—the first had gone well, but the second hadgiven rise to headlines emphasizing, or perhapsoveremphasizing, Fund criticism. At the same time,it should be pointed out that the country’s economicand financial situation was more challenging by thetime the second PIN came around, in 1998, than itwas the first time.33. In any event, PINs are evidently drafted verycarefully. Indeed it is apparent that the authors takegreat pains to avoid loose, exploitable, words orphrases—above and beyond the question of the dele-tion of “highly market-sensitive” material that is aFund Board precondition for their being authorized inthe first place. However, this has led to a concern thatpoints in the opposite direction from the criticism justmentioned. That is to say, a common complaint wasthat the language then became so bland, so “unplain,”that it failed to communicate well—which, after all, isthe ostensible purpose of the exercise.67 No onewould mistake the PIN for a press release, even if itsoriginal name was pressinformation notice. As oneexperienced businessperson in Latin America put itwhen shown a PIN for his country, “the language isvery much like that of auditors—precise, but not easyto pierce if you are not familiar with it.” 34. In terms of substance, we should underline that therevisions to the summing up of the Board discussionthat are made before a PIN is issued have on occa-sion—and contrary to the spirit of the PIN exercise—gone beyond the exclusion of market-sensitive infor-mation. Deletions have also included material that wasthought to be politically embarrassing by national au-thorities. We know that the staff and the Board areconscious of this weakness, including the possibilitythat the summing up may itself be watered down exante to avoid having such changes pressed when thesummary is sent to the Executive Director concerned. 35.Two Board reviews of PINs have taken placesince their inception, and it is apparent that after thefirst review there was a noticeable drop in the num-ber of deletions. However, to what extent this droprepresents greater restraint on the part of Directors

54

67By way of example, fairly randomly chosen, the reader is in-vited to peruse the following PIN sentence for the main message:“[Directors] stressed, however, that fiscal consolidation at bothprovincial and federal levels had been reflected in substantial ef-forts to contain the growth in outlays for medical services and toimprove the efficiency of the health care system, but they weredoubtful that that approach could be relied upon to contain the in-creases in health care costs that were likely in the future.”

Transparency, Publicity, Communication

or more concern on the part of the staff to avoid pro-voking changes is not clear to the team. Evidently,while the PIN represents progress in explainingFund surveillance, it is still work in progress both interms of its own definition of what it aims to be andin terms of enhancing Fund transparency. 36.Anticipating the discussion of Article IV consul-tation reports that follows, it is also worth recallinghere that a number of countries release the staffstatement that is given to the national authorities atthe conclusion of an Article IV mission.37.This is at the least curious. The concluding state-ment represents, first, only the staff ’s view, and sec-ond, its view before it leaves the country and actu-ally writes the consultation report that will beconsidered by the Board. That report is the morecomplete and considered view of staff. It has the im-primatur of management. And the consultation is notin principle concluded until the Board has consid-ered that report and said whatever it has to say. Then,of course, a PIN may be issued on the basis of thesumming up of the Board discussion. In fact, virtu-ally without exception, the countries that have re-leased the concluding statement have also autho-rized PINs. Any possible contrast between thosestatements and what is said in the correspondingPINs has not been an issue to date.38.When queried on this situation, country repre-sentatives (including here Board members) allowedthat it was at the least untidy, but in the words ofone, “at this stage, more transparency, more publica-tion, is better than less.” Others, whose authoritieshad not released the concluding statement but had is-sued a PIN, were concerned at the evident break-down in orderly procedures and the consequentialdevaluation of the Board discussion.39. As a final comment, it may be that for publication“hawks,” issuing a concluding statement has the ad-vantage that its language does tend to be more directthan that typical of a PIN and to that extent communi-cates better. It also represents a single view rather thana summary of differing views, and this would alsopresent a better package in terms of communicationprocess, if not necessarily in substance—that is, peeras opposed to staff surveillance.

Article IV Consultation Reports

40.Our discussions of transparency issues and whatthe Fund might publish focused very much on thesereports—the basic document of the country consul-tation process. The report embodies an analysis ofeconomic and financial developments, a discussionof policies, a discussion of staff (or the authorities’)views, and the authorities’(or the staff ’s) reactionsto them. In addition to the basic question of whetherto publish or not, our search for views on what might

constitute appropriate confidentiality included ascer-taining whether there might be analogies outside theFund surveillance framework per se. We begin witha general discussion of confidentiality in principleand practice, and then turn to the question of publi-cation of the Article IV staff reports.

Confidentiality: Principles

41. We were told on quite a few occasions that if theFund published the results of its consultations, thenthe quality of discussion would surely suffer. Thiswould be even more the case if there was any attemptby the Fund to deepen the consultations through moreanalysis and exchange on hypothetical risks facing acountry. Such hypothesizing is done rather little now,and the officials with whom we discussed this possi-bility were noncommittal about undertaking it—withor without the protection of confidentiality.42. As noted already, there was almost total agree-ment—and not just among officials, we should em-phasize—that some part of the consultations betweenthe Fund and a member might well be kept confiden-tial. One yardstick, in the terms of the PIN, is materialthat is “highly market-sensitive.” This evidently refersto the exchange rate, but doubts were expressedwhether it really extended beyond that, to interestrates for example. More broadly, a number of offi -cials, and some academics, thought that there werealso politically sensitive issues that countries were en-titled to keep confidential, including in relation to aPIN—for example, impending legislative proposals.One close and experienced observer of the Fundscene thought that in any event for many countries thetruly unique and valuable part of the reports was thestatistical tables, given that the Fund staff puts a lot ofeffort into getting country statistics into internation-ally comparable shape in the main financial areas—fiscal, monetary, and balance of payments. Of course,statistical tables—but not forecasts—are now pub-lished in the RED, although it was not clear howmany outsiders realize this. He wondered whetherthere was much need for the publication of Fundopinion. A number of other commentators thoughtthat a start at least would be to publish the statistics,although it was less clear whether this injunction cov-ered forecasts as well. Quite possibly not.43.Some more conceptual discussion of the confi-dentiality question took place with central bankers.They have had to think about the basis for confiden-tiality quite a lot in recent years, given the thrust to-ward increased transparency in monetary policy de-cision making. One take on this issue was thatconfidentiality was justified to the extent that it wasnecessary to do the job mandated for the institutionby the legislature. Arguing for confidentiality inthose terms implies being very clear on what are the

55

IV IMPACT

objectives of the Fund’s surveillance mandate. An-other view, in another country, was that transparencywas particularly important in describing the consid-erations that lay behind specific policy decisions.However, it was not considered at all necessary todivulge—in the published minutes of monetary pol-icy discussions, for example—analyses of hypotheti-cal situations, or “what if” questions. An awkward-ness of this analogy for our purposes is that little ofthe material in a regular Fund consultation report ac-tually culminates in a specific decision by the Board.44.Finally, one interviewee, a U.S. academic, raisedthe possibility that progress in spelling out confiden-tiality guidelines might be made by considering thedistinction between microeconomic risk events andmore macroeconomic ones. Thus, the Fund might bewell advised to be very public as regards systemicrisks—for example, where a banking system hasbeen borrowing heavily in foreign currency andlending on in local funds—but not so in regard tohow far particular institutions themselves were fromthe precipice. In this latter situation, which can alsowell represent the tricky situation sometimes facedby a banking supervisor, it could be bad luck (or, asthe academic put it, “a stochastic event”) that wouldmake the difference. To the team’s mind, however, itwould also make an important difference how goodthe Fund was, or was allowed to be, in practicingearly intervention. Clearly, if it rang the alarm belllate, this could itself bring down the banking system,as well, of course, as individual banks. Indeed, onecentral banker drew explicitly the analogy betweenFund surveillance and that of a banking supervisorin terms of the dilemmas faced. In that regard, henoted, transparency looked better on paper than itdid in practice. Of course, we can add, the analogymay not be perfect because presumably a supervisorcan take direct steps to remedy a bad situation,whereas the Fund can only cajole, whether publicly,semipublicly, or confidentially.

Confidentiality: Practicalities

45. In terms of the practicalities of making surveil-lance analysis and conclusions more public, a num-ber of points were made.

• It was obvious, said some, and as alreadynoted, that at least statistical data could be re-leased. But, we also note, it is doubtfulwhether data alone would help much in pro-viding a basis for accountability on the qualityof policy advice.

• It was suggested that lags in publication couldhelp to reconcile the polar views on keepingthings confidential, provided however that thelag was not very long. The Fund has recently

decided to cut publication lags, but the lags arestill years rather than, say, months.

• Perhaps, some wondered, there could be ashorter, confidential, report alongside the con-sultation report itself. However, others sawthis as nonviable since it would become knownthat there were “Report I and Report II,” andthe same issues of publication and confiden-tiality would arise again.

• Some interest was shown, obviously in regardto situations of particular concern, in a phasedprocess where the Fund was, initially, a “silentwhistleblower” engaging with a country in a di-alogue “on behalf of the rest of the world.”Then, the advice could become gradually moreopen through other vehicles such as speechesand WEO. As one commentator put it, it washighly desirable to avoid the “nuclear option” incases of imminent vulnerability, potential conta-gion, moral hazard, etc. So, mechanisms for amore progressive approach needed to be found.

• A point made related to the above was that instarting publication, timing is important. As in the case of a PIN, it is desirable to establishthe process in a period of relative marketcalm, rather than to launch it at a time that isless settled.

• Peer pressure was considered by some to be apotential avenue of persuasion short of goingpublic. But no one could point to examples ofsuccess. As one Asian central banker put it,“countries are just too polite to each other.”Whatever the criticisms of Fund missions tothe effect that they are too accommodating tothe countries they deal with, they are not seenthis way in relation to other available vehiclesof surveillance. In fact, Fund reports were gen-erally seen as more rigorous than those ofother international organizations—in particu-lar those of the OECD, which are published,but only after country vetting, or those of theEuropean Commission.

• Finally, one seasoned observer, formerly astaff member but now outside, emphasizedthat it is all well and good to have rules abouthow and with what to go public, but there al-ways remains the question of the will to do it.This is not an academic point, given the factthat the Fund has time and time again showngreat reluctance to deploy the sanctions againsta member that are already at its disposal.

Publication of Staff Reports

46.As can be readily gathered from Board discus-sions and the G-22 report, views were quite polar on

56

Communication Within Government

the desirability of publication of Article IV staff re-ports.68 All the same, the views we collected wereprobably more varied and nuanced than in the abovebecause we were not seeking the official view butrather insights on the question.47. In regard to considerations favoring publication,most emphasis was placed on considerations relatingto accountability and incentives. Much was made,above all in the private sector, of the importance ofmaterial being available on which to judge the qual-ity of Fund advice. It was contended that this expo-sure would help to ensure that advice stayed good, orgot better if not good. On the other hand, consider-ably less was made of the view that publicationwould help markets function better generally—onthe basis of presumably more complete informa-tion—although this view was also evident in the pri-vate sector. As one businessman in Latin Americaput it, he thought that the additional informationwould provide a better perspective on the risks heand others faced, and then markets would work bet-ter. Then again, another observer, from a debt-ratingagency, was concerned about “volatility feedback”resulting from publication—at least as a possibilitythat could not be disregarded.48.On the other side of the ledger, concern was par-ticularly intense over the effect of publication incausingcrises rather than preventing them, and theconsequential “moral hazard” issue facing the Fund.More generally, the point was made more than oncethat the Fund was predominantly a provider of ad-vice to its members rather than a “whistleblower.”Furthermore, publication of that advice, in particularin regard to possible future policy moves, would ad-versely affect the nature of the relationship, whetherthe danger of causing a confidence crisis wasthereby increased or not.49.All this being said and recorded, no doubt not forthe first time in the extended debate that has takenplace in different forums, there was general accep-tance among propublication ranks (as also recordedin the G-22 report) that some material emanatingfrom consultation discussions would probably needto be kept confidential. We discuss this issue inChapter V. But before concluding this section it isimportant, given the thrust of our recommendations,to convey our understanding as to what kind of dis-tribution Article IV reports actually do receive aftercopies are produced.50. What is very clear is that Article IV reports findtheir way quite readily outside the circle of autho-

rized users. It was obvious from our interviews thatoutside circulation was quite common69—whetheramong financial institutions, debt-rating agencies,the media, etc. Furthermore, not all the distribution iscompletely unauthorized. In particular, national au-thorities evidently quite often make the reports avail-able to debt-rating agencies, which in turn use themin coming to their assessments of sovereign risk.These agencies told us that they find them useful.51.Another feature is that because of this reality, orperhaps as a general precautionary measure, the re-ports themselves are apparently unlikely to reflectthe full extent or depth of the consultation discus-sions that actually take place. This was confirmedboth by country authorities, some of whom said thatthey spoke to Fund staff on the understanding thatnothing that was not already effectively publicwould appear in the written staff report, and by Fundstaff, who made it clear that they practiced a sub-stantial degree of self-censorship. According to a se-nior staff member, “nothing really confidential ap-pears in a staff report.” Looked at from anotherangle, it is interesting to note that in a couple ofcases academics had, as a condition of being inter-viewed, insisted to their authorities on being givenaccess to the relevant Article IV reports. Their reac-tion was that there was not much material in the re-ports that they did not know already from other, reg-ular, sources of economic and financial news anddebate. So they wondered what the fuss was about. 52. Interpreting why, given the fact of general leaks,there may be concern about publication even infairly “open” societies, our impression is that it mayimportantly reflect worry about adverse media head-lines that then have to be dealt with. As was alsonoted in our discussions, in many places the Fund isseen as very powerful, so its advice simply has thatmuch more popular impact. Wide release could,therefore, cause embarrassment even if all the in-vestment houses in the world had already obtainedthe relevant report and made of it whatever theycould.

Communication Within Government

53.The issue here, simply put (and spelled out moreformally in the specific question (x) put to the team),is whether the advice got to the right people in gov-ernment. At one level, it can refer to whether the ad-vice was distributed widelyenough. At another it can

57

68We should note that in April 1999, the Board did agree on acompromise pilot project of publishing Article IV consultation re-ports for about 20 countries on a voluntary basis—to be reviewedafter 18 months. Our own views on publication come in Chapter V.

69Bear in mind that they are sent, among other things, to 182countries. In this regard, we were told of one instance where thepublic had access in a country’s central bank library to all ArticleIV reports.

IV IMPACT

be whether the advice went high enough in the deci-sion-making apparatus.54.What is apparent from our discussions is thatpractices vary a great deal in both respects. What isless apparent, without a far more extended andsearching discussion than the team was able to un-dertake, is how much difference this has made.However, communication within the governmentdid not appear to be very widespread. In a couple ofinstances (that is, two out of the four cases discussedin detail earlier), there had been a clear breakdownin communication between the central bank andsome other parts of the government in the periodleading up to a crisis. In one case this was com-pounded by frequent changes in government. On aless pathological note, it is worth recording thatthere was a general tendency for Fund advice to re-main, if not entirely at least very greatly, the prop-erty of the ministry of finance and the central banktogether. This may well reflect bureaucratic preroga-tives, besides any more general considerations relat-ing to who really needed to know what. At the sametime, it was also pointed out in response to this lineof questioning that distribution of Fund material be-yond the two institutions referred to could compro-mise confidentiality. Prominent officials in otherministries, even when visited by the Fund team andin cases where their area of work was the subject ofrecommendations in the staff report, did not neces-sarily get to see those recommendations—or the fullArticle IV report. However, it also seemed that the

recent tendency in countries has been to circulateFund surveillance conclusions (but more likely themission’s concluding statement than the Board sum-ming up) more widely within government ratherthan less. This would follow particularly from thefact that the surveillance conclusions have in recentyears ranged beyond issues of fiscal, monetary, andexchange rate policy. Still, there was no indicationanywhere that the circle of participants went beyondeconomic agencies, or that any thought that this wasdesirable or necessary.55.As regards more vertical dissemination, our im-pression is that this would generally be quite limitedand almost invariably in a summary form—some-times reported upon orally around the cabinet tableand sometimes circulated in the form of a cabinetmemorandum. A minority of countries have devel-oped the habit of putting the staff concluding state-ment (but not, it appears, the PIN as well) on theagenda of a cabinet meeting. At the same time, it isnot unusual for the Fund’s views to be used in inter-nal cabinet debate to bolster a policy position. A reg-ular pattern seemed to be for the views to be used tobolster the case for fiscal restraint. As one senior fi-nance ministry official in Asia put it to us, while hewas not convinced that it was necessary to go as faras the Fund wished in cutting the deficit, ministerstended to want to do less than was needed in this di-rection. So perhaps, he suggested, the net result ofexposing politicians to the Fund’s views was aboutright.

58

1. As can be seen in the preceding chapters, theteam reviewed a wide range of surveillance issueswith a wide variety of parties—as called for underits terms of reference. Correspondingly, our findingsand recommendations also cover many different as-pects of surveillance—both in regard to the Fund’sdealings with its members and in regard to its inter-nal procedures.2. However, before going through the individualrecommendations together with the reasoning be-hind them, we want to present a broader picture ofwhat we found.3. Not surprisingly, and paralleling the experienceof member countries, Fund surveillance has beenplaying catch-up to the globalization of financialmarkets and the vastly increased importance of pri-vate capital flows. This has shown up very clearly ininternal post mortems on the Mexican, and then theAsian, crises. One challenge for the Fund, alongwith everyone else, has been to reach an adequateunderstanding of the forces at work on the globalcapital scene—including their role in the phenome-non of contagion. In that regard, we think that in itsex post internal reviews the Fund has identified quitewell the areas where action needed to be taken in re-gard to broad surveillance approaches, both by coun-tries and by the Fund itself. However, it appears tohave had some difficulty translating its broad inten-tions into corresponding action in the field. Notwith-standing the lessons of Mexico, the implications,particularly in terms of capital account risks and vul-nerabilities in the domestic financial sector, did notregister uniformly, and with the appropriate inten-sity, throughout the institution. We say this even al-lowing for the fact that translating broad intentionsinto solid actions inevitably takes time. Indeed, oneof the weaknesses of the Fund’s internal workingsthat drew our attention was the difficulty it has intransferring relevant knowledge and experienceacross the organization.4. It is also noteworthy that this process of catch-upto globalization is continuing at a rapid pace—as ev-idenced by the emphasis in recent months on the de-velopment and monitoring of international standardsof conduct over a range of economic and financial

activities and on the development of the ContingentCredit Line as an additional source of liquidity. Ac-cordingly, the team found itself to a degree focusingon a moving target. We discuss later in this chapterthe various implications for surveillance of thesenew international initiatives. However, we werepleased to see that many of those implications onlyserved to reinforce points we had emphasized inlooking at the surveillance framework that was al-ready in place.5. Another challenge in evaluating the effective-ness of surveillance was the evident gap—one thatis, it should be emphasized, more apparent outsidethe Fund than inside—between the general impres-sion, or expectations even, of the impact of surveil-lance and the Fund’s actual degree of “clout.” It istherefore worth reiterating the staff ’s observation inthe 1997 biennial review that “Fund surveillance isonly one influence, and generally not the predomi-nant one, on members’policy and performance.” Wecan confirm this in regard to the range of countriesthat we surveyed—in particular through the study ofthe four countries that went into crisis in the recentperiod; see Chapter IV. So while Fund influence willno doubt vary by country and within a country overtime, depending on particular circumstances, itsurely has to be accepted that surveillance is hardlyever going to be a primary influence on a country’spolicy actions. 6. The best the Fund can realistically hope to do iscontribute over time to building or maintaining aconsensus across the membership on the broad pol-icy framework; and to tip the balance in situationswhere policies are in fact genuinely in the balance.However, the latter requires the existence of a rea-sonably clear domestic political will or consensus tochange policies if necessary—something that was,for example, certainly not apparent in the four “vul-nerabilities” cases that we focused on. Indeed, wewere left with the impression that a financial crisismay be, unhappily, a crucial ingredient in concen-trating the necessary political will—or, alternatively,that domestic political considerations will almost al-ways dominate over Fund warnings, even in the faceof a looming crisis. Part of the task regarding sur-

V Conclusions and Recommendations

59

V CONCLUSIONS AND RECOMMENDATIONS

veillance may therefore be seen as seeking mecha-nisms to concentrate that will before a crisis strikes.Here, more transparency, in the sense of publicity, insurveillance conclusions is obviously an importantissue, and some of our recommendations aim to en-courage further movement in that direction.7. Turning to a review of the broad thrust of ourconclusions, one of the key elements is the need tofocus surveillance more directly on the interna-tional aspects of a country’s situation, the linkagesacross countries, and the lessons countries canlearn from other experiences. Member countrieshave expressed a lot of interest in such material,and here the Fund has substantial talent and advan-tages that are not yet used to best effect. Further-more, to maximize the effective use of those re-sources in the international interest, moreflexibility in monitoring country situations is calledfor. In particular, more resources should be devotedto multilateral and cross-country issues and less toindividual Article IV assessments of industrialcountries—other than the very largest, where theimplications of their policies for the internationalfinancial system should be the focus.8. As regards the general substance or quality ofadvice, we found it was generally good in the Fund’score areas of macroeconomic and financial analysis.It was less surefooted when it ventured into morestructural areas. But, in any event, Article IV consul-tations themselves should, while becoming more in-ternational and less “bilateral” in approach, also bemore focused on a relatively tight range of core is-sues. Ensuring this focus needs not only strong guid-ance from management, but also Board support inlight of the ever-present pressures and inducementsto expand the scope of the Fund’s activities in coun-try surveillance. Our judgment is that this expansion,which continually adds to the surveillance load, de-tracts from its effectiveness. 9. We believe that the time has come to publish allArticle IV consultation reports. Those reports al-ready get significant informal, but privileged, distri-bution outside. However, broad, regular, distributionshould yield important gains in both accountabilityand understanding. Besides improving the account-ability of the Fund generally, publication should helpalso in maintaining or improving the quality of ad-vice by exposing its views more to outside assess-ments. This will also help to limit “clientism”—theinevitable pressures on staff to avoid raising clearlywith a given country difficult but important issuesthat should be addressed in the interests of interna-tional financial stability.10.We also tackle a range of internal issues wherewe believe that shifts of policies or procedurescould benefit the surveillance effort. In a number ofthese, in particular the problem of inadequate

knowledge transfer, increased publicity for consul-tation reports will help. However, in questions ofstaffing and internal organization, more direct stepsappear to be called for.11. Mention should be made here of a quite distinctinternal issue, that of the role of the Board. It is be-yond our scope to deal with the underlying gover-nance issues that exist. However, as regards surveil-lance, we are strongly of the view that the Board’scontribution would be greatly improved by the ef-fective use of a committee structure.12. In the next five sections we present our detailedfindings and recommendations. The first deals withobjectives and priorities, with particular emphasis onwhere surveillance might usefully be expanded andwhere cut back. We then look at the quality of ad-vice—appraising Fund counsel in some particularareas, but also examining broader questions relatingto the Fund’s general approach in developing its ad-vice. This is followed by a section that tackles anumber of organizational issues: internal organiza-tion and resources; the style and intensity of contactswith member countries, and the extent to which suchcontacts might usefully vary across members, i.e.,“selectivity”; and finally here, the role of the Boardin surveillance. The fourth section deals with com-munication, above all with the much-debated issuesof transparency and confidentiality in surveillance.Finally, as already indicated, we comment on twonewer developments that have substantial implica-tions for all aspects of surveillance—the growingemphasis on monitoring rules of conduct over arange of financial and economic activities and, fi-nally, the Contingent Credit Line.

Objectives and Priorities

International Added Value


13.Fund multilateral surveillance is of generallyhigh quality, and its work in this area, particularly asexpressed through WEO and the ICMR, receivedmuch favorable comment. Some particular issues aredealt with later in this chapter, but the basic pointthat these publications are well regarded should berecorded here. 14. If there is a broad shortcoming, it is the lack ofintegration of multilateral and bilateral surveillance.The staff ’s comparative advantage is in analyzing in-ternational systemic issues and economic interde-pendence, not least when the latter gives rise to vul-nerabilities. A number of our recommendationsbelow focus on how the multilateral and interna-tional expertise of the Fund staff can be harnessed tostill greater advantage.

60



15.This is the main focus of our attention in the areaof international value added. Simply put, we, andmany of those we spoke to, found the Article IV con-sultation process to be insufficiently international infocus—too “bilateral,” in fact. This is also the impli-cation of the Fund’s own “third lesson” in relation tosurveillance and the Asian crisis.70

16.Given the shift in the international environmenton account of the vastly increased importance of pri-vate capital flows, the increasing openness ofeconomies through trade, and the increased sophisti-cation in many countries in the field of domesticmacroeconomic policies,71 the Fund’s comparativeadvantage now lies at least as much in bringing in-ternational considerations and experience to bear asit might in giving broad advice about domestic poli-cies per se. However, this is not the way Fund con-sultation discussions are shaped at present.

There are three different aspects of Fund value-added to consider here:

• the desirability of a Fund mission explicitlybringing to the consultation table its view ofthe international economic and financial situa-tion and prospects, with particular emphasis onthose features that are most relevant for thecountry in question;

• a more systematic attempt by staff to explorebeforehand and then volunteer pertinent expe-riences from other countries, whether in termsof issues or policies; and

• more direct focus on how a member mightincur and absorb shocks coming from the out-side, whether they come through direct eco-nomic interrelationships and spillovers orthrough more general contagion from marketdisturbances. Even if, as may well be the case,the member is not inclined to respond directly,or at all, to these hypothetical issues, it wouldstill be worthwhile to raise them consistently,thus leaving an imprint.

17.For the very largest countries or policy areas(i.e., the euro area), the focus in this third aspectwould be as much on the implications of their ownpolicies for the rest of the world as on the likely im-pact from developments elsewhere.18. Besides the direct advantages of explicitly includ-ing such considerations in the consultation process,this more international focus would have the impor-

tant collateral gain of forging greater linkage withinthe Fund between the multilateral and bilateral aspectsof its surveillance work. These links, as discussed inmore detail in the section on organization, below, arequite weak.

We recommend that consultation guidance berestructured to give explicit attention to interna-tional aspects along the lines indicated above.


19. In the Fund canon, “regional” surveillance is notgenerally seen as a separate category, but here itmerits separate discussion. Internal reviews of sur-veillance have given some attention to regional as-pects. However, this attention has been more on themechanics of the institutional side, focusing onevolving regional economic and financial associa-tions such as the European Union, and monetaryunions such as the CFA franc zone and the EasternCaribbean Currency Union. At the same time, an-other set of regional-type issues—regional economicand financial spillovers and contagion, and the po-tential for constructive application of regional peerpressure—has been driven increasingly to the foreby recent events in Asia and Latin America.20. The first, institutional, set of questions has to dealwith how the surveillance process copes with evolv-ing regional arrangements. For the euro area, as notedin Box 2.2, the challenge of finding the appropriateset of surveillance arrangements is particularly com-plex, involving as it does political as well as technicalconsiderations. However, despite this complexity, thereplacement of 11 monetary and exchange rate poli-cies with one, formulated exclusively at the euro arealevel, should lead to significant resource savings.Clearly surveillance over monetary and exchange ratepolicies—the raison d’être of Fund surveillance underArticle IV—will have to take place via the Europeaninstitutions. However, even in the case of fiscal pol-icy, given the intensive, and intensifying, degree ofmonitoring at the euro area level, we believe that theFund is much more likely to have an impact if it con-centrates its efforts at this level.72

We recommend that surveillance of the euroarea center around the ECB and other EU bod-ies. Surveillance of individual participants inthe euro area should largely take place at theeuro area level, and through EU institutions.

21.As regards the second set of regional issues, theFund has taken some initiatives to address them

61

70IMF, Annual Report, 1998, p. 37.71Our conclusion is that in a preponderance of cases it is not an

understanding of economic policy principles that inhibits a coun-try from doing “the right thing,” but rather, as we have already in-dicated, the challenge of summoning up the critical mass of polit-ical will to do so in a timely fashion.

72The same is true for structural policies. But, in any case, as setout below, we recommend that the Fund aim to reduce the resourcesdevoted to the surveillance of structural policies across the board.


through its participation in regional meetings, in-cluding at the highest level of management, and, in-deed, in taking the initiative to arrange a meeting ofministers of finance and central bank governors ofthe Americas in September 1998. On the evidence todate, what surveillance results should be expectedfrom such gatherings? In our view, probably notmore than broad familiarization and consciousnessraising. This is because even where country policiescause concern among their neighbors, peer pressureis unlikely to be applied very actively—and in anycase convincing evidence of any significant impacthas yet to be produced. Furthermore, the Fund is not,and practically speaking cannot be, more than a fa-cilitator of such meetings,73 with any associated ex-hortations from senior management that might giveincreased focus to the event. At the same time, theFund documentary contributions have not been asmuch to the point as they could, and should, be.They have been focused more on supplying materialthat exposits individual country situations than inhighlighting the regional interactions and issuesagenda that could usefully be explored if partici-pants really had a mind to do so in such a forum. Thesame is often true of internal Fund material; for ex-ample, that presented at the Board’s Informal Coun-try Matters sessions, which were originally intendedto have a regional focus.22.That being said, we suspect that even if the Funddid prepare more compelling documentation alongthe lines just indicated (which it should of courseaim to do in any case, since it is the right thing to doeven if it ruffles a few country feathers), the diplo-matic niceties of large gatherings would continue toconstrain direct discussion, and hence any resultsfrom peer pressure.74 Accordingly, we do not seethat regional surveillance in this form can in the endrealistically be a substitute for country surveillance.However, country surveillance should in any eventemphasize regional and spillover issues.

We recommend that (consistent with an in-creased focus on international aspects) theFund bring spillover issues, whether regional ormultilateral, directly to the table in its variouscountry consultations and in Board discussions.

23. Conceivably, this kind of focus could also lead,over time, to more fruitful regional exchanges—at themargins of such meetings if not at formal sessions.

Cutting Back

More Concentration on the Core of Surveillance

24.The recommendations just made fit squarely intothe Fund’s statutory mandate to exercise “firm sur-veillance over the exchange rate policies of mem-bers.” However, the same cannot be said of other el-ements that have come to be included in bilateralsurveillance.25. In recent years, the surveillance agenda has ex-panded rapidly. This expansion into nonfinancial,structural areas was described earlier, in Chapter I,and will not be repeated here. What should be noted,however, is that the pressures for expansion comefrom more than one source and may also reflect dif-ferent priorities.

• Governments and/or their legislatures. Here,the outstanding example has been the recentU.S. law authorizing the latest increase in theU.S. quota in the Fund. This legislation re-quires the U.S. government to press for awhole series of initiatives to be undertaken bythe Fund. These include promoting, besides“market-oriented reform, trade liberalization,and economic growth,” also “democratic gov-ernance and social stability” in member coun-tries. Doing this will involve, among otherthings, “establishing or strengthening elementsof a social safety net” and “the maintenanceand improvement of core labor standards” aswell as pursuing “macroeconomic stabilitywhile promoting environmental protection.”Another very recent example is the June 1999G-8 communiqué, which calls upon the Fundto give “more attention” to “the developmentof sound social policy and infrastructure in de-veloping countries” and “to give particular pri-ority to core budgets such as basic health, edu-cation, and training to the extent possible, evenduring periods of fiscal consolidation.”

• Nongovernmental organizations. In some areasNGOs are also pressing for similar expansionsof the Fund’s role to those just described (seealso our report on discussions with NGOs inChapter III).

• The Fund’s own view of its evolving role. It hasincreasingly concentrated on medium- to long-term issues such as stabilization sustainabilityand growth, which it now considers key to sur-veillance. In this regard, it perceives a need toexamine a wider range of issues, micro or struc-tural in nature. In addition, the Fund has ofcourse been attempting to deal with the de-mands for any additional activity in surveillancealong the lines just noted in the previous bullets.

62

73As it also is at G-7 meetings. 74It is interesting to note that in the most thorough published

analysis of surveillance, Guitián (1992), who was then a seniorstaff member, saw the key challenge as how—in the absence ofany coercive power—surveillance is to be made effective throughpeer pressure. See Manuel Guitián, The Unique Nature of the Re-sponsibilities of the International Monetary Fund, IMF PamphletSeries No. 46 (Washington: International Monetary Fund, 1992).


26.Our basic concern is that raising a broad set ofissues in a consultation will not enhance, but ratherdilute, the quality and impact of surveillance advice.For one thing it can, as we were told, lead to confu-sion on the part of national authorities as to what infact is the basic thrust of that advice. For another,while the staff are well equipped to deal with macro-economic management and international financialquestions and perhaps in time even structural ques-tions relating to the financial sector, they cannotcredibly lay claim to nearly the same level of exper-tise in nonfinancial structural and/or microeconomicquestions, social policy questions, environmentalquestions, and the host of other issues that are beingput on the agenda. And this apparently shows in con-sultation discussions. To be sure, we were told ofcases where a team had indeed been effective in pro-viding advice in noncore areas. But the basic issue israther whether Fund resources as a whole would bebetter applied, and surveillance thereby becomemore effective, if the focus were more targeted to thecore topics. We believe that there would be an im-provement. This is especially so given the evidentneed to focus, in addition to domestic macroeco-nomic policies and exchange rate matters directly,more on financial system and capital account issuesin light of their propensity to generate exchange ratedisturbances. This conclusion is strengthened furthershould the Fund find merit in the increased interna-tional perspective that we believe it should alsobring to the bilateral consultation process. 27.Evidently, the pressure for expanding the scopeof surveillance comes in important measure fromevolving views as to what needs to be addressed inFund programs rather than from the evolution of sur-veillance itself. The design of programs is outsideour mandate, and we shall not deal with the contro-versial questions as to what needs to be included inthem. However, while we can certainly see the logicof the argument that because the Fund might beobliged to involve itself in a wide range of economic(and perhaps even noneconomic) issues in programs,these should also be put on the agenda for surveil-lance, we are not persuaded by it. The reason it is notpersuasive is because we believe that surveillance,even in the relatively narrow, relatively “tradi-tional,” sense in which we see it, is difficult enoughto do successfully without adding to the menu arange of other kinds of interventions whose rele-vance depends on hypothetical future needs.28. Accordingly, given the Fund’s competencies, andgiven the evident need for further improvement in theFund’s work on exchange rate policies, and the asso-ciated macroeconomic and financial framework, par-ticularly in regard to the international dimensions, itis our view that the quality (if not the quantity) ofsurveillance would be better served if the Fund were

to concentrate more centrally on the above areas. Thealternative, if one takes literally the demands nowplaced on surveillance, is a vast expansion, withoutany foreseeable limit, of varied and specialized re-sources devoted to Fund monitoring of economies.29. It is important to emphasize that a more centralfocus on what might be considered the traditionalcore of surveillance does not mean that importantlinkages between macroeconomic management andother policies should not be recognized and ad-dressed as part of surveillance. However, the crite-rion for addressing other issues should be the extentto which they actively and directly impinge upon theeffective conduct of macroeconomic policy. To givea specific example, labor market policy clearly canhave implications for macroeconomic management.But the Fund should normally confine its advice andanalysis to those implications—and not attempt toresolve the more general question of what “good”labor market policy might be.75 Moreover, the bur-den of proof should rest upon advocates for includ-ing additional items to show how they would im-prove the effectiveness of surveillance. This toowould help focus the overall surveillance agenda.

We recommend that surveillance focus, aboveall, on the core issues of exchange rate policyand directly associated macroeconomic poli-cies, in particular the international implica-tions of such policies. Other analysis shouldonly be undertaken if directly relevant.

30.We recognize that the views just expressed re-garding a tighter focus for surveillance may not fittoo well with the kind of role foreseen for the Fundin the application of international standards or codesof conduct. In particular, this role, as currently en-visaged, may involve monitoring standards in fieldsthat are less clearly financial and more microeco-nomic (e.g., accounting standards and governance).It seems to us that here the Fund has been viewed asnot so much the chosen instrument as, practicallyspeaking, the only available instrument, given itscurrent surveillance infrastructure and global mem-bership. We discuss this issue in the section belowon recent developments. 31.Our view implies that the Board should act as askeptical, restraining, influence on any widening ofthe surveillance agenda beyond its core. In otherwords, the presumption the Board should adopt isthat of a limited approach, focused on well-specifiedcore topics in which the Fund has a clear compara-tive advantage.

63

75Of course, we recognize that what is directly relevant tomacroeconomic management will differ from country to country.Labor market policy may be directly relevant in one country butnot in others.


32.Turning to individual bilateral consultations, wesee a similar need to set priorities. At the moment,staff feel obliged to produce reports that cover theentire surveillance agenda of the Fund. Even withour proposed refocusing and reduction of thatagenda, the pressure will remain to produce a com-prehensive report that touches relatively lightly on alarge number of issues, rather than one that focusesin more depth on a limited number of key issues. Webelieve that what is required is a systematic structurethat ensures that bilateral surveillance takes the latterapproach.

We recommend that a systematic process bedeveloped whereby the Board would discussand sign off on the main issues to be raised atforthcoming individual Article IV consultationdiscussions.

33.Proposals in this regard would be submitted bythe staff, and the Board could then augment or sub-tract from them, and at least discuss or take note ofthem. Such proposals would not be the detailedbriefing for the mission—best in our view left withmanagement. Rather, they would come before thebriefing and would allow the Board to reflect on thetwo or three main issues that the consultation shouldprincipally address.76

34.Such a process would also afford the Boardgreater ownership of the consultation process thatculminates in a Board discussion (something that ap-pears lacking at present), while providing greater as-surance that the process would be focused on what istruly important.77 This recommendation has to beseen also in the context of our recommendations re-garding the role of Board committees in surveillancethat are developed later, in the section below on or-ganization. However, anticipating that discussion, itis worth emphasizing here that this process is aimednot at generating more work or a slower pace ofwork in getting the Fund’s surveillance businessdone, but at helping to streamline work by setting anearlier, better focus on what are the real priorities ineach country’s surveillance.

Less Attempt to Find Optimal Policies in All Areas

35. Not surprisingly, every staff report suggests somemodifications to current policies—for example, asmall tightening in fiscal policy, or an acceleration offinancial market reform. No country is perfect. How-ever, this may lead to two problems. First, it leads tostaff making suggestions for relatively marginal im-provements. But while countries may sometimes findthis helpful, it is unlikely to be the best use of re-sources. Second, it tends to reduce the impact of Fundcriticism when matters are seriously awry. Sinceevery report contains some criticism, and since allcriticisms are to some extent muted by the subtle andspecialized phrasing in which they tend to becouched, it is easier for countries to disregard warn-ings that really do require immediate policy action.

We recommend that staff focus policy adviceon issues of serious or immediate concern anddistinguish such advice clearly from analysisof whether relatively small or judgmental pol-icy shifts would be helpful.

Implementation

36.One point that struck us in comparing our reviewand conclusions with earlier internal reviews is theextent to which the problems—and to some extentthe solutions—that we identify have already beendiscussed. There is, it seems, something of a discon-nect (or at least a very long lag) between the broadpolicy directions outlined by the Board in reviewand policy discussions, and actual practice in day-to-day surveillance work, particularly bilateral consul-tations. Accordingly, we are inclined to think that ifthe periodic reviews of surveillance are in future tolead to more effective surveillance in practice, sub-stantially more attention should be given to the ex-tent to which decisions by the Board affecting sur-veillance have been implemented.

We recommend that in the next internal re-view of surveillance, more attention be givento measuring in some detail (by topic andcountry) the extent to which the specific oper-ational guidance that has been put forwardon behalf of the Board is actually followed inFund consultation reports, and, equally im-portant, if not why not.

37.This will require a less general, more fully docu-mented, approach by the staff. We also wish to em-phasize that a fully adequate discussion of thesematters involves reviewing not only what the Funddid or did not do, but also what kind of response wasforthcoming from the country in question. The Asiancrisis review went some distance in this direction,but was of course limited in country scope.

64

76This recommendation would also address a related problem;perfectly naturally, individual Directors are inclined to raiseparticular issues during a Board discussion that they feel staffmight pursue in future consultations. There is certainly nothingwrong in this. But if all such requests are pursued—and cer-tainly the incentives are for staff to do so—then the net effect isa continual expansion of the agenda. The proposed systemwould help Directors to consider whether such suggestions wereconsistent with a focused approach to surveillance, and with theresources available.

77Note that this issue also comes up in relation to matters oftransparency, in particular “self-censorship,” discussed in the sec-tion of this chapter on communication.

Substance

Substance

Clientism

38.The problem of insufficient frankness is one ofboth substance and language. Fund staff are some-times unwilling to probe deeply into areas where theauthorities are sensitive, although these are alsolikely to be the areas where deeper examination ismost warranted. Alternatively, the staff may diag-nose a problem but present their analysis in ratherindirect language. This allows the staff to claim thatthey have covered the issue. However, they will nothave alerted the international community, and possi-bly not the authorities either, to the significance ofthe problem. The consequences of a lack of direct-ness can be particularly serious in the area of vulner-abilities. For example, the Fund staff have beenmore prepared by and large to tell a member that itsfiscal position, while basically sound, should ideallybe tightened further than they have been to focusclearly on problems in a banking system.39.This problem has long been recognized, yet ithas persisted. It was raised in successive reviews ofthe Mexican and Asian crises. Most staff, includingthose directly responsible for country work, are wellaware of the problem and do try to minimize it.However, in light of the extremely strong incentivesfor the staff to maintain a close relationship withcountry authorities, and the fact that countries tendto react badly to criticism, especially when they havedifficulties, it seems unlikely that further exhortationalone will improve matters much. Nonetheless, itshould be possible to improve the incentives to thestaff for more candid advice.40.Proposals elsewhere in this chapter to focus sur-veillance more closely on identifying and addressingthe major priorities, and in specifically tackling po-tential vulnerabilities, will help. Furthermore, thepublication of staff reports should provide importantsupport over time for greater frankness. With publi-cation of views, the evidence will be more out in theopen and can be judged accordingly.41. Such steps should also help to change perceptionswithin the Fund. However, a view that exists in the in-stitution is that a report that is incisive but offends theauthorities is damaging to a mission chief’s careerwhile one that is bland and later turns out to be lack-ing in some important respect will be overlooked.

We recommend that the Board, management,and senior staff attempt to alter the incentivestructure by making it clear that they will, ifnecessary, back up staff who give frank advice.

42.We recognize that the line between a lack ofdiplomacy in presenting advice and frankness maysometimes be difficult to draw, and that the latter

should not be seen as an excuse for the former. How-ever, given on the one hand the Fund’s express, andwell-founded, wish to offer more candid advice, andon the other hand the built-in incentives to avoid of-fering it, such an affirmation is very unlikely to re-dress clientism bias too far.

Template

43. We do not believe that, in general, the criticismthat the Fund applies an overly rigid “template” or“model” to every country is justified. There is nothingwrong with the fact that the Fund has a frameworkwithin which staff analyze macroeconomic issues; in-deed, it would be worrying if this were not the case. Inbilateral surveillance,78 the staff use mainstreamopen-economy macroeconomic models, and their par-ticular approach makes a point of enforcing consis-tency in the analysis of macroeconomic and financialflows. This can only be to the good. However, therehas been a more limited emphasis on stock variables.This results in weaknesses in the analysis of capitalflows; some specific points relating to capital accountissues are described below. 44. It is appropriate to raise here a related problem.Staff appear in general to be reluctant to give advicethat is country-specific—that is, advice that takesinto account the political and institutional con-straints within which policymakers have to operate.In other words, Fund advice often focuses on identi-fying the first-best general policy. There is nothingwrong with that, and politicians need to hear it in asdirect a form as possible. But when policymakers,quite reasonably, respond that they live in a second-best world, staff are apparently not as good at sug-gesting how the first-best might actually be imple-mented in practice, or at developing and analyzingalternative, specific, policies.79

We recommend that surveillance devote moreattention to policy implementation and to theidentification and analysis of alternative pol-icy options.

65

78In modeling for multilateral surveillance purposes, the Funduses its own multiregion macroeconomic model (MULTIMOD).See Douglas Laxton, Peter Isard, Hamid Faruqee, Eswar Prasad,and Bart Turtelboom, MULTIMOD Mark III: The Core Dynamicand Steady-State Models, IMF Occasional Paper No. 164 (Wash-ington: IMF, 1998). MULTIMOD is generally recognized asbeing on the cutting edge in this field.

79As noted in Chapter II, this is in part a problem of success.Once, it may have been sufficient to focus with many countrypolicymakers on what the first-best was, since they didn’t know.Now, in many countries policymakers know full well what thefirst-best policy is, but would like advice and assistance in identi-fying the best available alternative policy option or modification,even if the Fund will also underline (as it certainly should) that itis indeed only a second-best.


45. This is consistent with our general recommenda-tion that surveillance give much greater emphasis tocross-country experiences.80 The ability to give suchadvice depends to some extent on having staff withsufficient policymaking expertise and experience,which may be difficult to obtain within the Fund;some recommendations in this area are set out below.Another consideration here is the need for the Fundto take account of the differences in economic and in-stitutional structures between different countries.

Capital Account Issues

Policy

46.The Fund has been severely criticized for beingtoo enthusiastic about capital account liberalization.Without entering into this debate at length, we notethat it does appear that the Fund has in the past beenkeen to liberalize as fast as possible, and as a conse-quence has given insufficient attention to the propersequencing of capital account liberalization—in par-ticular, the need for domestic financial sector reformas a precondition for external liberalization. How-ever, policy appears to have shifted significantlysince 1997; while the potential benefits are still rec-ognized,81 much greater attention is now being paidto the risks of liberalization if the financial sector isweak. We believe this more nuanced attitude isbroadly appropriate. However, for it to be properlyimplemented in the context of bilateral surveillancethere will need to be more detailed and sophisticatedanalysis of the financial sector than the Fund has de-veloped so far, a topic we discuss below.

Analysis

47.Distinct from its policy recommendations, theanalysis of capital account issues in surveillance hassuffered from a number of weaknesses. There are anumber of related problems here.

• A traditional Fund method of analysis—finan-cial programming—tends to treat a substantialpart of capital flows as a residual. This leads toa lack of attention to the capital account andthe forces driving its various main compo-nents. However, in recent crises, reversals in

capital flows have been the force behind cur-rent account reversals.

• Concentrating on the current account and itssustainability,82 as much Fund analysis does,can lead to a lack of attention to autonomousforces driving the capital account—and there-fore driving the current account as well.

• A lack of attention to the capital account isalso likely to lead to an inadequate apprecia-tion of the domestic macroeconomic effectsacross countries of shifts in capital flows.

48.These problems arise even though there are staffwho have a detailed knowledge of capital account is-sues and problems; indeed, some Fund economistsare in the forefront of research on these topics. How-ever, they are mainly in the functional departments(especially Research) while area departments takethe lead in bilateral surveillance. Without adequateinterdepartmental cooperation or coordination, theseinternal arrangements have led to inadequate knowl-edge transfer within the Fund. But with the increasedimportance of international capital flows, and in par-ticular the role played by sudden reversals of capitalflows in recent crises, it is clearly essential that bilat-eral surveillance—not just research and multilateralwork—devote as much attention to analysis of capi-tal account issues as it currently does to current ac-count issues.

We recommend that Article IV staff reportsgive greater attention to the forces driving thecapital account, and to capital account issuesin general.

A number of our other recommendations are alsoparticularly relevant in this area:

• The increased attention to vulnerabilities thatwe recommend should focus more attention onthe analysis of the possibility and conse-quences of large capital outflows.

• Improving interdepartmental relations (partly ajob for senior management), and ensuring thatknowledge is transferred between differentparts of the Fund, should help to ensure thatthe expertise that does exist on these issues isdeployed in bilateral surveillance.

Financial Sector

49. In response to the Mexican and Asian crises, andto the clear desire of the international community,the Fund has greatly increased the emphasis given to

66

80To give a concrete example, it would probably be more usefulto a country experiencing large capital inflows to have a Fundanalysis of other countries’policies and experiences, and what thelessons might be for the country in question, than to be told thatcapital account liberalization was a good thing.

81Here it is worth recalling the point made in Chapter IV—thatstressing the importance of proper preconditions and sequencingshould not at all be taken to imply that capital account liberaliza-tion is a bad thing.

82We are not suggesting here that the current account and itssustainability is not an important subject of analysis.

Substance

surveillance of the financial sector. We believe thatthis is appropriate; the close linkage—in both direc-tions—between domestic financial sector crises andexchange rate/balance of payments crises justifiesthe Fund’s involvement. However, we note that im-proving financial sector surveillance was identifiedas an important task as long ago as 1995, and thepace of improvement has been slow. So far, area de-partments have tended to treat this remit as anotheradd-on, which they do not have the time and exper-tise to fulfil in more than a pro forma way, especiallyif countries drag their heels in providing informa-tion. In short, the infrastructure and the push havebeen lacking.50.The proposed Financial Sector Stability Assess-ments (FSSAs) to be undertaken by MAE shouldhelp to remedy this. However, to be successful, theFund will need to be able to draw on the appropriatespecialized staff; we discuss this in more detailbelow. Moreover, it will be important to ensure thatthe work done by MAE is properly integrated intothe Article IV process.

Vulnerabilities

51.At the moment, Fund documents tend to addsome general remarks about “downside risks” to anelaborate central scenario for a country’s economicprospects over the next year or two. However,notwithstanding the analytical and procedural com-plications that are evidently involved in going fur-ther, it is important for the Fund to strive to bring theanalysis of vulnerabilities more to center stage.

We recommend that surveillance devote sub-stantially more attention to vulnerabilities.

52. In the following, we sketch how this added focuson vulnerabilities might be brought about. Some fur-ther implications for the Fund’s work are discussedin the sections of this chapter.

We envisage that this process would proceed inthree stages:

• identification of vulnerabilities prior to ArticleIV missions, in part with the aid of early warn-ing systems (EWS);

• information gathering from the private sector;and

• presentation of vulnerabilities analysis to na-tional authorities.

Identification of Vulnerabilities

53.The starting point for a Fund team going on anArticle IV mission would be to examine, togetherwith functional departments (notably RES andMAE), potential vulnerabilities of the country in

question. As part of this process, the team would ex-plore the latest position of the country in relation tothe set of indicators used in the Fund’s EWS.83 Sucha discussion would become an integral part of the re-view of the mission brief, but the subject would lenditself better to a meeting of the team with staff fromRES and other functional departments than to writ-ten exchanges. 54.Although it remains very difficult to predict thetiming of a crisis, it is not a random event. Giventhe high stakes—and the poor record of both theFund and outside observers to date—even a modestability to predict crises would be very valuable.84

In any event, one important advantage of the EWSapproach for bilateral surveillance is its quasi-ob-jective nature; a poor score will force area depart-ments and, we hope, country authorities to focustheir attention on problem areas even if they wouldprefer to avoid the topic. So, even if the indicationsfrom EWS are no better than the staff ’s informedjudgment (and we would not normally expect thatthey would be), they may have a role in combatingpressures for “clientism.” Of course, this additionaltool should not rule out staff using their own judg-ment and analysis to identify what they consider tobe the country’s principal vulnerabilities. Further-more, the application of country-specific analysiscan, in turn, help to improve the construction ofEWS.

Meetings with the Private Sector

55. To follow up this part of its preparation, partici-pants in the mission would make a point of searchingout information from the private financial sector—through analysis of market commentary or interviewsalong the lines already conducted by a mission fromthe capital market divisions in RES when it visits aglobal or regional financial center before preparingthe ICMR. Such information may, for example, relateto the buildup of large open positions in the currencyconcerned, or to the interpretation of recent move-ments in short-term indicators additional to those usedin EWS, notably bond spreads.

67

83We explain the main elements in EWS and comment on itsuse, primarily in multilateral surveillance in Box III.1, while fo-cusing here on its application to the Article IV process.

84One observer has likened the difficulties of forecasting bal-ance of payments crises to those of predicting earthquakes. SeeBarry Eichengreen, Toward a New International Financial Archi-tecture (Washington: Institute for International Economics,1999). We find this analogy excessively unfavorable to EWS fortwo reasons: first, earthquakes are even more difficult to predictthan financial crises; and second, their occurrence cannot be in-fluenced by human action—only their consequences can be miti-gated—while balance of payments crises can, in principle, beforestalled by timely warnings.


Presentations to Country Authorities

56.The mission would take up the issue of vulnera-bilities against its central scenario for the country’sprospects. It would not only assess the risks but alsocomment upon the feasible policy responses.57.We are aware that not much reaction can be an-ticipated from many national representatives. Theteam’s interviews indicated that officials, while rec-ognizing the merits of the exercise, were wary of en-gaging in a dialogue about such matters. The knowl-edge that the consultation report was going to bepublished would presumably reinforce such wari-ness. However, we nonetheless believe that even justpresenting this material on a regular basis is valuableand important for the surveillance process.58. With time, actual discussion of vulnerabilities (asopposed to the consultation presentation itself) shouldbecome a standard part of Article IV consultations.Such an outcome will be helped along by two paralleldevelopments. One is the greater familiarity with theanalysis of vulnerabilities and EWS that will developas past crises are dissected in the multilateral publica-tions. The other development is the introduction of theContingent Credit Line. This would require the analy-sis of vulnerabilities and the risks of contagion to takecenter stage in some bilateral consultations—and tohave the discussions reported to the Board in caseswhere a country wanted to qualify for a contingent fa-cility. We return to the implications of this in the sec-tion on recent developments below.


59. While there are competitors, the comprehensive-ness of the Fund analysis tends to make WEO andICMR indispensable to anyone interested in an objec-tive and detailed perspective on the global economyand the increasing interdependence among its compo-nents. We reported in Chapter II that some officialsand academics saw value in more regular publicationof the analysis, particularly the forecasts, contained inWEO. We also note that in December of both 1997and 1998, a combined “interim” WEO/ICMR wasproduced. We suggest that the Fund take a further stepin this direction.

We recommend quarterly publication of theWEO forecast.

60.This is not intended to imply that the Fundshould produce much more material than it does atpresent; simply that rather more should be pub-lished. In terms of forecasts, descriptive material,and short-term policy analysis, it is clear that most ofthe necessary material already exists internally, andindeed is presented to the Board at the regularWEMD sessions (and the quarterly private sector fi-

nancing notes). We see no reason not to publish thismaterial with greater frequency.85

61.As regards the longer and more systemic back-ground studies that currently form part of WEO andthe ICMR, these could be reserved for a larger pub-lication—either WEO and ICMR separately or withthe two documents combined, such as the one thatappears at the time of the Annual Meetings—orthey could be incorporated quarterly as and whenavailable. Again, we are not suggesting any in-crease in the amount of material produced by theFund. 62.We also believe that this change in publicationshould be accompanied by two changes in the ac-countability for the publications that wouldstrengthen their objectivity.63.There are presently two problems for the objec-tivity, and hence the authority, of the multilateralpublications. On the one hand, the forecasts are notclearly the responsibility of the main authors in RES.On the other, members of the Executive Board attimes appear to lean more heavily on the staff tomodify judgments of policies in their respectivecountries than is healthy for the long-run reputationof WEO and the ICMR.64.As regards the former point, we have noted inChapter III that forecasting in the Fund is to a largeextent a “bottom-up” rather than a “top-down” ex-ercise, that is, global forecasts are built up from thecountry analysis of area departments. Although aconsistency check is obviously provided by RES,we suspect that this process is likely to result in ageneral bias toward optimism (although, clearly,there are no doubt occasions when area departmentforecasts are more accurate than RES ones). 65. Furthermore, this leads to the serious problem thatno one is willing to accept ownership of the WEOforecast. This was evident from our interviews withstaff; RES staff disclaimed responsibility for forecast-ing errors in the published WEO—although the rest ofthe world is given to understand that it is their prod-uct—saying that the forecasts were really the respon-sibility of area departments. Area departments, bycontrast, regard the forecasts as the collective respon-sibility of the staff. We do not regard this situation asappropriate.

We recommend that ultimate responsibility forthe WEO forecast be vested clearly in RES.

68

85The current WEO forecast schedule might not fit too wellwith a move to quarterly publication, since at present WEO ap-pears in late April and late September, leaving a gap of only fivemonths. Either this timetable could be adjusted slightly, or thenew summer WEO could be particularly lean—perhaps simplycontaining forecasts and descriptive material.

Organization

66.Of course, RES will continue to rely to a verygreat extent on the work done, and the materialprovided, by area departments; we are not suggest-ing a shift in resources here. And where there aredisagreements, it will doubtless wish to take theirviews into account. However, in the end, if RES isto be accountable for the quality of the forecast ex-ercise, it should be clearly responsible for makingthe final decision on contentious issues.86 Timewill tell if it is correct.67.The second issue in regard to accountability—and candor—is the nature of the Board’s involve-ment before publication. While the Board has the in-tention of letting WEO and the ICMR be issued onthe staff ’s responsibility, it is evident that Board re-view can become Board pressure for change to pro-tect individual country interests. This obviously re-duces the overall value of what is published.

We recommend that the Board make it clearthat the presumption is that the staff draftshould be published as it stands.87

68. Indeed, simply as a matter of practicality, the re-view and Board clearance process would need to bestreamlined considerably if the WEO forecast wereto be produced quarterly.88

Organization

69.This section deals with the organization of sur-veillance under three main heads: first, the organi-zation of surveillance relations with countries, in-cluding the organization of missions; second,internal aspects; and third, the role of the ExecutiveBoard.

Country Surveillance

Greater Flexibility in Monitoring

70. Members vary a great deal in their characteristics,and more selectivity in how surveillance is applied istherefore appropriate. It may be contended that thereare risks involved(it would have been easy to argue,for example, that Korea in 1995, a successful countrywith what the Fund and many others believed werelargely sound policies, required relatively little atten-tion.89 However, our broad judgment is that signifi-cantly fewer staff resources could be devoted to in-dustrialized countries overall without any loss ineffectiveness or impact. This reflects these countries’risk characteristics, and the greater public availabilityboth of data and economic analysis. In large part,Fund efforts here are duplicative of other work per-formed in both the public and private sectors.

We recommend the following:

• For all industrial countries but the verylargest, full Article IV consultations shouldbe less frequent than annually.90

• For the very largest industrial economies,91

in light of their systemic impact, annual con-sultations are still called for. However, to im-prove the payoff, surveillance should focusmore on the international implications ofthese countries’domestic policies and corre-spondingly less on advice regarding the poli-cies themselves.

• The particular issues in surveillance of theeuro area are discussed in more detail in Box2.1. But regardless of exactly how surveil-lance is organized, there is a strong case forcutting back the resources allocated to theeuro area, which are now more than fourtimes those devoted to the United States.

71. The changes recommended for industrial coun-tries overall should yield significant savings in re-sources. Some could be applied to lessening theproblems of overloaded surveillance agendas that arenoted below. However, it should be possible also totransfer resources to the other areas that we identifyas priorities: the international dimension and harness-

69

86One potential objection to this recommendation is that itwould simply reverse the problem; area departments would feela loss of ownership of the Fund’s forecasts for the countries forwhich they are responsible. While this might occasionallycause difficulties on the rare occasions where there were sharpdisagreements (although we do not see any great problem ingeneral in being quite transparent about divergences of viewand the reasons for them), we do not see that it would under-mine the conduct of bilateral surveillance to the extent to whichthe present situation undermines the credibility of the WEOforecast.

87Of course, the Board would still discuss the draft in advanceof publication (and WEO could record the views in an Appendix,as was done with the May 1999 issue). But the staff should decidehow to deal with the points raised at the discussion.

88WEO is currently discussed by the Board four to fiveweeks before publication, and Board members require threeweeks to review it before the Board discussion. As a conse-quence, while it only takes about two months for the staff toproduce a draft WEO document—even with the current elabo-rate internal review procedures—it takes another two monthsbefore publication can proceed, with relatively little valueadded.

89It is also worth bearing in mind that the deterioration inKorea’s circumstances in 1996/97 was in any event not signaledby current, nonselective surveillance procedures.

90This would not imply that staff would only visit countriesevery two years; we would expect short, smaller missions to visitmore frequently than this. See also our recommendations on morecontinuous surveillance below.

91The United States, the euro area (i.e., consultations focusingon the area as a whole), and Japan.


ing cross-country experience.92 In saying this, we arenot arguing for any particular allocation of resourcesbetween area departments on the one hand and func-tional departments on the other (although, clearly,there will need to be some reallocation across areadepartments). But we do see it as important that theresources within area departments—again, taken as agroup—should in important measure be refocused todeal with these priorities.

Small States

72.Small states93 clearly have concerns that theireconomic policy problems are qualitatively differ-ent, and that the Fund fails to take sufficient accountof this. Moreover, the impact of surveillance is prob-ably greatest for small states (see also our discussionof technical assistance, below). It seems appropriateto try to respond to these concerns. 73. While we did analyze the possibility of setting upa special coordinating division to take the lead indealing with such members, we abandoned this ideaas being too complicated to implement unless onealso shifted small states out of their respective areadepartments. However, what we do propose is that theFund establish a project to analyze the economic ef-fects of small country size from the macroeconomicmanagement point of view—to see whether their situ-ations are indeed substantively different in ways thatare relevant to the Fund’s surveillance competencies.Then, surveillance might be reviewed in this light.74. Of course, the basic challenge still remains: howshould the level of surveillance resources devoted tosmall states be decided in the light of two conflictingconsiderations—that these states are numerous and at-tach great importance to surveillance, but their shareof the total world economy is very small. Clearly, asubstantial reduction in these resources would notpose any systemic threat to the world economy, whilefreeing up resources for use elsewhere. To the extentthat surveillance should focus on issues of concern tothe international financial system, such a reductionappears to be justified. However, the negative impacton individual countries might be significant, amount-ing as it would to an effective reduction in develop-ment aid or technical assistance. This trade-off is es-sentially a political question; but it should berecognized as one that essentially is not about the de-ployment of resources within surveillance, but ratherconcerns choices between surveillance and aid. Assuch, it is largely outside our remit.

More Continuous Surveillance

75. More continuous surveillance has for some timebeen a Fund objective, and a range of measures suchas midyear visits and Board informal country andWEMD sessions have now become a regular feature,along with the long-standing practice of ad hoc bilat-eral meetings at the time of the Interim Committeeand Annual Meetings. These features are further aug-mented by the work of functional departments—inparticular the visits of the capital markets groups fromRES, whether to countries or to capital market centers. 76.Can the process be improved further? We think itcan. While missions themselves apply modern tech-nology to communicate with headquarters, the sur-veillance cycle and process as a whole has not fullyadapted to it. The process appears very much aswhen it was designed in the 1960s, despite the in-crease in the quality and availability of telecommu-nications and the ease of transmission of data and re-search on the Internet. That is to say, in the normalcourse of events the episodic nature of relativelylarge-scale physical visits still dominates—with agear-up and wind-down, and a relative lessening offocus as staff turn to other endeavors. But while mis-sions are no doubt an essential way of taking stockfrom time to time, they should not be allowed toovershadow these other means of communicationand information gathering. The staff need to developsuch channels with their country contacts. Thismight mean taking a leaf out of the book of the waythese things are done in the private sector, whichconsciously maximizes electronic contact and mini-mizes the length of physical visits. We should em-phasize that we are not trying to devalue actual visitsas part of continuous surveillance. Rather, we aresuggesting shorter, more targeted, and perhaps (butnot necessarily) more frequent visits.77.Our discussions with country representativescertainly indicated some receptiveness to updatingthe process in this way, with consequential savingsin the time spent by missions and by their hosts. Ac-cordingly, we think that the Fund should actively ex-periment with receptive members in enhancing therole of telecommunications in data gathering and inother areas to see what improvements and efficien-cies in surveillance can be gained—as a complementto or substitute for actual visits. In the same way, ex-perimentation should also be undertaken in makingconsultation missions shorter by trying to increasethe amount of work that can be done in advance.

We recommend a more conscious focus on theuse of electronic means of communication,initially on an experimental basis, with a view to maintaining close contact with policy offi -cials and to reducing the length of consulta-tion missions.

70

92We note that the Fund is already increasing the resources de-voted to financial sector surveillance.

93Without aiming at any precise definition of “small states,” wecan note that there are some 40 Fund members with populationsof less than one million.

Organization

Technical Assistance

78.While Fund technical assistance as such is outsideour remit, surveillance is often to a greater or lesserdegree technical assistance. This is particularly thecase for small states. However, as we observed and asnoted in the 1999 internal review of technical assis-tance, surveillance and technical assistance are not al-ways as well coordinated as they might be.

We endorse the recommendations of the inter-nal review in this regard—in particular, thatarea departments should, through the consul-tation process, seek to integrate the supportprovided to members through technical assis-tance and surveillance activities.

External Contacts

79.Fund missions have been expanding their rangeof contacts outside the traditional circle centered onthe central bank and the ministry of finance. But thisexpansion does not appear to have been systematic.And while it is understandable that staff would tendto meet people who they feel speak the same lan-guage—academics and financial market participantsrather than trade unionists, nongovernmental organi-zations, or others who are often critical of theFund—this tendency should not be allowed to domi-nate completely.80.Although decisions will still need to be taken ona case-by-case basis, we believe there is a role forcentral guidance to encourage breadth of contact.

We recommend that the External RelationsDepartment monitor and assess Fund prac-tices and experiences to date with a view togiving guidance with the Fund’s explicit back-ing regarding external contacts.

81.The purpose of such contacts is twofold; both toexplain to such bodies what the Fund is and what itdoes, and to allow those representatives to give theirviews in areas relevant to the Fund. Meetings shouldbe focused on the surveillance agenda, with the staffmaking very clear what in that context the Fund doesand does not do, and why.94 We have already madeproposals earlier in this chapter for tightening thisfocus in the Fund’s own work in the interests ofmore effective surveillance. Furthermore, given theinterest in shorter missions where feasible, such con-tacts also would have to be well focused. However,we think that an allocation of a small part of a mis-sion’s time for such endeavors is worthwhile.

82. In the same vein, the Board may wish to con-sider inviting from time to time informed outside ob-servers of the Fund to meet with the Board.83.On a related topic, in some countries we were in-formed that representatives of the government, usu-ally the central bank or the department of finance,made a point of accompanying a mission to all meet-ings, whether with government or private sector rep-resentatives. This may be for administrative conve-nience. Nevertheless, it can in some cases beinappropriate for effective surveillance. However,missions are understandably reluctant to tell govern-ments so.

We recommend that the Fund make clearwhere necessary that meetings with nongovern-mental representatives should take place with-out the presence of government officials.

Relations with World Bank and Others

84.We found that coordination between the Fundand the World Bank remained uncertain, particularlyin the financial sector area, where the greatest over-laps exist. We are aware that this topic has beenstudied numerous times, and we do not have any fur-ther suggestions to make. However, as is discussedfurther below, we believe that the Fund should makemore effort to make use of work done by others, par-ticularly in areas where it does not have expertise.

Internal

85. In earlier chapters, particularly that on conductand methods, we described a range of surveillancechallenges internal to the institution that emergedfrom our interviews. These are addressed further here.

Skills Mix

86. In light of the evolving demands of surveillance,in particular the emphasis on more specialized andmore sophisticated policy skills, more diversitywould seem desirable in three principal areas.

• More financial sector expertise. If the Fund isto fulfill the new demands placed on it in thisarea, it clearly needs more staff with authorita-tive experience in the financial sector. Thismeans both people with regulatory and super-visory experience and people with private fi-nancial sector experience.

• More policy expertise. Policymaking experi-ence appears to be particularly helpful in sur-veillance, and this suggests that a sustained ef-fort should be made to encourage it throughrecruitment—as well as through secondment or

71

94We believe that this can also help to correct a general and sig-nificant lack of understanding of what the Fund is and what it cando.


interchange programs—with relevant policy-making institutions in member countries. Thismay also mean giving greater weight to poli-cymaking expertise in deciding the level atwhich individual staff are recruited to theFund.

• More outside experience in general. The dan-gers, given the large proportion of staff whospend virtually all or most of their careers inthe Fund, are insularity, conformity, and lackof hands-on experience. This suggests greateruse of external consultants, interchange withmember governments, academia, and privatesector institutions, and so on.

87.The Fund is aware of its needs in the first area—financial sector expertise. However, more of a con-scious program may be needed in regard to the othertwo.

We recommend consideration of the followingspecific policy actions:

• More emphasis on policy experience, andtherefore less on academic credentials, at allrecruitment levels, including the EconomistProgram; and

• Fund staff should be positively encouragedto spend one or more assignments outside theFund before reaching management grades.Current programs in this area should be ex-panded, if that proves necessary to ensureadequate opportunities.95

Accountability

88.The Fund has a strong sense of hierarchy, andone downside of this is a lack of individual account-ability for the quality of the output. Surveillance ma-terial that is submitted to the Board is regarded asthe collective output of the Fund, with the full en-dorsement of management. As a result, responsibil-ity is pushed upward, to management or to the Fundcollectively.96 While it is clearly helpful to the cohe-sion of the institution for there to be a sense of col-lective responsibility, it is damaging to internal ac-countability and incentives if junior or midlevel staffdo not feel individual “ownership” of Fund policies

or outputs to which they contribute directly. Wethink that there could be more individual account-ability, with the associated good incentives, withoutdamaging cohesion.

We make the following specific recommenda-tions:

• There should be as much accountability aspossible for papers, staff reports, and the as-sociated policy recommendations. In particu-lar, the staff member most directly responsi-ble for authorship of a paper or staff report,or with greatest knowledge of the countrybeing discussed, should be the main presen-ter at any Board discussion.

• In the surveillance context, it should gener-ally be the case that one staff member—nor-mally the Division Chief in the area depart-ment—has overall responsibility foroperational dealings, including leading mis-sions, with an individual member state, andshould be held accountable for them.

89. The Fund is not alone in facing the problem of en-suring that staff are rewarded according to the qualityof their output. We do not underestimate the difficultyof this task, which is particularly marked in publicsector organizations like the Fund, where pay differ-entials are smaller and separations rarer than in theprivate sector. However, at the moment the percep-tion—both inside and outside the Fund—is that suc-cess and failure, certainly in the surveillance context,do not translate into career prospects as directly asthey might. We are aware that the Fund administrationis conscious of these difficulties and has taken somesteps to mitigate them. We are also aware that the con-centration of elevated ratings under the performanceevaluation system is not a totally fair reflection of theextent of differentiation in performance recognitionthat does go on. Nevertheless, there still remains aproblem in sharpening incentives.

Interdepartmental Relations and the Review Process

90.We view communication among departments asbeing relatively poor. A lack of collegiality betweendepartments inhibits learning from others’experi-ences and, more generally, the transfer of knowledgewithin the institution that is so important for its ef-fective functioning. As a result, the Fund sometimesappears to be less than the sum of its—often veryimpressive—parts.91.This is particularly apparent in regard to the in-terdepartmental review process through which allsurveillance documents must pass. We certainly donot dispute the necessity for a review process; it is

72

95It has been suggested to us that the role of resident represen-tative offers this kind of experience. We agree that it is differentfrom being in Washington, but we are not convinced that the ex-perience overall is sufficiently different.

96The same process takes place within departments; material—even of a relatively mundane nature—is not generally circulatedto other departments until it has been collated, reviewed, and ap-proved by departmental front offices. This issue is also addressedbelow.

Organization

clearly essential both for quality control and forknowledge transfer. But, as currently constituted, itis extremely cumbersome, being time-consumingand generating a considerable amount of paper, andperhaps excessively adversarial on occasion. And asa result, as was very clear from our discussions withstaff, it is often frustrating for the participants, inboth area departments and functional departments,and less constructive than it might be. We think thatthe process could be streamlined, and the number ofparticipants considerably reduced, while maintain-ing and indeed increasing the value of the exercise.We think that this could also help knowledge trans-fer, with less focus on checking and more on addingideas.92.While we refrain from making formal recom-mendations on a detailed organizational matter ofthis nature, we do have a number of specific sugges-tions that the Fund might wish to take into accountin any examination of its internal processes.

• At present all drafts and comments passthrough front offices. As well as reducing ac-countability and adding an extra level of re-view, this also takes time away from the otherresponsibilities of front office staff, discussedbelow. It is not clear why drafts could not becirculated by their principal authors directly toother departments, with the internal depart-mental review taking place simultaneously.

• There could be greater use of informal com-munication rather than written memoranda.The issue behind many written commentscould be resolved more efficiently and less ad-versarily by phone or e-mail. It would beworth experimenting with meetings and themore flexible exchange, even debate, that willthen occur—short, focused, and only with thenecessary participants—instead of paper mes-sages.

• As noted above, the individuals primarily re-sponsible for reports should be accountable fortheir quality. It should therefore be their re-sponsibility to solicit comments from other de-partments and to decide how to take account ofthem. Other departments should normally re-strict their comments to their areas of responsi-bility, and avoid drafting comments.

93. In this context, it is worth mentioning the role ofPDR, which provides an important central monitor-ing and quality control function, and whose signing-off role should therefore be retained; although thisdoes not imply that PDR, like other departments,could not reduce the resources devoted to theprocess. It is in large part PDR’s responsibility to en-sure that general policies are implemented, and that

staff reports do indeed address the key issues, focuson international aspects, and are frank and direct.Correspondingly, our proposals to limit the scope ofconsultations should help to reduce the burden ofchecking off that all consultation bases are touched.

Front Office

94.The Fund departmental management structureappears to be rather top-heavy; in particular, the cur-rent number of staff in departmental front offices ap-pears large compared to other organizations, publicand private sector. And, as noted above, front officestaff seem to spend a disproportionate amount oftime on the interdepartmental review process—bothin approving the drafts of their own departments,and preparing and collating comments on other de-partments’ drafts—an activity which could bestreamlined considerably.95. If, as we see it, many in the front offices of areadepartments act in effect as senior division chiefs, itis appropriate to see whether they should be movedinto divisions, with a corresponding reduction in thesize of the front office. This would help internal ac-countability. Another concern that we were left withis that while an important role of the front office ingeneral should be to take an overall view—particu-larly as regards knowledge transfer—we are not atall convinced that this key role is being filled underpresent arrangements, where keeping up with thearea’s own country surveillance agenda dominateseveryone’s time and thinking.

We recommend that front offices in area de-partments be made clearly accountable for en-suring that bilateral surveillance incorporatecross-country and multilateral perspectives.

96.This will involve keeping up with relevantanalysis and research both in other Fund depart-ments and outside.

Overwork

97.There is no doubt that many Fund staff arechronically overworked. While the willingness ofstaff to work long hours is commendable, and con-tributes to the Fund’s impressive capacity to respondquickly to crises, this inevitably reduces the generaleffectiveness of surveillance, as well as that of otheractivities. In particular, it is conducive to “tunnel vi-sion”—the inability to look outside, take the longview, or examine alternatives—simply because thepressure of deadlines in getting through the regularsurveillance calendar is in many instances so great. 98.On a related topic, there is a need for more andhigher-quality research assistant level staff to relievejunior professional staff of low-quality work. This is

73


being pursued, but we note that at the present pace itwill be some time before the professional/researchassistant ratio is comparable to that of other similarinstitutions. 99.The greater focus and prioritization in surveil-lance that we recommend, together with a number ofthe other recommendations, could help to relieve thepressure on staff. However, for this relief to be real-ized, the organization as a whole needs also to prac-tice constant vigilance and restraint so as to avoidgetting drawn into interesting but lower value-addedareas.

We recommend that the overall volume of sur-veillance work be reduced relative to the num-ber of personnel engaged in it.

General Organizational Issues

100. The Fund faces a rapidly changing external envi-ronment. This constantly generates new demands andchallenges, both for surveillance and for the Fund’sother functions. With the current debate over reformof the international financial architecture, this is thecase now more than ever. In this context, we think it isworth examining whether the internal workings of theFund—its organizational structure, how it is man-aged, the recruitment, composition, and skills of itsstaff, and its internal functioning and procedures—arewell adapted to these new challenges. 101. The challenge of rapid change is not unique tothe Fund. A number of the issues discussed above, inparticular those related to accountability, incentives,hierarchy, and management structures, are ones thatare faced by most large public sector institutions,and that have been sharpened by recent changes inthe role of the public sector. These issues are particu-larly difficult in “policymaking” institutions like theFund, and we do not wish to minimize the complexi-ties here, or suggest that management and seniorstaff are not seized of the issues. However, we donote that many policymaking institutions—as wellas large organizations in the private sector—in theFund’s member countries have undertaken funda-mental management and organizational changes inthe past 10 years; the example of New Zealandsprings immediately to mind, but many others havegone at least part way down similar roads. 102. The Fund has not in recent decades undertakena fundamental review of its organizational and man-agement structure in the light of its objectives.However, we are not aware that the Fund is so dif-ferent from other policymaking institutions thathave benefited from such reviews. Given thechanges taking place in the international financialarchitecture—and the Fund’s evident integral role—current circumstances might provide an opportunityfor such an examination.

External Relations and Relationships and Review

103. This is a broad area, and we limit ourselves tosome observations directly related to surveillance.104. The Fund has moved a considerable distance inthe amount it publishes on surveillance (as well ason other subjects). However, the basic point we wantto underline here is that it is desirable for the Fundnot only to communicate the results of its work andthinking to others, but also to take pains to show thatit listens to and learns from others. 105. In this regard, we were struck by the fact thatstaff reports typically do not contain references ei-ther to academic literature or, perhaps more tellingly,to the output of other organizations, or to the finan-cial press.97

We recommend that the staff:• systematize and organize their use of outsideinformation with a view to applying it in sur-veillance; and

• refer to and/or summarize work produced byother organizations where relevant.

106. Finally, we note that while the Fund should ofcourse continue with its internal reviews of surveil-lance, and while publication of consultation reportsshould help in terms of accountability and maintain-ing or improving quality, we also believe that a sys-tematic external review process (as distinct from ex-ternal evaluation) would also help to providecontinuous upward pressure on quality.98 For exam-ple, external review of some proportion of individualstaff reports by invited outside reviewers (normallyacademics with country-specific knowledge or pos-sibly former officials) would be relatively inexpen-sive, but could provide a salutary check on the inter-nal process and conventions.

We recommend that the Fund experiment withexternal review of a sample of staff reports.

Executive Board

107. In the earlier chapter on conduct and methods,we summarized our discussions regarding the role ofthe Board—pointing to a number of problems thathad been registered. Some of those problems are

74

97One staff member pointed out that the 1996 staff report onKorea was produced just after an Economistsurvey, which had adetailed description of the financial problems of Korea’s corpo-rate sector. However, the staff report did not cover this topic, andmade no reference to the Economistsurvey.

98The external evaluation process, of which this evaluation is apart, can only look at issues in a one-off fashion. It is not the samething as an ongoing process of external review and evaluation ofthe day-to-day surveillance output of the Fund.

Communication

congenital, reflecting the way that the internationalcommunity has designed the basic arrangements forFund governance. Nevertheless, everything weknow about institutional governance indicates to usthat a group of 24 is, to put it mildly, extremely largefor useful exchanges of views, discussion, and groupdecision making. Accordingly, the Board starts witha major impediment. This impediment is particularlystriking in the area of surveillance, given the hugeworkload and given the crucial role that the Board issupposed to play in the Article IV consultationprocess.108. If a regular, functioning, Board committeestructure were easy to set up, it surely would havebeen by now. The glaring absence of meaningfulcommittee work speaks volumes for the constraintsunder which Directors apparently operate, de facto ifnot legally, as country and constituency representa-tives. That may well mean that in tackling the ques-tion of the Board’s role in surveillance we are look-ing at symptoms rather than causes. However, on theassumption that there is a general desire to have theBoard operate more effectively in the process, andthat a look from the outside is worth something inthis regard, we strongly urge that serious considera-tion be given to instituting a committee structure thatspecifically aims to make the Board’s contributionmore effective.99

We recommend that all Article IV staff reportsbe discussed in the first instance in a commit-tee rather than by the full Board.

109. We should emphasize that this does not meanthat the full Board would not get a chance to discussindividual countries; nor do we mean that each staffreport should be discussed in depth twice, thus in-creasing rather than reducing the workload. Rather,those discussions would take place on the basis of acommittee report that provided focus and, wherepossible, grouped countries in the same region orfacing similar issues. Indeed, by allowing a commit-tee to set the agenda for a focused discussion of thefull Board, our proposal should strengthen theBoard’s involvement, and perception of ownership,in the surveillance process. Consistent with our gen-eral recommendations, the full Board should aim tofocus its discussions on the international environ-ment and cross-country issues, as well as on themain domestic issues.

110. We believe that, if implemented effectively,such a system could both improve the Board’s contri-bution to the surveillance process—and consequentlythe process as a whole—and reduce the time spent byindividual Directors discussing surveillance material.While there may be some additional work for staff asa result, we think this would be a worthwhile trade-off. It will, however, require the Board—collectivelyand individually—to adopt the working practicesnecessary to make a committee structure function ef-ficiently. Without this determination, a committeesystem would not improve matters. 111. An illustrative committee structure is describedin Box 5.1. Clearly, other structures would be possi-ble. However, if significant benefits are to be real-ized, a radical change in the Board’s working proce-dures is necessary.

Communication

Publication of Article IV Reports

112. Transparency, whether to enhance accountabil-ity, to help markets function better, or to increase theeffectiveness of surveillance, now appears well es-tablished as an internal principle in the Fund. Thedifficult issues in regard to surveillance have comenot in relation to this principle, but rather in agreeingon how far or in which direction transparency can beextended without in some respect compromising theeffectiveness of surveillance.113. On this latter score, the chief reservations againstextending transparency through publication of the Ar-ticle IV consultation reports have been twofold.

• The Fund’s role as policy adviser would beweakened severely by publicity. Governmentswill be unwilling to engage in frank discus-sions about policy options if the results ofthose discussions are to be made public.

• Publication of policy analysis and views couldprecipitate exactly the sort of crisis that it wasmeant to avert. Indeed, this might occur even ifthe analysis is actually incorrect or merelyoverly pessimistic.

114. However, in weighing the benefits and costs ofpublication from the viewpoint of surveillance, theteam found these costs to be distinctly less onerousthan they might appear in general.115. Our main reason for discounting the damagethat might result from publication is the fact that Ar-ticle IV reports are not, in the event, kept very confi-dential. When one considers the worldwide distribu-tion that such reports have to receive, it would besurprising indeed if they were not fairly readily ac-cessible to a determined, informed, interested party.

75

99Agreement on two or three lead speakers to focus each dis-cussion at Board surveillance meetings, as they are presently con-stituted, would be an improvement over current arrangements.However, we believe that the gains in terms of a more activeBoard contribution to Fund surveillance activities would be muchgreater through a committee structure.


Debt rating agencies may obtain them directly froma country being rated. Others may get hold of themthrough third parties. Our discussions with a widerange of interviewees fully confirmed this view.116. To be more precise, it is a sound assumption thatthe major international market participants alreadyhave, in various ways, fairly ready access to these re-ports. Furthermore, to the extent that they find themuseful to act upon, they will do so. It then followsthat any adverse market impact from publicationwould be quite modest. Note, however, that the mod-est nature of the impact applies both ways—not only

in mitigating the adverse effect of actually precipitat-ing a crisis through publication, but also in lesseningany favorable effects through improved market infor-mation and, presumably, market discipline.117. There is a second consideration as to why thecosts of publication would be low. The fact that Arti -cle IV reports are very likely to have a significantcircle of unauthorized readers also appears to affectwhat is put in them. Again, our wide-ranging discus-sions led us to the view that either governments donot engage in as frank and as confidential discus-sions as many suppose, or, alternatively, that mis-

76

Box 5.1. Possible Board Committee Structure for Surveillance

We propose that most surveillance work be conductedthrough committees. In particular, the first considerationof all Article IV staff reports should be at the committeelevel. We suggest that these committees be organized on aregional basis. This would have the additional advantagethat the committees could have an input into regional sur-veillance as well as into the bilateral Article IV process.

The main points of such a structure might be as follows:

• There would be one Executive Board committee foreach area department, so a total of six committees.

• Each committee would have eight members. Ideally,four of these would be from within the relevant geo-graphic region, and four from outside.1 Each Execu-tive Director would thus be on two committees; hisor her own regional committee, and one other:2

• The committees would consider Article IV staff re-ports. In addition, they could commission and con-sider additional analysis from the area departmenton regional issues.

• The committees would, on the basis of proposalsfrom staff, agree on the two or three key issues onwhich upcoming Article IV consultations wouldfocus (see the section of this chapter on objectivesand priorities).3

• Nonmembers of the committee could attend as ob-servers, but normally would not speak. They couldsubmit written views if they wished, but this wouldnot be encouraged.4

• The committees would propose to the full Boarddraft Article IV conclusions, normally on a lapse-of-time basis. These conclusions would onlybe discussed by the full Board if there were a significant difference of view within the com-mittee or if there were significant objections from nonmembers of the committee. Any Boarddiscussion would focus on these areas of disagreement.

• Every quarter, the full Board would discuss eachregion, on the basis of a report from the regionalcommittee. This report would discuss develop-ments in the region and would highlight major re-gional issues and potential problem countries. Re-ports, which would be the responsibility of thecommittee, would be drafted by staff from Direc-tors’ offices, with input from area department staffas required.

The advantages of this procedure would be the following:

• A substantial reduction in the amount of time spentby Executive Directors on Article IV staff reports.Each Executive Director would be on two commit-tees out of six, and would thus consider approxi-mately one-third of all staff reports in the firstinstance.

• More constructive discussion on reports, since theywould be discussed in a committee of 8 rather thana Board of 24.

• More focused discussion in the full Board, whichwould only discuss important issues, problem coun-tries, and issues where there was a significant dis-agreement among Directors.

• Greater attention to regional surveillance and cross-country issues.

• An opportunity for the Board, through the commit-tee process, to have a more systematic input intothe issues considered by staff in bilateral and re-gional surveillance and into the agenda of the fullBoard.

1Since area departments do not each contain four Boardconstituencies, this will have to be adjusted in some cases.

2Chairs would rotate and be from outside the region. Alter-natively, it would be worth considering having Deputy Man-aging Directors in the chair, if this were thought to improvethe process.

3In practice, staff could prepare a brief report to the com-mittee each quarter, stating that Article IV consultations werescheduled and what the main focus of discussions would befor each.

4This is a small but important point. If all Directors feelthey must participate in all committees, little increase in focuswill result, while the demands on the time of Directors andstaff will increase.

Communication

sions (quite possibly with the tacit agreement of offi -cials) filter those discussions to minimize potentialembarrassment in the written report.100

118. Clearly, this should affect one’s view as to theimportance of the first concern with publication thatwas mentioned—namely, that the frankness of theconsultations would suffer. Equally clearly, anyonewho still feels that they are likely to find bombshellsif they get their hands on a Fund report is likely to bedisappointed.119. One conceivable conclusion from this is that it isnot particularly worthwhile to publish the reports—their contents already having been absorbed by mar-ket participants, in addition to having been leaked oc-casionally to the media. However, this would in ourview not be the correct conclusion. Admittedly, thereis little to be gained in terms of market functioning oreven Fund impact through markets. However, thereare still definite gains to be had. One, of a subsidiarynature, is that consistent publication would establish amore level playing field as regards the informationavailable—a generally good thing. More broadly,publication would increase the Fund’s accountability.120. One aspect is the increased accountability to thepublic in general—the taxpayer, the voter. This alsoenhances the Fund’s broad political legitimacy. An-other, of particular interest to the team, is the sys-tematic availability of more comprehensive informa-tion on the kind of advice given. This in turn shouldallow more informed criticism, favorable or other-wise, of that advice. And this should, correspond-ingly, act to enhance the quality of that advice, or atleast keep it up to the mark. In particular, referringback to earlier discussion in this chapter, therewould be a wider range of judgment as to whetherthe advice was sufficiently frank in the circum-stances (“clientism”) or whether the Fund diagnosisor model was adequate or too rigidly applied (“tem-plate”). To the extent such criticisms held water, theincentive to improve would be greater.121. One possible objection to publication is thatthis would provide material that could be used tocriticize, even unfairly, the member government, andthat some are better able to deal with such criticismthan others. We appreciate this point but do not re-gard it as conclusive. Dealing with criticism is notnecessarily a bad thing. Furthermore, it is now gen-erally accepted that the Fund is not infallible, andpublication of its advice is as likely to draw criticism

of that advice—again, fair or unfair—as it is to beused to criticize the recipient.

We recommend that the Fund should publishon a regular basis the complete text of all Arti -cle IV reports.101

122. It should be added here that we believe thatthese reports should not be vetted before publication,either by the member in question or by the Fund, toremove “highly market-sensitive” material. This isconsistent with our position that the reports alreadyreceive substantial self-censorship, given thechances that they would be leaked. Also, of course,such vetting would muddy the waters as regards ac-countability. While we appreciate the reasons for the“highly market-sensitive” reservation, in the circum-stances that we have analyzed in this section, we arenot persuaded by them. However, if the reservationis kept, we think that the process would becomemore internally transparent, and hopefully more rig-orously consistent, if proposals for exclusion werebrought back to the Board for active considerationrather than being left to the Director in question andthe staff to resolve, as is now the case with the PIN.123. We are aware that there is now a pilot projectthat involves the voluntary publication of abouttwenty Article IV staff reports over a year. It is notclear to us what this process will demonstrate, giventhe self-selection bias that is involved in what ispublished, although we are aware that efforts arebeing made to gather a “robust sample” of countries.It would, however, set a precedent along the linesthat we recommend pursuing on a comprehensivebasis, except that the project envisages, as in thecase of the PIN, the exclusion from publication ofmaterial that is deemed highly market sensitive.

Where Confidential Exchange Fits In

124. Given the importance that some members haveplaced on their ability to have a confidential exchangewith the Fund, we return here to that question.102

125. The basic point made in this section is not thatconfidential exchange should be ruled out. Rather, itis that the consultation process—involving as it doesan inevitably broad international peer review of a

77

100An interesting recent example—since it relates to a large in-dustrial country whose economy is generally performing well andwhich might therefore be expected to be less sensitive—is theUnited Kingdom. A number of Executive Directors complainedthat the most recent staff report made only passing reference toEMU—clearly the most pressing medium-term macroeconomicissue—and expressed their belief that this reflected the staff ’s viewthat the U.K. authorities would not welcome such a discussion.

101We recognize that there are some legal issues here. The Fundcould publish a staff report—provided that it did not contain theviews of the Executive Board or any confidential material pro-vided by the country—without the country’s permission. How-ever, if the report contained the Board’s views, significantly morerestrictive criteria would apply. We take the view that if the Boarddecides in principle in favor of this approach then these issues canbe resolved.

102See also the note by Chairman John Crow, “Confidential Ex-change—An Elaboration,” included as Appendix V of this report.


member’s policies in the light of that member’sstatutory obligations to the Fund, in other words,“firm surveillance”—is not realistically to bethought of as a confidential exchange. Furthermore,as we have suggested, given the on-the-ground re-ality of consultations, little in terms of the qualityof the consultation should be lost by making the reports public. And, on balance, with the increasedopenness the impact of surveillance would be enhanced.126. We also believe that any need for confidentialexchange, as there might well be, for example, overhypothetical policies, should be dealt with outside theframework of the Article IV consultations. Any suchdiscussions could only be reported to the Board in aquite general way if their substance was expected toremain confidential. Such confidential discussions (asone of our interlocutors put it, “to bounce off ideas”)could be held at mutual convenience and in any formor venue that seems appropriate. They could eventake place in the margins of the formal Article IV dis-cussions (as, we suspect, they already do).

The Role of the PIN

127. Where would our recommendation leave thePIN? While its general significance would clearly bediminished, it would in one crucial respect not be su-perseded—namely, in regard to the expression ofBoard views that concludes the consultation process.Of course, it might well be that views expressed inthe Board will sometimes be at variance with thosein the staff report. Then, outsiders would see thismore clearly. This should not be a particular causefor concern. Of course, if differences were to persist,this would be a cause for concern that the Fundwould need to address—as it would even if the re-ports were not published.128. The issuance of a PIN should also, of course, bestandard, not voluntary. 129. In particular, the PIN should set out clearly theconsensus or majority view of the Board on the mainpolicy issues (rather than, as now, emphasizing thediversity of views). Furthermore, perhaps now thatsome two years of experience have elapsed, therewill be more confidence in moving the PIN to amore straightforward level of communication.

We recommend that the Fund intensify its efforts to make the PIN more reader-friendly.

Recent Developments

130. As we have worked on this report, a number ofnew proposals related to surveillance have emergedfrom policymakers inside and outside the Fund. Wewere not able to discuss these new initiatives in

depth with country officials and other interviewees,since they were—and are—still in the process of de-velopment. However, in this section we discusstwo—the application of international standards incountry surveillance, and Contingent Credit Lines—that are likely to have particularly important impli-cations for the topics that we cover.

International Standards

131. The Fund has made significant progress in im-plementing better statistical standards for an impor-tant segment of its members. Similarly, in two of itsareas of core competency—fiscal and monetary pol-icy—the Fund has developed a “Code of Good Con-duct on Fiscal Transparency” and is well on its wayto formulating a similar code for monetary policy incooperation with major central banks and the BIS.These efforts are intended to help in making impor-tant aspects of surveillance more standardized andobjective through the detailed checklists providedby the codes.132. However, it is worth noting that even in theFund’s core area of expertise, this will present chal-lenges. Monitoring standards of fiscal transparencyis not the same thing, and does not necessarily re-quire the same skills, as monitoring the stance of fis-cal policy. Monitoring standards is likely to requiresubstantially more institutional knowledge, cross-country experience, and perhaps background in pol-icy formulation and implementation (and corre-spondingly rather less in the way of basicmacroeconomic modeling skills). This reinforces anumber of our earlier recommendations:

• the need for a more diverse skills mix amongFund staff;

• the increased focus in surveillance on knowl-edge transfer and cross-country experience;and

• the need to integrate better the work of func-tional and area departments, with missionsmaking greater use of functional departmentstaff.

133. These points are perhaps most important in thefinancial sector area, which has proved once moreduring the Asian financial crisis to be closely relatedto the Fund’s macroeconomic concerns, and which,as discussed above, is clearly important to more ef-fective surveillance. For a limited number of coun-tries—perhaps a dozen a year—specialized missionsthat would prepare Financial Sector Stability Assess-ments (FSSAs) are now envisaged. This will requireadditional staffing, largely in the form of temporarysupport from central banks and other financial su-pervisors, in addition to intensified cooperation with

78

Recent Developments

the World Bank, particularly in nonindustrial coun-tries. The Fund’s plans in this area seem realistic andappropriate, provided the Fund can succeed in at-tracting temporary contractual expertise to assist itspermanent staff, since it would take a long time tobuild up in-house expertise in the areas required forthe FSSAs.134. However, this task seems demanding enough.We are concerned that if, as the G-7 finance ministerspropose,103 Article IV consultations are expected tocover, in addition to the financial sector issues ad-dressed by FSSAs, monitoring of country compli-ance with numerous other international standards,then overstretch is inevitable. In our view, in areassuch as securities market regulation, accounting andauditing, bankruptcy legislation, and corporate gov-ernance, the Fund staff lack the professional capacityto participate in developing the relevant internationalstandards, and arguably even to monitor them. Inmost of these cases, other international institutionssuch as the World Bank, the OECD, the Basle Com-mittee for Banking Supervision, and private or inter-national organizations—the International Organiza-tion of Securities Commission (IOSCO) andprofessional associations of lawyers, accountants, orauditors—have developed international standards onwhich international monitoring might be built. But inseveral cases the standards remain incomplete, eitherbecause there are outstanding disagreements or be-cause the geographical coverage of the preparatorywork has been far less than global.135. More important, monitoring these standards isin most cases not a simple matter of “ticking boxes,”as it is sometimes presented; it requires a consider-able degree of professional expertise, and the abilityto ask probing questions in a variety of areas. With-out in any way detracting from the capacities ofFund staff, we think it is unreasonable to ask them toacquire the requisite expertise across such a widerange of topics, especially given the overstretch thatexists already. In other words, we do not believe it tobe realistic to incorporate the monitoring of interna-tional standards outside the Fund’s core areas of re-sponsibility104 into Article IV consultations. 136. Instead, we are attracted by the idea developed,for example, by Eichengreen (1999), that not onlythe formulation of the international standards, butalso their monitoring, should to the largest extent

possible be the responsibility of the above-men-tioned organizations and associations. The Fundwould simply report their views on country compli-ance in an annual publication and on its web site (asthe Annual Report on Exchange Arrangements andExchange Restrictionsalready does). Not onlywould this reduce the problem of overstretch, itwould also soften a potentially serious conflict of in-terest for the Fund, since compliance with standardswill be one of the criteria that the Fund will have toconsider in connection with elegibility for the Con-tingent Credit Line (see below).

We recommend that outside the Fund’s coreareas, monitoring international standards tothe maximum extent possible be delegated tothe responsible international bodies, with theFund’s role largely confined to that of a clear-inghouse for information.

Contingent Credit Line

137. The Contingent Credit Line (CCL) is a direct re-sponse to the contagion observed in and around theAsian crisis. In particular, some argued that while fi-nancial crises in certain countries were clearly relatedto macroeconomic imbalances or weaknesses in thebanking system, the crisis quickly spread to othercountries where weaknesses were less obvious or im-mediate. In other words, some international investorswere simply pulling out of all emerging markets,without looking closely at individual countries’fun-damentals. For example, immediately after the Russ-ian devaluation and default of August 1998, therewas a dramatic widening of interest rate differentialsbetween almost all emerging markets and the largeindustrial countries—even though economic funda-mentals in many emerging market countries wereclearly wholly different from those in Russia. 138. The CCLproposes to provide a form of insuranceto Fund members with sound fundamentals, by allow-ing them to “prequalify” for the use of Fund resources,if they are affected by a crisis of confidence not obvi-ously linked to their own policies and performance.Ideally, if the CCLsucceeds in its objectives, the re-sources themselves need never be used, since theiravailability should in itself give the markets confi-dence that the country will not suffer a financial crisis. 139. A decision to set up the CCL, initially for oneyear, was adopted by the Executive Board in April1999. For a country to qualify, the Board will have tosatisfy itself that four sets of conditions have been met:

• that the member is unlikely, on the basis ofcurrent policies, to need Fund resources;

• that its policies have received a “positive assess-ment during the latest Article IV consultation”

79

103Communiqué of April 25, 1999.104To reiterate, we consider that the core areas of responsibility

are exchange rate policy, macroeconomic policy, financial sys-tems, and capital account issues. While the Fund’s current viewof the core seems more expansive than that—see, for example,the discussion in Chapter I of this report—it does seem that onthis question of international standards, the Fund’s view and oursdo coincide.


and beyond; and that the member adheres to“relevant internationally accepted standards”;

• that the member has appropriately involvedprivate sector creditors in limiting externalvulnerability; and

• that a satisfactory economic and financial pro-gram has been submitted.

140. The CCLwill blur the previously clear distinc-tion between Fund surveillance and application ofconditionality. We have in our interviews and analy-sis accepted the time-honored distinction in thesetwo types of relationships between the Fund and amember; surveillance is the peacetime activity,while the conditionality attached to the use of Fundresources involves much tighter monitoring, typi-cally based on quantitative performance criteria.105

However, this neat distinction between surveillanceand conditionality would not be applicable follow-ing the introduction of a CCL. This would clearlyhave major implications for surveillance proceduresin all the dimensions that we have tried to evaluate:conduct and method, substance of advice, and im-pact. We believe that these implications tend to rein-force our principal conclusions:

• Surveillance would have to become more con-tinuous, since if policies went off track be-tween annual consultations, the Fund wouldhave to withdraw access to the CCL.

• Staff appraisals and Board conclusions wouldhave to become clearer and more unambiguous,since the Fund would have to determine whethera member is or is not eligible for the CCL.

• Surveillance would have to become more fo-cused on the macroeconomic and external is-sues that are the core competency of the Fund,since it is these that will be most relevant to acountry’s ability to qualify for the CCL. More-over, more attention will need to be paid to ex-ternal vulnerabilities.

• Since it will presumably be made public whichcountries have qualified for the CCL, the casefor publishing the Article IV staff reports,

which will form the principal basis on whichthe Fund decides whether a country shouldqualify, will be strengthened.

141. The new facility also has attractions in terms ofincreasing the impact of surveillance. We have notedthat in general, impact has been low; and that inlarge part this is simply a consequence of the factthat as long as members do not expect to have todraw on Fund resources, external advice is unlikelyto have as much impact as internal forces. But sincea positive assessment of economic policies and per-formance is a prerequisite for qualifying for use ofthe CCL, the clout of surveillance is likely to in-crease significantly. From this perspective, a well-designed CCLwould add to the potential effective-ness of surveillance.142. However, we do have some concerns. Our ear-lier discussion suggests some grounds for skepticismas to the ability of the surveillance process, as cur-rently constituted, to discharge the demanding taskshighlighted by the adoption of the CCL. In particu-lar, we see three possible problems:

• the difficulties of applying internationally ac-cepted standards in an increasing number ofcomplex areas, discussed above;

• the additional tension between the Fund’s tra-ditional role as a policy adviser and the need tobecome in effect a rating agency; and

• the likelihood that the Board will come underpressure to become lax in its judgment as towhether a member has met the criteria for quali-fying for the CCL, and the even greater diffi -culty of withdrawing the CCLif a member’spolicies or situation deteriorate materially.

143. As regards transparency, we have tried aboveto define what we understand as the remainingscope for confidentiality prior to the adoption of theCCL. Since the whole idea of the CCL—a publicdemonstration of the Fund’s faith in a country’spolicies and prospects, designed to engender marketconfidence—requires publication of the decision tocommit resources to a member under the new facil-ity, the CCLcan only enhance the need for trans-parency. On the other hand, some national authori-ties may be unhappy at the prospect of discussingtheir reaction to possible unfavorable disturbancesin the knowledge that such exchanges would be in-cluded in a published document. Even when suchdiscussions become essential in prequalifying forthe CCL, some role may need to be preserved forconfidential exchanges, though the Board will needto be more fully informed than through subse-quently published Article IV reports. This would en-hance the role of the informal country matters andWEMD sessions in the Board.

80

105Some of the countries we have looked at more closely havepassed from one state to the other as they entered into Fund pro-grams: Thailand in July 1997, Korea in December 1997, andBrazil in November 1998. In accordance with our terms of refer-ence, and since the Fund involvement in—and influence over—amember’s policies changes qualitatively around such dates, wehave confined our study to surveillance in the classical sense. Ofcourse, as mentioned in Chapter I, the advice given in the surveil-lance context is to some extent likely to foreshadow program con-ditionality. But if, as in these countries, the program follows a cri-sis, circumstances will have changed significantly, and sonecessarily will the Fund’s policy prescriptions.

Recent Developments

144. Finally, there is a clear risk that the Board, giventhe tradition for individual Executive Directors to de-fend the interests of their respective constituenciesand to show considerable deference to each other, willbe reluctant to deny CCLstatus to a country that seeksit. Even more so, it will be very difficult, given thepossibility of adverse market reaction, to “down-grade” a country whose performance has deterioratedsince the initial commitment (of course, this problemexists already with respect to Fund-supported pro-grams). In our view, this emphasizes the need forgreater frankness in staff reports, greater transparency,and more pointed discussions in the Board. Moreover,formulations such as “some Directors felt . . . ; how-ever, other Directors stressed . . .” will not be useful in

a published PIN, no matter how accurate the render-ing of the discussions. The Board will have to assumea greater degree of collective responsibility in caseswhere a CCLdecision is proposed than has been cus-tomary in discussions of Article IV reports.145. On balance, the addition of the CCLfacilityshould add to the effectiveness of surveillance byraising the demand for high-quality assessments byFund staff and the Board and by adding financialclout to surveillance, hence strengthening its im-pact. But the CCLand the associated reliance oncomplex, largely judgmental, considerations and ona number of international standards will also furtherexpose some of the weaknesses in past surveillanceto which we have drawn attention in this report.

81

1. Purpose of the Evaluation

The Executive Board of the International Mone-tary Fund has decided to request independent exter-nal experts to conduct an evaluation of Fund surveil-lance over members’policies under Article IV of theArticles of Agreement. The purpose of the evalua-tion is to assess the effectiveness of Fund surveil-lance and to make recommendations for improve-ments consistent with the purposes of the IMF asdefined in Article I.

2. Focus of the Evaluation

The evaluation will focus on the effectiveness of allaspects of Fund surveillance over members’policiesand will be carried out by three independent externalexperts, as indicated in Section 3 below. The expertsare requested to include in their evaluation all chan-nels and instruments of Fund surveillance, includingbilateral surveillance, regional surveillance, multilat-eral surveillance and the content and format of theWorld Economic Outlook, and surveillance of inter-national capital markets and financial systems and ofthe provision by member countries of economic andfinancial data to the IMF and the public. The expertsare requested to consider four broad topics in theirevaluation of Fund surveillance, including on thebasis of case studies as indicated in Section 4 below:

A. The effectiveness of surveillance in identifyingthose macroeconomic, structural, and financialweaknesses and imbalances in member coun-tries and the world economy that are an obsta-cle to achieving sustainable noninflationaryeconomic growth and external viability.

B. The substance of Executive Board and Fundstaff surveillance recommendations: are thesepolicy recommendations of the appropriaterelevance, realism, and timeliness?

C. The impact on members’policies of the Fund’ssurveillance recommendations: has an appro-priate impact been achieved, and what doesimpact depend upon?

D. The conduct and methods of surveillance: howhelpful are the procedures of surveillance, theresources and staff skills employed, the meansof interaction with member country authori-ties, and the dissemination methods of Execu-tive Board surveillance conclusions?

In focusing on the above topics, the experts maywish to be aware of the following more specificquestions that are of interest to Executive Directors:

(i) How did the Fund’s advice correspond tothe short-term objectives and medium-termstrategies of existing policies?

(ii) How did this advice correspond to theanalysis and advice of other domestic andinternational institutions? Did the Fund’sadvice add value?

(iii) Has Fund surveillance paid sufficient atten-tion to regional surveillance, to interactionamong countries, and to the external effectsof policies in major countries?

(iv) Did surveillance have different impacts indifferent groups of countries?

(v) Have the frequency and general focus of theFund’s surveillance been appropriate withhindsight? Has advice been consistent? Hasadvice helped foster noninflationary eco-nomic growth?

(vi) How successful have been the specific ef-forts made since early 1995 to strengthensurveillance? What effect have these effortshad in the context of the difficulties emerg-ing in some Asian countries in 1997? Howhas the provision of information by the au-thorities affected surveillance?

(vii) Did the advice take into appropriate ac-count the institutional, political, and socialframework? Did it pay adequate attention tothe uncertainties and political constraintsthat lead to “small” deviations from first-best policies?

(viii) Was the Fund’s advice implemented? If not,why?

Appendix I Terms of Reference

82

Terms of Reference

(ix) Did the Fund balance openness and sensi-tivity appropriately from the authorities’perspective? How helpful were the docu-ments that emerged from bilateral surveil-lance? How useful were the methods tomake surveillance conclusions available tothe public? Should the Fund go further inpublishing country reports?

(x) How did governments disseminate surveil-lance conclusions within and among gov-ernment institutions? Was the circle of par-ticipants limited to economic agencies?

(xi) How do you assess the role of the ExecutiveBoard in surveillance?

3. Evaluators and Their Independence

Mr. Ricardo Arriazu, Mr. John Crow, and Prof.Niels Thygesen, working as a team, have agreed toconduct the evaluation; Mr. Crow will serve as chair.They shall conduct their work freely and objectivelyand shall render impartial judgment and make rec-ommendations to the best of their professional abili-ties. At their full discretion, the evaluators may wish

to take into account the views of member countryauthorities, parliamentarians, academic experts, rep-resentatives of other international organizations, rep-resentatives of the business and financial marketcommunities, representatives of civil society and themedia, and Fund Executive Directors and staff.

4. Selection of Countries for the Case Studies

The evaluators are requested to base their conclu-sions, in part, on the study of a limited number ofcountry cases. The selection of country cases will bethe responsibility of the evaluators. Countries cho-sen for study should be representative of the Fund’smembership in terms of size, geographic location,and the variety of the macroeconomic, structural,and financial issues encountered. The effectivenessof Fund surveillance in these countries should beevaluated over a time period that is long enough toallow such insight as is possible into acceptance andoutcome of surveillance recommendations and dur-ing which the countries received no, or at most spo-radic, disbursement of Fund resources.

83

For the purposes of the surveillance evaluation ex-ercise, the following are the relevant parts of theFund’s Articles of Agreement.

Article I

The purposes of the International Monetary Fundare:

(i) To promote international monetary coopera-tion through a permanent institution whichprovides the machinery for consultation andcollaboration on international monetaryproblems.

(ii) To facilitate the expansion and balancedgrowth of international trade, and to con-tribute thereby to the promotion and main-tenance of high levels of employment andreal income and to the development of theproductive resources of all members as pri-mary objectives of economic policy.

(iii) To promote exchange stability, to maintainorderly exchange arrangements amongmembers, and to avoid competitive ex-change depreciation.

(iv) To assist in the establishment of a multilat-eral system of payments in respect of cur-rent transactions between members and inthe elimination of foreign exchange restric-tions which hamper the growth of worldtrade.

(v) To give confidence to members by makingthe general resources of the Fund temporar-ily available to them under adequate safe-guards, thus providing them with opportu-nity to correct maladjustments in theirbalance of payments without resorting tomeasures destructive of national or interna-tional prosperity.

(vi) In accordance with the above, to shorten theduration and lessen the degree of disequilib-rium in the international balances of pay-ments of members.

The Fund shall be guided in all its policies and de-cisions by the purposes set forth in this Article.

Article IV

Section 1. General Obligations of Members

Recognizing that the essential purpose of the in-ternational monetary system is to provide a frame-work that facilitates the exchange of goods, services,and capital among countries, and that sustains soundeconomic growth, and that a principal objective isthe continuing development of the orderly underly-ing conditions that are necessary for financial andeconomic stability, each member undertakes to col-laborate with the Fund and other members to assureorderly exchange arrangements and to promote astable system of exchange rates. In particular, eachmember shall:

(i) endeavor to direct its economic and finan-cial policies toward the objective of foster-ing orderly economic growth with reason-able price stability, with due regard to itscircumstances;

(ii) seek to promote stability by fostering or-derly underlying economic and financialconditions and a monetary system that doesnot tend to produce erratic disruptions;

(iii) avoid manipulating exchange rates or theinternational monetary system in order toprevent effective balance of payments ad-justment or to gain an unfair competitiveadvantage over other members; and

(iv) follow exchange policies compatible withthe undertakings under this Section.

Section 3. Surveillance Over ExchangeArrangements

(a) The Fund shall oversee the international mone-tary system in order to ensure its effective operation,and shall oversee the compliance of each memberwith its obligations under Section 1 of this Article.

Appendix II Articles of Agreement

84

Articles of Agreement

(b) In order to fulfill its functions under (a)above, the Fund shall exercise firm surveillanceover the exchange rate policies of members, andshall adopt specific principles for the guidance ofall members with respect to those policies. Eachmember shall provide the Fund with the informa-tion necessary for such surveillance, and, when re-quested by the Fund, shall consult with it on themember’s exchange rate policies. The principlesadopted by the Fund shall be consistent with coop-

erative arrangements by which members maintainthe value of their currencies in relation to the valueof the currency or currencies of other members, aswell as with other exchange arrangements of amember’s choice consistent with the purposes ofthe Fund and Section 1 of this Article. These prin-ciples shall respect the domestic social and politicalpolicies of members, and in applying these princi-ples the Fund shall pay due regard to the circum-stances of members.

85

The list below is set out as follows. We do not sep-arately identify Fund staff or Executive Directors byname. We list the people we met on our country vis-its by country, identifying their position at the timeof the interview, or relevant former affiliation. Sepa-rately, we list officials of other governments and in-ternational institutions, academics we saw outsidespecific country visits, private sector, and NGO rep-resentatives. A number of those we met were formermembers of the Fund staff or management, or FundExecutive Directors; these are identified by an aster-isk. We apologize for any errors or omissions.

Staff

We met with a large number of Fund staff (morethan 50). Among senior management, these includedthe Managing Director, Deputy Managing Directors,and senior management from the relevant area andfunctional departments. We also met with the staffdirectly responsible for surveillance for the countriesin our sample, and, formally and informally, withnumerous other staff in a variety of departments andat a variety of levels.

Executive Directors

We met with a number of Executive Directors(representing the majority of Executive Board con-stituencies), their Alternates, and Advisors.

Brazil

Ana Teresa H. De Albuquerque, Secretaria doTesouro Nacional

Edmar Bacha, Creditanstalt, formerly Secretaria doTesouro Nacional

Fabio O. Barbosa, Secretaria do Tesouro NacionalGustavo Bussinger, Banco Central do Brasil

*Daniel L. Gleizer, Credit Suisse, First BostonEduardo Refinetti Guardia, Secretaria do Tesouro

Nacional

Francisco L. Lopez, Governor, Banco Central doBrasil

Gustavo Loyola, Tendências Consultoria Integrada,formerly Governor, Banco Central do Brasil

Demosthenes Madureira de Pinho Neto, BancoCentral do Brasil

*Alvaro Manuel, Ministry of PlanningAlkimar R. Moura, Escola de Administraçao de

Empresas de Sao Paulo, formerly Banco Centraldo Brasil

Claudio Ness Mauch, Banco Central do BrasilMailson F. Da Nóbrega, Tendências Consultoria In-

tegrada, formerly Minister of FinanceMarcelo Piancastelli de Siqueira, Secretaria do

Tesouro NacionalRoberto Egydio Setubal, President, Brazilian

Bankers Association Jose Tavares, Ministry of Planning

Chile

Vittorio Corbo, Universidad Católica de ChileAlejandro Foxley, Senator, formerly Minister of

FinanceGuillermo LeFort, Banco Central de ChileCarlos Massad, Governor, Banco Central de ChileJoaquin Vial Ruiz-Tagle, Ministerio de HaciendaJuan Villarzu, Empresa de Obras Sanitarias de

Valparaiso, formerly Secretary General to thePresidency

Roberto Zahler, Zahler & Co., formerly Governor,Banco Central de Chile

China

Weiping Di, State Development BankYou Guo, China Everbright BankHongbo Huang, State Administration of Foreign

ExchangeWeiping Huang, Renmin University of ChinaHongmei Han, State Administration of Foreign

ExchangeXuejun Kang, Ministry of Finance

Appendix III List of Interviewees

86

List of Interviewees

Mingzhi Liu, People’s Bank of ChinaZhengming Liu, People’s Bank of ChinaYiping Peng, China Everbright BankJie Shao, State Administration of Foreign ExchangeHong Sheng, Beijing UNIRULE Economy Re-

search InstituteGouqing Song, Beijing UniversitySining Tang, State Administration of Foreign

ExchangeBenhua Wei, People’s Bank of ChinaFulin Wu, China Everbright BankPing Xie, People’s Bank of ChinaJunmei Yang, Ministry of FinanceXian Zhu, Ministry of Finance

Czech Republic

Richard Falbr, Confederation of Trade UnionsOta Kaftan, Czech National BankMiroslav Hrncír, Czech National BankVaclav Klaus, Member of Parliament, formerly

Prime Minister and Minister of FinanceIvan Kocarnik, Ceska Pojistovna, formerly Minis-

ter of FinancePavel Kysilka, Czech National BankVera Masindova, Czech National BankLudek Niedermayer, Czech National BankPavel Stepanek, Ceska SpositelmaJosef Tosovsky, Governor, Czech National Bank,

formerly Prime Minister

Hong Kong SAR

Gary Coull, Global Emerging Markets, Crédit Lyonnais

James Lau, Hong Kong Monetary AuthorityGeorge Pickering, Bank for International

SettlementsAndrew Sheng, Hong Kong Security and Futures

Commission

India

Shankar N. Acharya, Ministry of FinanceMontek Singh Ahluwalia, Planning CommissionShri Chidambaran, formerly Minister of FinanceTarun Das, Ministry of FinanceSandip Ghose, Reserve Bank of IndiaOmkar Goswami, Confederation of Indian IndustryV. Govindarajan, Ministry of Finance

*Bimal Jalan, Governor, Reserve Bank of IndiaVijay Kelkar, Ministry of FinanceRohit Modi, Ministry of FinanceH. Prasad, Ministry of Commerce

T.R. Prasad, Ministry of IndustryYashwant Sinha, Minister of FinanceM.R. Srinivasan, Reserve Bank of IndiaSatya Pal Talwar, Reserve Bank of India

*Sawak S. Tarapore, formerly Reserve Bank of India*Asuri Vasudevan, Reserve Bank of India

Japan

Yoichi Funabashi, The Asahi ShimbunToyoo Gyohten, Special Adviser to Prime Minister,

formerly Deputy Minister of FinanceKyoto Ido, Ministry of Finance

*Takatoshi Ito, Hitotsubashi UniversityMasaaki Kanuo, Japan Economic Research Center

*Michio Kitahara, Bank of JapanRichard Koo, Nomura Research InstituteYutaka Kosai, Japan Economic Research Center

*Haruhiko Kuroda, Ministry of Finance*Takashi Murakami, Bank of Japan*Takehiko Nakao, Ministry of Finance*Yoshio Okubo, Ministry of Finance*Eisuke Sakakibara, Ministry of FinanceSeiji Shimpo, Economic Planning Agency

*Masahiko Takeda, Bank of JapanKazuo Ueda, Bank of JapanMikio Wakatsuki, Japan Research Institute, for-

merly Bank of JapanKoji Watanabe, KeidanrenMasaru Yoshitomi, Long Term Credit Bank of Japan

Saudi Arabia

Mohamed Aba Al-Khail, Gulf International Bank,formerly Minister of Finance

Abdulwahab Attar, Minister of PlanningIbrahim Al-Assaf, Minister of FinanceTameel Al-Hojailan, Secretary General, Gulf Coop-

eration Council*Muhammad Al-Jasser, Saudi Arabian Monetary

AgencyAbdullah Al-Kuwaiz, Gulf International Bank,

formerly Deputy Minister of FinanceHamad Al-Sayari, Governor, Saudi Arabian Mone-

tary AgencyJobarah Al-Soraisry, Ministry of FinanceKevin Taecker, Saudi American Bank

South Africa

Jim Buys, Anglo American Corporation of SALtd.Estran Calitz, Professor of Economics, University of

South Africa, formerly Ministry of FinanceDennis Dykes, NEDCOR

87

APPENDIX III

Alec Erwin, Minister of Trade and IndustryEvan P.J. Franklin, SAReserve BankBernie L. de Jager, SAReserve BankChristo Liebenburg, NEDCOR, formerly Minister of

FinanceMxolisi Lindie, Department of FinanceTrevor Manuel, Minister of FinanceTito Mbowene, SAReserve BankGuy Mhone, Department of LaborJay Naidoo, NEDLACRaymond Parsons, SACOBFrancois le Roux, Department of FinanceChris Stals, Governor, SAReserve BankMatthys Strauss, SACOBTimothy T. Thahane, SAReserve BankBen van Rensburg, SACOBLambertus (Bertus) van Zyl, SAReserve Bank

Sweden*Krister Andersson, Skandinaviska Enskilda Banken,

formerly Sveriges RiksbankClaes Berg, Sveriges RiksbankSune Davidsson, Liberal Party of SwedenBengt Dennis, Skandinaviska Enskilda Banken,

formerly Governor, Sveriges RiksbankThomas Franzén, National Debt OfficeIngemar Hansson, Ministry of FinanceUrban Hansson, Ministry of FinanceLars Heikensten, Sveriges RiksbankStefan Ingves, Sveriges RiksbankTomas Nordstrom, Ministry of FinanceSvante Öberg, National Institute for Economic

Research*Eva Srejber, ForeningssparbankenAnn Wibble, Industriforbundet, formerly Minister of

Finance

Thailand*Pisit Leeahtam, Ministry of FinanceKleo-Thong Hetrakul, Bank of ThailandTarrin Nimmanahaeminda, Minister of FinanceNukul Prachuabmoh, formerly Governor, Bank of

ThailandDavid Proctor, Bank of AmericaAmmar Siamwalla, Thailand Development Re-

search InstituteAmaret Sila-On, Financial Regulation Authority,

formerly Minister of IndustryChatu Mongol Sonakul, Governor, Bank of Thailand

*Vij it Supinit, Parliament of Thailand, formerlyGovernor, Bank of Thailand

Jon Vanasin, Financial Regulation AuthorityVicharat Vichit-Vadakan, Financial Regulation

Authority

Amnuay Viravan, Saha Union Co., Ltd., formerlyMinister of Finance

Worawut Wesaratchakit, Financial Regulation Authority

Chaiyawat Wibulswasdi, formerly Governor, Bankof Thailand

United States

Steven N. Braun, Council of Economic AdvisorsTom Connors, Federal Reserve BoardJeffrey Frankel, Council of Economic AdvisorsAlan Greenspan, Chairman, Federal Reserve BoardKaren Johnson, Federal Reserve BoardDonald Kohn, Federal Reserve BoardLarry McDonald, U.S. TreasuryNuriel Roubini, Council of Economic AdvisorsDavid Stockton, Federal Reserve BoardEdwin Truman, U.S. Treasury

Small States

Representatives of Antigua, the Bahamas, Barbados,Botswana, Dominica, the Eastern Caribbean Cen-tral Bank, Jamaica, St. Vincent and theGrenadines, Samoa, the Seychelles, and Trinidad.

Officials of International Institutionsand Governments Other Than Those of Sample Countries

Johan Barras, DG2, European CommissionWillem Buiter, Bank of EnglandHervé Carre, DG2, European Commission

*Andrew Crockett, General Manager, Bank forInternational Settlements

Jon Cunliffe, HM Treasury, United KingdomE. A. Evans, Secretary, The Treasury, Australia

*Günther Grossche, Secretary, Monetary Committee,European Union

André Icard, Bank for International SettlementsOtmar Issing, European Central BankDonald Johnston, Secretary General, OECDMervyn King, Bank of EnglandJürgen Krueger DG2, European CommissionJohn P. Martin, OECDThorvald Moe, OECDJohn Murray, Bank of Canada

*Gus O’Donnell, HM Treasury, United KingdomTommaso Padoa-Schioppa, European Central BankRinaldo Pecchioli, OECDJean Pisani-Ferry, Tresor, FranceStephen Potter, OECD

88

List of Interviewees

Arnaud Schneiweiss, Tresor, FranceKumiharu Shigehara, OECDPhilip Turner, Bank for International SettlementsIgnazio Visco, OECDJohn West, OECDWilliam White, Bank for International Settlements

*Nigel Wicks, Chairman, Monetary Committee, European Union

*John Williamson, World Bank

Academics

Ralph Bryant, Brookings InstitutionRichard Cooper, Harvard UniversityWendy Dobson, University of Toronto

*Barry Eichengreen, University of California,Berkeley

Martin Feldstein, National Bureau of Economic Re-search/Harvard University

Benjamin Friedman, Harvard UniversityPeter Kenen, Princeton UniversityFrederic Mishkin, Columbia UniversityRichard Portes, London Business School/Center for

Economic Policy ResearchJeffrey Sachs, Harvard UniversityJan Art Scholte, Warwick University

NGOs

Gemma Adaba, International Confederation of FreeTrade Unions

Jo Marie Griesgraber, Center of ConcernStephen Pursey, International Confederation of

Free Trade UnionsChristine Real de Azua, Accounting for the

EnvironmentCarol A. Webb, Friends of the Earth

Private Individuals

Jeffrey Anderson, Institute for International Finance

Kevin Barnes, Institute for International FinanceEric Barthalon, ParibasFred Bergsten, Director, Institute for International

Economics*Sterie T. BezaJohn Chambers, Standard & Poor’sRobert Chote, Financial TimesWilliam Cline, Institute for International Finance

*Charles Dallara, Managing Director, Institute forInternational Finance

*Richard ErbGregory Fager, Institute for International Finance

*Joaquin Ferrán*David Folkerts-Landau, Deutsche Morgan GrenfellLacey Gallagher, Standard & Poor’s

*Morris Goldstein, Institute for InternationalEconomics

John Hartzell, Dresdner BankRandall Henning, Institute for International

EconomicsHelena Hessel, Standard & Poor’sChristopher Huhne, Fitch-IBCA

*Desmond Lachman, Salomon Brothers*John Lipsky, Chase Manhattan BankAnders Ljungh, Morgan StanleyDavid Malpass, Bear StearnsCatherine Mann, Institute for International

EconomicsKen Pinkes, Moody’s

*Jacques PolakAdam Posen, Institute for International EconomicsLex Rieffel, Institute for International Finance

*Douglas Smee, CitibankBritt Swofford, BancOne

*Leo Van HoutvenKal Wajid, Institute for International Finance

89

*Former Fund staff, management, or Executive Director.

1. We recommend that consultation guidance be re-structured to give explicit attention to interna-tional aspects.

2. We recommend that surveillance of the euro areacenter around the ECB and other EU bodies. Sur-veillance of individual participants in the euroarea should largely take place at the euro arealevel, and through EU institutions.

3. We recommend that (consistent with an increasedfocus on international aspects) the Fund bringspillover issues, whether regional or multilateral,directly to the table in its various country consul-tations and in Board discussions.

4. We recommend that surveillance focus, above all,on the core issues of exchange rate policy and di-rectly associated macroeconomic policies, in par-ticular the international implications of such poli-cies. Other analysis should only be undertaken ifdirectly relevant.

5. We recommend that a systematic process be devel-oped whereby the Board would discuss and sign offon the main issues to be raised at forthcoming indi-vidual Article IV consultation discussions.

6. We recommend that staff focus policy advice onissues of serious or immediate concern and distin-guish such advice clearly from analysis ofwhether relatively small or judgmental policyshifts would be helpful.

7. We recommend that in the next internal review ofsurveillance, more attention be given to measur-ing in some detail (by topic and country) the ex-tent to which the specific operational guidancethat has been put forward on behalf of the Boardis actually followed in Fund consultation reports,and, equally important, if not why not.

8. We recommend that the Board, management, andsenior staff attempt to alter the incentive structure

by making it clear that they will, if necessary,back up staff who give frank advice.

9. We recommend that surveillance devote more at-tention to policy implementation, and to theidentification and analysis of alternative policyoptions.

10. We recommend that Article IV staff reports givegreater attention to the forces driving the capitalaccount, and to capital account issues in general.

11. We recommend that surveillance devote substan-tially more attention to vulnerabilities.

12. We recommend quarterly publication of theWEO forecast.

13. We recommend that ultimate responsibility forWEO forecasting be vested clearly in the Re-search Department.

14. We recommend that the Board make it clear thatthe presumption is that the staff draft of theWEO/ICMR should be published as it stands.

15. We recommend that:

• For all industrial countries but the very largest,full Article IV consultations should be less fre-quent than annually.

• For the very largest industrial economies, inlight of their systemic impact, annual consulta-tions are still called for. However, to improve thepayoff, surveillance should focus more on theinternational implications of these countries’do-mestic policies and correspondingly less on ad-vice regarding the policies themselves.

• There is a strong case for cutting back the re-sources allocated to the euro area, which arenow more than four times those devoted to theUnited States.

16. We recommend a more conscious focus on the useof electronic means of communication, initially

Appendix IV Recommendations

90

Note:This list is attached for reference purposes only. Recommendations should be read in the context of the analysis and discussion in Chapter V.

Recommendations

on an experimental basis, with a view to maintain-ing close contact with policy officials and to re-ducing the length of consultation missions.

17. We endorse the recommendations of the internalreview that area departments should, through theconsultation process, seek to integrate the sup-port provided to members through technical as-sistance and surveillance activities.

18. We recommend that the External Relations De-partment should monitor and assess Fund prac-tices and experiences to date with a view to giv-ing guidance with the Fund’s explicit backingregarding external contacts.

19. We recommend that the Fund make clear wherenecessary that meetings with nongovernmentalrepresentatives should take place without thepresence of government officials.

20. We recommend:

• More emphasis on policy experience, andtherefore less on academic credentials, at all recruitment levels, including the EconomistProgram;

• Fund staff should be positively encouraged tospend one or more assignments outside theFund before reaching management grades. Cur-rent programs in this area should be expanded,if that proves necessary to ensure adequate opportunities.

21. We recommend that:

• There be as much accountability as possible forpapers, staff reports, and the associated policyrecommendations. In particular, the staff mem-ber most directly responsible for authorship of apaper or staff report, or with greatest knowledgeof the country being discussed, should be themain presenter at any Board discussion.

• In the surveillance context, it generally be thecase that one staff member—normally the Divi-sion Chief in the area department—has overallresponsibility for operational dealings, includingleading missions, with an individual memberstate, and should be held accountable for them.

22. We recommend that front offices in area depart-ments be made clearly accountable for ensuringthat bilateral surveillance incorporate cross-country and multilateral perspectives.

23. We recommend that the overall volume of sur-veillance work be reduced relative to the numberof personnel engaged in it.

24. We recommend that the staff:

• systematize and organize their use of outsideinformation with a view to applying it in sur-veillance; and

• refer to and/or summarize work produced byother organizations where relevant.

25. We recommend that the Fund experiment withexternal review of a sample of staff reports.

26. We recommend that all Article IV staff reportsbe discussed in the first instance in a committeerather than by the full Board.

27. We recommend that the Fund publish on a regu-lar basis the complete text of all Article IVreports.

28. We recommend that the Fund intensify its effortsto make the PIN more reader-friendly.

29. We recommend that outside the Fund’s coreareas, monitoring international standards to themaximum extent possible be delegated to the re-sponsible international bodies, with the Fund’srole largely confined to that of a clearinghousefor information.

91

Questions were raised at the Board’s informalmeeting of August 3 regarding what, in the surveil-lance evaluation team’s eyes, could represent confi-dential exchange between the Fund and the mem-ber, given the team’s emphasis on having the staffassess “vulnerabilities” in the consultation discus-sions. This note reiterates and elaborates on ourthinking.

As regards the vulnerabilities exercise, the impor-tant point to our mind is that the staff should presentits assessment for the member’s consideration as anintegral part of the consultations. That assessmentwould, in large measure, be conducted with refer-ence to the kinds of indicators entering into earlywarning systems, and would occur for all consulta-tions. We also recognize that in the consultation ex-ercise, the member may wish to respond to the as-sessment fully, partially, or not at all.

As regards confidential exchange, this could ap-propriately, as our report suggests, deal in generalwith hypothetical matters. And one part of any suchexchange could of course, if the member chooses, bea discussion of issues arising from the vulnerabilitiespresentation, if the authorities did not judge it desir-able to respond to such matters in the consultation it-self. Other hypothetical matters (e.g., possible orpending government or legislative action) could alsoof course be discussed in such an exchange.

We are also of the view, on practical grounds, thatthe product of such confidential exchange wouldnot be conveyed directly to the Board. By defini-tion, it is not part of the formal consultation proce-dure—the results of which would be made public.Rather, it would be retained within the staff andtransmitted to the Board in a general way, and atmanagement’s discretion.

Appendix V Confidential Exchange:An Elaboration

Note by Chairman John Crow

92

Part 3

Statement by the Managing Director on the Report of External Evaluators on Fund Surveillance

Executive Board Meeting, August 27, 1999

Staff Response to the External Evaluation of Fund Surveillance

1. Management has been asked by the Chairman ofthe Evaluation Group of Executive Directors tocomment on the External Evaluation of Surveil-lance. We offer our views as a contribution to the in-depth debate the Executive Board intends to have asit seeks to strengthen the process of Fund surveil-lance. We welcome the report, and express our ap-preciation for the careful work and considered judg-ments of the panel. We will discuss the majorrecommendations of the report, but will not com-ment on other details of the text, beyond noting ourgeneral agreement with most of it. We note also thestaff ’s comments, which in our view constitute aconsidered response to the evaluation, presenting de-tailed reactions to and commentary on the individualrecommendations of the evaluators. Our views differat most in the nuances from those of the staff.2. In paragraph 2 of their report, the evaluators use-fully define six goals of surveillance: (i) policy ad-vice; (ii) policy coordination and cooperation (amongcountries—but also, it becomes clear from the text,within countries); (iii) information gathering and dis-semination (to member countries and the public); (iv) technical assistance (particularly in supplyingmacroeconomic expertise to smaller, developingcountries); (v) identifying vulnerabilities (especiallyto governments, but also perhaps in the course of timeto markets); and (vi) “delivering the message,” that is,disseminating professional conclusions on policymatters to members, an extension of the advice role.Of these, we regard the first goal, policy advice, andthe third, information gathering and dissemination, asprimary, and the others as derivative, or as implied bythe policy advice and information roles.

We have ourselves been struck in our discussionswith many members how much they value not onlythe Fund’s surveillance of their own economies, butalso the information on the major economies and theworld economy that they receive from the surveil-lance process; for countries not in the OECD, thegreat bulk of the membership, Fund surveillance pro-

vides their only systematic window on the worldeconomy, and they value it highly. They value, too,the opportunity provided by Board discussion to com-ment on and seek to influence global developmentsand policies in the leading industrial countries. 3. The evaluators inject a valuable note of realismin their discussion of the impact of surveillance on acountry’s policies, for instance in paragraphs 5 and 6of Chapter V of their report. Outside the context ofprograms, Fund advice should be viewed as “aninput that could on occasion be significant, depend-ing upon the stage of the domestic policy debate”(Chap. V, para. 7). Similarly, to the extent that Fundsurveillance documents are made public, they willfrequently be only one voice among many comment-ing on the global economy or particular countries—though their relative importance increases for thesmaller and systemically less important countries.These observations emphasize the necessity to en-sure that the Fund’s analysis is first rate.4. The report was written at a time when surveil-lance was undergoing major changes within theFund, largely as a result of the crises of the last fiveyears. Thus, as noted in both the evaluation itselfand in the staff ’s comments, some of the recommen-dations, including those relating to enhancing sur-veillance of the financial sector, capital account is-sues, and policy interdependence and contagion, arealready in the process of being implemented. Simi-larly, the experiments on publication of Article IVreports will help the Board decide whether to acceptthe evaluators’recommendations on publication ofArticle IV reports. Management supports publica-tion, while recognizing that we need to find a way ofpublishing that takes account of valid concerns overthe implications for the frankness of discussionswith the staff, and possible market reactions.

We will next comment on the main recommenda-tions of the report.5. The authors recommend that wegive greater em-phasis to aspects of analysis that the Fund is particu-

Statement by the Managing Director on theReport of External Evaluators on Fund Surveillance

Executive Board Meeting, August 27, 1999

95

STATEMENT BY THE MANAGING DIRECTOR

larly well equipped to provide: (1) aspects of the in-ternational financial situation most relevant to themember country; (2) relevant experience from othercountries; and (3) how the member can best absorbwhatever shocks might emanate from abroad (i.e.,how to deal with vulnerabilities). This—including theemphasis on the analysis of vulnerabilities—is an ex-cellent idea, which deserves greater emphasis in oursurveillance activities. 6. Greater focus in Article IV consultations on theFund’s core issues,macroeconomic and exchangerate policies, plus the financial sector. This recom-mendation runs counter to most of the pressure on usfrom the Board and the membership, who havegiven increasing emphasis to the interactions be-tween macroeconomic and structural policies, rang-ing even beyond the financial sector. The evaluatorsbase their recommendation in part on evidence thatFund advice outside the core areas is not regarded bythe authorities as commanding the same authority asthat on macroeconomic issues. The staff presents acareful response to the recommendation for morefocus; we conclude, as they do, that while the coreissues should continue to be at the center of the sur-veillance process, other areas will—depending onthe country—have also to be included in the surveil-lance process. At a minimum, social sector issuesand poverty will have to be discussed in most ArticleIV reports. In order to make surveillance in theseareas more effective, the Fund will have to furtherstrengthen its cooperation with other institutions,primarily of course the World Bank.7. Article IV consultations for small and medium-sized industrial countries should be less frequentthan annual.We recognize that the relative valueadded by the Article IV consultation to the policy-makers in these countries, most of which are mem-bers of the OECD, is probably less than that in de-veloping countries that receive less attention fromother agencies and financial market participants.However, we agree with the comments of the staffon this recommendation and add one more factor:that much of the strength of the Fund as an institu-tion derives from the symmetry with which coun-tries are treated in the surveillance process, witheach country having the right to comment on eco-nomic policies in every other country. We will needto consider the possibility of less frequent Article IVconsultations for small and medium-sized industrialcountries on a case-by-case basis, taking into ac-count both the country’s views on the desirability ofannual surveillance for itself, and the views of othercountries on the benefits of such surveillance. 8. Regional surveillance should receive more em-phasis. We welcome the suggestions in this area,and accept responsibility for the fact that staff pre-sentations in country matters sessions do not have a

regional focus: management has hitherto taken theview that given the particular orientation of thesesessions, it would be more appropriate to concen-trate on country-level problems. There are twosenses in which the term regional surveillance isused: first, that common policy issues and spillovereffects should receive more attention; and secondthat, as in the EU, analysis should be focused moreon region-wide rather than national aggregates andpolicies. We agree that regional surveillance in thefirst sense deserves greater attention; and take up thecase of the EU next.9. The panel recommends that surveillance of EUcountries be carried out mainly at the EU level,as isnow being done for monetary policy for the EMUcountries. We believe that surveillance for EU coun-tries will move toward the EU level over the courseof time, as other aspects of decision making move inthat direction. Meanwhile, as we maintain country-level surveillance for EU countries, we need to con-sider whether too many resources are being devotedto that effort. 10.The panel recommends quarterly publication ofthe World Economic Outlook (WEO).We do not sup-port this suggestion, which would be intensive instaff and Board time, preferring rather the approachtaken so far of bringing out a special issue of WEOwhen needed. We could, however, support more fre-quent (than annual) publication of capital marketsinformation, including possibly through expandingthe quarterly report on capital flows sent to theBoard by the Research Department. On the issue ofWEO projections, we see great value in the presentiterative process in which the area departments andthe Research Department evolve towards a mutuallyacceptable view. (We note for the record the high re-gard for the WEO and International Capital Marketsreport shown by the evaluation group.) 11. The review process. This problem is discussedalso in the evaluation of research. We have under-taken several internal reviews of the review process.The review process is an essential contributor to thecreation of a coherent Fund position on issues. How-ever, we do need to find a way of reducing theamount of resources devoted to it; that can be donein part by ensuring that departments concentrate ontheir own areas of expertise, with “no comment” be-coming a fully acceptable and frequent comment.12.Role of the Board. While this is by and large amatter for the Board to consider, we have the follow-ing three observations.

• In making the recommendation for a committeestructure, the evaluators do not take account ofthe varying levels of participation of each chairat different meetings. The Executive Directorsor their alternates tend to take part in the most

96

Statement by the Managing Director

important meetings, or those of particular inter-est; advisers and assistants take part in othermeetings. This arrangement ensures the benefitsof the committee approach, without removingthe right of individual countries to take part inall discussions.

• For several reasons, most practically given timepressures, we would be reluctant to bring brief-ing papers to the Board for discussion; one way

for the Board’s concerns in a particular countryconsultation to be brought to the attention of thestaff is for the Executive Director to ask col-leagues before each mission whether there areissues that in their view deserve special atten-tion, and then inform the staff.

• The experiments now under way to increasethe efficiency of Board discussions could alsohelp focus surveillance.

97

1. The evaluators have produced a candid and com-prehensive report. It is realistic about what surveil-lance can—and cannot—achieve, and it provides anhistorical overview of surveillance at the Fund thatis informative even for knowledgeable readers. 2. The evaluators’main findings—summarized intheir executive summary—have much in commonwith the views of Fund staff. Many of the prescrip-tions of the report—notably, better regional surveil-lance and a stronger international dimension to bilat-eral surveillance; and greater emphasis on financialsector and capital account issues, external vulnera-bility, policy transparency, and the dissemination oftimely information—are already on the Fund’sagenda. Staff broadly agree with the thrust of therecommendations in these areas.3. Staff welcome the largely positive evaluation ofthe conduct of both bilateral and multilateral sur-veillance in core areas. Staff also appreciate thequestions raised about the institution’s capacity todeal adequately (in the context of bilateral surveil-lance) with nonfinancial structural issues and the re-lated concern that the expanding coverage of theseissues could detract from the overall effectivenessof surveillance. The focus of surveillance is an issuewith which the staff have also been grappling, andwe share many of the concerns raised by the evalua-tors. Improving work on nonfinancial structural is-sues, strengthening financial sector and capital ac-count surveillance, and incorporating more of thevarious international and regional dimensions of acountry’s macroeconomic situation into the Fund’sbilateral surveillance without reducing the qualityof surveillance pose a major challenge—especiallyin light of the increasing demands on staff fromother areas of work and stringent resource con-straints. Nevertheless, most staff recognize the ana-lytical necessity of the broadening of surveillance tocover certain areas that are critical to assessingmember countries’situations. To focus bilateral sur-veillance solely on the exchange rate and directlyassociated macroeconomic policies, as the reportrecommends, risks missing important issues andmisreads the needs and demands of the Fund’smembership.

4. On methodology, many staff members wouldhave liked to see more space devoted to the method-ological underpinnings of the work leading to theconclusions and recommendations. Relatedly, thereare comments and suggestions that arose from theinterviews that staff found particularly thought pro-voking—for example, concerning the forecastingrecord of departments, the use of cross-country com-parisons in bilateral surveillance, staff advice on fis-cal consolidation, and the emphasis on first-best so-lutions. However, these were based on limitedobservations and it was difficult to gauge their rele-vance to the Fund’s work more broadly. The pointsadvanced in the report on these issues/merit moresystematic examination, including in the context ofthe Fund’s internal work.5. On process, the staff read with interest many ofthe observations on the Fund’s internal organization.Many of the proposals in the report, including thoseconcerning a shift of the responsibility for WEOprojections to the Research Department, the internalreview process, and the role of the Board, have im-portant implications for the Fund’s internalprocesses and would need to be explored further be-fore these proposals can be properly evaluated. Thisview is revisited in some of the sections below. 6. The response that follows focuses on severalmajor themes and recommendations. The magnitudeof the report dictates a need for selectivity in this ef-fort. Thus, we touch upon issues related to narrow-ing the focus of surveillance; the conduct of regionalsurveillance; surveillance of the industrial countriesand the euro area; shifting the responsibility forWEO projections to the Research Department(RES); the review process; the role of the Board; andtransparency and Fund surveillance. A first annexprovides extensive comments from RES on theWEO and the ICMR and the Department’s analysisof capital flows; the staff in general associates itselfwith these comments. A second annex provides morefactual and detailed comments from individual de-partments on a variety of other points.7. This note does not comment on the interestingobservations of the report on standards and the Con-tingent Credit Line: discussions on these issues con-

Staff Response to the External Evaluation of Fund Surveillance

98

The Focus of Surveillance

tinue at the level of the staff, management, and theBoard; policy and practice are undergoing rapidchange, and thus the report’s observations, and staffresponses, would almost certainly soon be out ofdate.8. In general, the view of the staff is that the re-port of the evaluators provides an informed outsideperspective that should serve as an important inputto the Fund’s deliberations on further strengtheningsurveillance.


9. There has been much debate, including in thepublic domain, over the broadening of the surveil-lance agenda in recent years. The call for increasedfocus in the Fund’s surveillance is not new and re-mains a challenge for the organization in light of theforces driving an expanding agenda. In its review oflessons from the Asian crisis for Fund surveillance,Executive Directors underlined that the focus of sur-veillance needed to extend beyond the core short-term macroeconomic issues, while remaining appro-priately selective.1 In this regard—bringing anappropriate focus to surveillance—there are impor-tant areas of agreement between the report and theproposals to bring more focus to surveillance pre-sented in the 1998 staff paper “Lessons for Surveil-lance from the Asian Crisis” and agreed by the Exec-utive Board:

• Surveillance should pay more explicit atten-tion to capital account/financial sector/conta-gion issues and external vulnerability.

• Surveillance at the country level should paygreater attention to policy interdependence andthe risks of contagion.

10.Notwithstanding the identification by the Fund’sinternal reviews of the need for more attention tothese issues in Fund surveillance, we agree that im-plementation of these general principles has takentime. This can be explained, in part, by limited re-sources and, in some areas, expertise.11. The staff find the demands arising from the sheerscope of surveillance issues to be challenging, par-ticularly in areas necessitating new expertise. We arealso well aware that the broadening focus of Fundsurveillance has significant resource costs. However,the staff recognize the analytical necessity of thebroadening of surveillance to cover certain areas ifsurveillance is to continue to be relevant. In this re-gard, the staff doubt very much that the Fund could

be more effective by focusing on core macroeco-nomic and financial issues to the exclusion of struc-tural issues. Selectivity, yes, but to focus only onmacroeconomic and financial issues misreads theneeds and demands of the Fund’s membership.12.As the report correctly points out, there is sub-stantial pressure from the international communityand shareholders to bring additional dimensions tothe surveillance process. Following the Mexican cri-sis of late 1994–95, there were calls by the interna-tional community for more intensive treatment ofmembers’financial sectors in Fund surveillance. Inthe immediate aftermath of the Asian crisis, the sig-nificance of transparency, data dissemination, finan-cial sector stability and hence standards, and theneed to extend beyond the core short-term macro-economic issues have received increasing emphasis.While the report acknowledges the forces that havedriven the broadening definition of “core issues” inFund surveillance, it does not examine these forcessquarely in coming to its recommendations on thefocus of surveillance. A recent Board discussion onMexico is illustrative in this connection for its heavyemphasis on progress in social policies over recentyears, as is the June 1999 G-7 Communiqué callingon the Fund to pay greater attention to social sectorissues.13.Staff particularly emphasize the following pointson the focus of surveillance:

• The report takes exception to the Fund’s in-volvement in a number of areas that are con-sidered to be outside the Fund’s statutory man-date, including trade liberalization, labormarkets, offshore banking supervision, tax re-form, expenditure streamlining, military out-lays, and income distribution. While staffwould agree that the Fund should remain pri-marily a macroeconomic institution, theFund’s Articles of Agreement and decisions ofthe Executive Board2 suggest a far broadermandate than the authors seem to have inmind.

• We believe that another key reason for the ex-panding scope of surveillance beyond tradi-tional macroeconomic analysis is the realizationby policymakers, international organizations,and the economics profession as a whole thatthere is a wider set of issues (microeconomic orstructural in nature) that must be consideredwhen analyzing the international monetary sys-tem, stabilization, medium-term sustainability,

99

1See IMF, Annual Report 1998, “Strengthening of the Architec-ture of the International Monetary System,” Chapter 7.

2See Articles I and IVand paragraph 3 of “Principles of FundSurveillance over Exchange Rate Policies” set out in ExecutiveBoard Decision No. 5392–(77/63), adopted April 29, 1977, asamended.

STAFF RESPONSE TO THE EXTERNAL EVALUATION OF FUND SURVEILLANCE

and growth—all key areas of Fund surveillance.Even if greater focus is achieved, a good macro-economic analysis needs to view the macro-economy in the broader context of the struc-tural, social, political, and environmentalsetting. To be relevant in Western Europe, forexample, Fund surveillance must be able to ad-dress labor market issues; in Central and East-ern Europe, privatization and enterprise restruc-turing; in Africa, civil service reform; and so on.Of course, the staff accepts that its attention tothose areas should draw on the best work avail-able outside the Fund.

• Relatedly, the Fund has come to learn the hardway that, over the longer run, the sustainabilityand viability of reform policies depend criti-cally upon the large number of factors that af-fect growth. These include, among otherthings, governance, the composition of expen-ditures, and poverty. Many of these may notfall in the class of issues that the report wouldconsider “directly relevant” to macro policies;yet, as experience shows, they can prove criti-cal over the longer term.

• The Fund’s membership covers a heteroge-neous set of countries. What is a core issuemay differ across countries. In a number ofcountries, structural issues outside the macro-economic and financial areas, characterized bythe report as “noncore,”3 are nonetheless at thecore of the problem. It would thus be odd if theFund were to sidestep these issues.

• On the need for greater attention to capital ac-count issues, the staff have been providinganalyses of capital flows for many years, al-though attention intensified following theMexican crisis and with the increasing integra-tion of capital markets.4 The staff, like others,have a learning curve.

• As the scope of surveillance has broadened,the staff are increasingly drawing on the exper-tise of other agencies where appropriate andfeasible. Examples of this approach includethe recent reliance on OECD work in framingcertain elements of Fund advice to countriessuch as Korea and the recent initiatives tostrengthen Bank-Fund collaboration in finan-cial and social sector work. As the report ac-knowledges, in some noncore areas, the Fund

has to step in, not so much as the chosen in-strument, but rather the only presently avail-able instrument. This issue has also come tothe fore in the discussion of transparency re-ports, where some working knowledge on thepart of Fund staff in noncore areas has beenfound desirable.

• The recommendation to focus on the “core” ofsurveillance, as the authors see it, leaves unan-swered how this would affect the content ofFund-supported financial programs. If Fundstaff did not have the opportunity to build theirintellectual capital in “noncore” areas in thecontext of surveillance, would they still be ex-pected to develop and apply conditionality re-lating to such issues in subsequent Fund-supported programs? How would “noncore”work currently carried out by the Fund beshared with other institutions?

14. For the reasons indicated above, and although weare in agreement on the appropriateness of containingthe scope of Fund surveillance, it is not easy to seehow a return to the traditional core areas could be ac-complished in the current international context. Moststaff believe that if surveillance is to continue to berelevant, its scope will need to change in response tothe evolution in the world environment. What may berequired is to sharpen the focus of surveillance in theindividual case (avoiding a “shopping list” approach),allowing sufficient flexibility to mission chiefs to de-termine what are the core issues in each countrywithin the Fund’s broad and evolving interests. How-ever, there are trade-offs in all of these approachesthat might usefully be noted: a narrow focus on the“core” risks missing important issues; a “shopping-list” approach is a recipe for superficiality; and acountry-by-country approach to focus raises issues ofconsistency of treatment across countries. In light ofthe above, the staff see the need for further examina-tion of these trade-offs in the Fund’s surveillance ac-tivities, including in internal reviews.

The Fund’s Conduct of RegionalSurveillance

15.The Fund’s approach to regional surveillanceemerges as an important theme, and staff agree thatregional surveillance should receive greater empha-sis. In this regard, the Fund has increased its in-volvement in regional surveillance in recent years, inpart linked to initiatives among member countriesthemselves, and some additional aspects of this in-volvement are worth noting. 16.The Fund has devoted increasing attention tosystematic participation in regional surveillance

100

3Apart from occasional references to poverty issues, the reportdoes not specify what is meant by “noncore” structural issues; wehave assumed the term refers to all structural issues other thanthose related to the financial sector and the capital account.

4Annex I provides instances from multilateral surveillance doc-uments that were available to the public.

Surveillance of Industrial Countries and the Euro Area

processes and to the development of these processes,not only in the euro area, but also in Asia, Africa,Latin America and the Caribbean, and the MiddleEast. In this regard, the report seems overly focusedon Europe as a model for regional surveillance. Asthe report notes, the fact that other regions—such asAsia—are not as integrated as Europe, and will mostlikely remain so, suggests a different role—onebeing explored by the Fund—for regional surveil-lance as follows:

• Regional surveillance in Asia is implementedthrough a range of channels. The Managing Di-rector participates in, and the staff prepare back-ground information for, meetings of ASEAN Fi-nance Ministers. The Fund has been designatedas the technical secretariat of the Manila Frame-work Group that was established specifically toundertake macroeconomic surveillance. Thestaff also work with various arms of ASEANand other regional groups. The establishment ofthe IMF’s Tokyo office was an explicit recogni-tion of the growing role of Asia, and workingwith various regional groups as part of surveil-lance is one of its major tasks.

• In the African region, the Fund has establishedperiodic discussions with WAEMU, theBCEAO, and BEAC as a backdrop for bilateralconsultation discussions with relevant mem-bers. Through the vehicle of cross-country ini-tiatives, the Fund is playing an active role in as-sisting regional organizations in Eastern andSouthern Africa to develop their capacity andpolicy focus in a harmonized way.

• Spillover effects of last year’s crisis in Russiahave been a focal point of most missions totransition economies in Central and EasternEurope and the Commonwealth of Indepen-dent States (CIS) recently.

• The Fund has increased its involvement withthe countries of the Gulf Cooperation Council(GCC) in the context of a regional framework.Recently, in the wake of the decline in oilprices, a framework of semiannual consulta-tions with the GCC countries at the regionallevel has been established.

• Beyond the reference to the Fund’s lead inSeptember 1998 in convening Western Hemi-sphere finance ministers and central bank gov-ernors to discuss common issues, includingcontagion effects of the crises in emergingmarkets, staff have been exploring with mem-ber countries in Latin America ways of en-hancing regional collaboration.

• Finally, surveillance of the Organization ofEastern Caribbean States in the Caribbean re-

gion includes a role for the Eastern CaribbeanCentral Bank—the regional monetary author-ity—again, as the backdrop for bilateral dis-cussions with members.

17.The Fund’s engagement in many of these areas isrelatively new and to a degree experimental, and isincreasingly seen as a means of coping with evolv-ing regional arrangements. Staff agree that moreneeds to be done and that country surveillanceshould emphasize regional and spillover issues andbring to bear more of a cross-country perspective.

Surveillance of Industrial Countriesand the Euro Area

18.The report proposes to reduce surveillance ofsmaller industrial countries, to shift euro area sur-veillance to the EU, and to focus more on the inter-national implications of the largest industrial coun-tries’ domestic policies and correspondingly less onadvice regarding domestic policies per se. Whilestaff can see the merit of some reallocation of re-sources away from industrial countries generally,we have reservations about the implications of mov-ing too far in this direction for the effectiveness ofsurveillance.19.The introduction of a two-tier surveillance struc-ture in which smaller industrial countries would bevisited less frequently raises questions about the ef-fectiveness of surveillance:

• The main effect of the shift to a two-year con-sultation cycle would be to reduce the Board’sinvolvement—which runs against the evalua-tors’ proposals elsewhere in the report to in-crease that involvement. At present, countrieson two-year cycles often have interim discus-sions that resemble Article IV discussions, ex-cept with regard to the documentation pre-pared for the Board.

• Implementation of the recommendation thatsurveillance of individual participants in theeuro area should largely take place at the euroarea level and through EU institutions would,at this stage, detract from the quality of theFund’s surveillance of individual participantsin the euro area. Indeed, the recommendationsits oddly with the report’s earlier acknowledg-ment of the relative candor of bilateral discus-sions and the reluctance of countries to engagein frank exchange and criticism in regional fo-rums. The proposal is also not consistent withmembers’obligations to collaborate with theFund in the context of bilateral Article IV con-sultations, and would largely remove Fund sur-

101


veillance from the effective levers of policy.Talking about the euro area’s fiscal policywould be to no avail in the absence of discus-sions about policies in individual countries.

• Similarly, we question the prudence of movingsmall and medium-sized industrial countries totwo-year consultation cycles, as the proposalseems to assume that these countries will nothave a crisis, or, if they do, they can take care ofthemselves. This proposal seems at odds withthe report’s emphasis on strengthening the roleof surveillance in identifying vulnerabilities andrisks of financial crises. If the Fund had adoptedthis recommendation earlier, it certainly wouldhave put into this group some countries thathave faced serious crises in the past decade. Theexperience of the Asian crisis is a reminder thattoday’s star economic performers can becometomorrow’s crisis countries.

• The fact that standard surveillance has not pro-duced adequate warnings might be seen as anargument for strengthening, not weakening,consultation procedures. Published data andanalyses complement (and provide importantinput for) the direct policy dialogue, but theyare not a substitute. As the experience in somecases has shown, these sources can sometimesbe quite misleading.

20.Staff are also skeptical about the scope for anddesirability of cutting back resources for industrialcountries, the room for savings in this area, and thesuggestion that the focus of surveillance for thelargest countries be mainly on the international im-plications of their policies.21.Specifically:

• In the staff ’s view, the evaluators are undulypessimistic with regard to the usefulness of theFund’s surveillance of industrial countries. TheFund’s value added lies importantly in provid-ing a disinterested outsider’s analysis of na-tional issues from a global perspective. Neitherfinancial markets, the press, nor academic re-searchers provide this type of information, atleast not on a regular basis. Examples includethe analyses of German unification and the re-cent issues facing Japan.

• A policy of less frequent surveillance of indus-trial countries would need the endorsement oftheir authorities. Only a few such countries havemade use of the voluntary option of a two-yearconsultation cycle, in the context of “selectiv-ity” endorsed by the Board in April 1997 andmade operational at the end of that year.

• The idea that the objectives of surveillance varyby type/size of country is contrary to the Fund’s

role as a global institution, and to the principleof uniformity of treatment that is essential to theFund’s relations with its members.

• It is not accurate that industrial countriesnever get into trouble and therefore have littleto gain from “full” surveillance. The exampleof Japan is an obvious one, as are the periodicexchange rate crises in Europe in the lastdecade. We also believe that surveillance hashelped intensify fiscal consolidation efforts ina number of countries, including Belgium,Italy, and Sweden.

• The evaluators recommend that fewer re-sources be devoted to industrial countries, butthat the WEO be published more frequently;yet the latter would depend critically on moreregular examination of developments in indus-trial countries.

Shifting the Responsibility for WEO Projections to RES andPublication of Quarterly WEO Forecasts

22.The report proposes to shift the ultimate respon-sibility for WEO projections to RES. Staff are con-cerned that the report downplays the complementar-ities that exist between bilateral and multilateralsurveillance that would be weakened by such ashift. Staff also view bilateral surveillance as thebedrock of the Fund’s work and the basis for multi-lateral surveillance. While we share some of theconcerns of the evaluators about the ownership ofthe WEO forecasts, the report appears to have un-derplayed aspects of the Fund’s internalprocesses—including informal as well as formalmechanisms to air and resolve differences of viewin this area. However, while most departments feelthat the interdepartmental mechanisms for coordi-nating the projections generally work well, the Re-search Department view is that in selected instanceswhere significant differences between the staff ofthe Research and area departments have emerged,the process of resolving these differences has notworked as well as it should have.23.The following specific issues were raised on thistopic:

• Staff are concerned that if RES were to havefinal responsibility for projections, it wouldimply that the country desks would have noownership of the IMF projections for theircountries. Most staff see this as unworkable.

• To allow different country forecasts to be usedin WEO, on the one hand, and in bilateral sur-

102

Review Process

veillance and program work, on the other,would be confusing to the public and wouldlimit the scope for country-specific knowledgeto be incorporated into the WEO forecasts.

• Staff acknowledge, however, that the currentownership of projections needs to be clarified,as some departments view projections as fore-casts of the area departments only, whereasother departments see them as forecasts of theentire organization—a situation that at times hasled to conflicting presentations to the public.

• RES shares the concern of other departmentsthat a full shift of responsibility would posesignificant operational and other difficulties,including less careful preparation of forecastsby area departments if they no longer havethe final responsibility; and the large re-source implications of the proposal. How-ever, RES also feels that there have been seri-ous problems that have occasionally arisenwith the WEO forecasts that deserve the at-tention of management and the ExecutiveBoard.

• On interdepartmental interactions, there is al-ready much more consultation with RES thansuggested by the report. The extent to whichthe Fund’s policy assessment of the situationof the major industrial and developing coun-tries is developed collectively, with input fromfunctional as well as area departments, is un-derplayed.

• Most staff see a degree of tension in interde-partmental relations as an inevitable andhealthy part of the multilateral surveillanceprocess. While there may have been problemson occasion in the past, affecting capital mar-kets work as well as the WEO, these are over-stated in the report, and have not gone as far asa breakdown in communications. Moreover,where difficulties have occurred, considerableefforts have been made to ensure closer coor-dination between departments.

• Staff note the inevitable tension also betweenglobal assessments and forecasts underlyingFund programs, given the difficulty in assum-ing that Fund programs might fail, eventhough some inevitably will. Shifting the re-sponsibility for forecasts will not resolve thesetensions, which could be better dealt withthrough alternative scenarios analyzing down-side risks.

• Staff of RES feel that the evaluators havemissed some important points about how in-formation about the Fund’s forecasts is con-veyed to the Executive Board. When there are

significant differences between the ResearchDepartment and the area departments over theWEO forecasts, the Executive Board is usuallyinformed of these differences, at least in quali-tative terms, in the course of the WEO orWorld Economic and Market Development(WEMD) sessions.

• Finally, staff do not support a shift to quarterlyWEO forecasts. Such a shift would imply asubstantial increase in the staff ’s workload, po-tentially compromising the analytical contentof these reports (which is a critical aspect oftheir comparative advantage). We also con-sider it important to retain some independencebetween the WEO and ICMR, and, when war-ranted by circumstances, ad hoc WEO/ICMRexercises could be produced to ensure thatcoverage of world developments is appropri-ately current (as was done in the early 1980sand also last year).5

Review Process

24.The evaluators provide some interesting obser-vations on the review process. Staff recognize thatthe review process is resource-intensive, and thatconstant effort is required to manage it effectively.However, staff see this process of internal criticismand peer review as playing a critical role in enhanc-ing the surveillance effort and improving the finalproduct. Thus, most staff would like to see effortsfirst focused on addressing specific problems thathave been identified and ensuring that the process istaken seriously, rather than unduly curtailing the re-sources devoted to internal review. On this issue,staff emphasized the following points.

• Most staff (including both functional and areadepartments) are not in favor of more infor-mality in the review process. There is alreadya great deal of informal give and take that isundocumented. More generally, in our view,moving away from written to oral feedbackmight serve to lessen responsibility, impairinstitutional memory, and make for less-considered comments.

• While replacing formal comments with e-mails and phone calls may streamline the re-view process, it is important that all depart-ments remain aware of each other’s comments.

• Replacing formal comments with meetingswould likely only increase the burden on re-viewing documents: staff from reviewing de-

103

5See Annex I for an elaboration on these issues.


partments would still need to write commentsto brief those participating in the meetings.6

Also, from the recipient perspective, meetingsare often much more time-consuming thanwritten comments.

• Other reactions were also registered by staff,including that some of the criticisms of the re-view process are already being addressed bymanagement guidelines; that cutting review re-sources would have to be balanced against thebenefits of the process (such as better checksand balances and better integration of technicalassistance activities with surveillance); andthat effective solutions would require delvingmore deeply into the role of reviewing depart-ments. This is an area to be pursued.

Role of the Executive Board andProposals for Restructuring

25.Staff read with interest the proposed changes inthe role of the Executive Board through the intro-duction of a two-step briefing process and of a com-mittee structure. Although it was clearly understoodby all that the Executive Board decides how to struc-ture its work, in our view there are key aspects of theFund’s internal processes—and channels of interac-tion—that cumulatively and collectively raise issuesto the Executive Directors in a timely fashion. Webelieve that there are important countervailing argu-ments, particularly concerning the resource implica-tions of the report’s proposals. The following spe-cific comments elaborate on the general thrust ofstaff views.

• Effective procedures already exist for the Ex-ecutive Board to express its views and provideguidance, for example through the regular con-sultation cycles and use of Fund resourcesmeetings (whose policy conclusions are regu-larly referred to and followed up in subsequentArticle IV discussions), country matters ses-sions, and WEO, ICMR, and WEMD sessions.

• The two-step briefing process potentially cre-ates significant additional work both for theExecutive Board and for the staff, especially ifconsidered in conjunction with the proposalsto implement quarterly WEO/ICMRs andmore frequent and continuous surveillance.

• Resource savings, if any, from the committeeapproach would most probably come at the ex-pense of the smaller countries, which would bethe ones most likely to be discussed solely atthe committee level, creating a “dual track” ofsurveillance that would undermine the globalnature of the present system and tend to intro-duce bias against small states.

• The new regionally based committee structuremight involve increased resource costs giventhe need for greater coordination between theBoard, committees, and the staff. In addition, itis not unlikely that many countries would bediscussed by the full Board.

• The new structure could tend to undermine theprocess whereby information is transmittedthrough Executive Directors to their capitals,and would require difficult and time-consumingbalancing so that participation by Directors oncommittees was appropriately representative.

• Despite the safeguards proposed by the evalua-tors, pressures toward “clientism” and a “re-gional perspective” could only increase rela-tive to the present system, amplifying andinteracting with the preexisting biases that theevaluators identify as operating at the area de-partment level.

Transparency and Fund Surveillance

26.On a final note, while supporting publication ingeneral, we are somewhat less sanguine about thesuggestion that “publication of staff reports shouldprovide important support over time for greaterfrankness.” The evaluators observe an opposite re-sult in OECD reports and the danger that policy dif-ferences could be smoothed over in the final reportis something that would need to be guarded against.While there are strong advantages to transparencyand accountability in publishing, it is not enoughsimply to assume away the problem that publishingmay reduce some country authorities’willingness tospeak candidly about policies. In fact, the pilot pro-ject for the release of Article IV staff reports is in-tended to try to address this and other issues of con-cern to Directors, and the staff would not want toprejudge the outcome of this experiment.27.Relatedly, staff are unpersuaded by the viewthat, because Article IV reports already circulate tosome degree outside the circle of authorized users(a tendency that some thought the report exagger-ated), they should be available to everyone. It istrue that the general publication of Article IV re-ports might help to create a level playing field for

104

6As one department put it: “One of the virtues of the Fund, incontrast to other bureaucracies, is the relatively small amount oftime spent in meetings. Certainly meetings can be important forsignificant and very controversial issues, but in our experiencethe balance seems about right.”

Transparency and Fund Surveillance

all users; unauthorized leaks, even if widespread,allow unfair access to those benefiting from theleaks. However, the report’s proposal—that theFund publish the staff report for each Article IVconsultation whether or not the relevant memberconsents—raises several legal issues, as the reportacknowledges, that are not so easily tackled. First,the Fund may not publish a staff report disclosingconfidential information without the member’sconsent. Second, even where confidential informa-

tion would not be disclosed, the Fund’s Articlespermit the Fund to publish its views (e.g., on amember’s policies) without the relevant member’sconsent only if they deal with “monetary or eco-nomic conditions and developments which directlytend to produce a serious disequilibrium in the in-ternational balance of payments of members.” Fi-nally, a special majority of the Fund’s ExecutiveBoard (i.e., 70 percent of the Board’s total votingpower) must vote for publication of such reports.

105

The Role of the Research Department

The evaluators clearly express a generally favor-able assessment of the Research Department’s con-tributions to Fund surveillance, especially to multi-lateral surveillance, where RES has the leadresponsibility. Of course, we appreciate this favor-able evaluation. Nevertheless, we have two specificconcerns about what the evaluators recommend andsay in their report.

The WEO and ICMR

First, while we appreciate the implicit complimentin the recommendation for quarterly publication of acombined WEO/International Capital Markets re-port, we have grave reservations about this sugges-tion. An essential element of the value added of theICMR is that it brings a “capital markets perspec-tive” to important multilateral surveillance issues,meaning that its focus is on the role of financial mar-ket participants, institutions, and regulations. For ex-ample, the analysis of the ERM (exchange ratemechanism) crisis provided in Part I of the 1993ICMR provides such a perspective on the factors in-volved in the ERM crisis—a perspective differentfrom the macroeconomic perspective on the crisisthat is presented in the WEO, focusing on the role ofmacroeconomic developments and policies. Neitherperspective contains the whole truth; both are rele-vant and valuable. Similarly the ICMR perspectiveon the Asian/Russian/Long-Term Capital Manage-ment crises is different from, and adds considerablevalue to, the WEO perspective on these crises.

Especially in an institution that places heavy em-phasis (some would say, excessively heavy emphasis)on macroeconomic analysis, combining the WEO andthe ICMR on a regular basis would, we believe, tendto undermine the analytical independence of theICMR and would, over time, undermine its uniquevalue added. Moreover, an essential element in thepreparation of the ICMR is the system of staff mis-sions to key financial centers and to a variety of

emerging market countries to gather information,analysis, and opinions (especially from private marketparticipants) on the issues to be covered in the report.It would be very difficult and disruptive to attempt tomake this system of missions conform to a quarterlyschedule of publication, while still leaving time forthe missions to be planned and executed and for thereport to be thought out, written, reviewed, consid-ered by the Executive Board, and published.

These concerns do not preclude the possibility ofsemiannual updates of the ICMR, perhaps some-times combined with interim updates of the WEO.Last December’s combined WEO/ICMR is a usefulexample of how this can be done, in consultationwith the Executive Board, when developments in theworld economy and global financial markets appearto warrant it. But the genesis of such reports wouldbe best determined on an ad hoc basis. Beyond this,the plan is to expand gradually the analytical contentof the quarterly private financing notes that are nowdistributed to the Executive Board (but not pub-lished) and to take up discussion of issues raised inthese notes, as may be needed, in the WEMD ses-sions. This, of course, will not satisfy demands fromoutside the Fund for more frequent updates on inter-national capital market developments and issues, butthose folks are not paying the bills for the exercise.

Regarding the frequency and schedule for theWEO, it must be recognized that the WEO plays acentral role in providing material for meetings of theInterim Committee and, correspondingly, the presentsemiannual schedule for the WEO reflects the sched-ule of meetings of the Interim Committee. Typically,there are nearly seven months between the late-September/early-October and the late-April meet-ings of the Interim Committee and, correspondingly,there are barely five months between the late-Apriland the late-September/early-October meetings.Given the usual schedule for preparation of theWEO, it is essentially inconceivable to think of aninterim WEO during the summer. Even allowing forsome compression of the normal WEO schedule (assuggested by the evaluators), an interim WEO pre-sented to the Executive Board in late July or earlyAugust, followed by a full WEO presented in early

Annex I Excerpts from Comments fromthe Research Department

106

Analysis of Capital Flows

September, makes no sense. (As far as the ExecutiveBoard is concerned, the WEMD sessions can beused, and have been used, to fill in the gaps.) Thefact is that the primary value added by the WEO, rel-ative to information and forecasts produced in theprivate sector, is not rapid revisions in assessmentsof the outlook, but rather timely and relevant analy-sis of the economic forces shaping the outlook andof the risks to the global economy. Too high a fre-quency would destroy its fundamental value.

When something of global economic significancehappens in the late summer or early autumn, it is fea-sible to produce an interim WEO in December orearly January. This was demonstrated in the cases ofthe ERM crisis, which broke in mid-September1992, as well as in the Asian and Russian/LTCMcrises of 1997 and 1998. It would be possible to planfor a regular edition of an interim WEO or interimWEO/ICMR at every year-end. However, with pres-sures on staff resources already beyond the sustain-able level, and with the need to plan, and think, andresearch to sustain the quality of both the WEO andthe ICMR, we believe that it is better to address theissue of a year-end WEO/ICMR on an ad hoc, as-needed basis. Experience with this procedure, webelieve, demonstrates responsible and constructivedecision making.


Second, RES has important concerns about Box3.2, in which the evaluators assess how the WEO andthe ICMR dealt with the issues of capital flows and fi-nancial crises. This is clearly one area where pub-lished information provides evidence and substance tosupport an evaluation of the effectiveness of theFund’s work but where too little attention has beenpaid to the evidence contained in the Fund’s reports.While we accept a number of the conclusions in thisbox as reasonable, or at least arguably plausible, westrongly believe that the evaluators go too far in threespecific statements that are effectively contradictedby a fair reading of what was actually said in theWEO and by other information provided to the Exec-utive Board (which has responsibility for the conductof surveillance). The following statements, we be-lieve, need to be reconsidered and significantly modi-fied in light of the facts.

. . . when doubts about the sustainability of capital flowswere voiced—more frequently in the ICMR than in theWEO—they did not put much emphasis on weakeningfinancial systems in the capital-importing countries. . . .[emphasis added]

. . . the importance of potential regional and interna-tional contagionof currency crises was given very little

attention prior to the crisis, even in internal analyses.[emphasis added]

The overall impression of Fund multilateral surveil-lance as expressed through the WEO and the ICMR isthat while these documents did make a number of perti-nent observations on capital flows and financial crisesthat are helpful in understanding subsequent develop-ments in Asia and elsewhere, the risks were not force-fully presented.[emphasis added]

The footnote attached to this last statement addsthe following:

In our interviews, staff suggested that such warningswere in fact present and indeed couched in languagethat was quite strong, in Fund terms. However, onewell-informed and disinterested observer was of theview that the basic drafting strategy was to say as littleabout risks as possible, while at the same time stillbeing able to claim, if the risks did become reality, thatthey had been addressed. [emphasis added]

The exposition of why we believe that these state-ments are not well supported and merit reconsidera-tion and revision is detailed and extensive. We gointo it not only because of its importance to the eval-uators’assessment of the multilateral surveillancework of RES, which is overall quite favorable, butalso because of the deficiencies that it suggests in themethodology that the evaluators have employed inreaching their broader conclusions and recommenda-tions. The evaluators have relied, to a great extent, onthe information gathered from their extensive inter-views with a wide array of individuals who areknowledgeable, to a greater or lesser extent, aboutthe Fund’s work on surveillance and its effectiveness.In contrast, the evaluators appear to have spent muchless time and energy on their own direct examinationand evaluation of the Fund’s surveillance work.

The evaluators’approach is surely efficient and, inmany respects, the right approach. It is important forthe evaluators not to base their judgments too heav-ily on their own assessment and prejudices, butrather to seek much wider views among people wellinformed about the Fund’s surveillance activities.However, outside observers also have their preju-dices, and it is important to countervail them withcareful scrutiny of the documentary evidence. Formuch of the Fund’s surveillance work, the documen-tary evidence—records of discussions with authori-ties, Article IV reports, minutes of Executive Boardmeetings, and so forth—is not publicly available.For the WEO and the ICMR, in contrast, the pub-lished versions are available. This makes it possibleto compare the judgments and conclusions reachedby the evaluators on this component of the Fund’ssurveillance work, based on their own judgmentsand the opinions gathered from informed outside ob-

107

ANNEX I EXCERPTS FROM COMMENTS FROM THE RESEARCH DEPARTMENT

servers, with the record of what was actually said inthe WEO and the ICMR. Faced with the evidence,readers may be left to judge whether the conclusionsof the evaluators are well-founded with respect tothis specific area of Fund surveillance, and they maybe allowed to make broader inferences about thebasis for the evaluators’conclusions in other areaswhere the documentary record of the Fund’s surveil-lance activities is not publicly available.

The fact is that on page 2 of the May 1997 WEOthere is a quite explicit warning that two among thefour key risks to global growth are those arisingfrom disruptions in capital flowsto emerging marketeconomies, including the possibility of contagion,and from fragilities in their banking systems.

Third, capital flows to emerging market countries. Thesurge in such flows in recent years reflects both the grow-ing shift to a more open global financial system and thesuccessful economic policies of many recipient coun-tries. But caution is warranted since both the global avail-ability of these flows and their cost are vulnerable tohigher global interest rates and to adverse developmentsaffecting systemically important capital-importing coun-tries. While the aggregate global flows do not seem ex-cessive, the reliance on capital inflows by some coun-tries, and the associated narrowing of their interest ratespreads, may not be sustainable.

Finally, the fragile banking systems are of concern in abroad spectrum of countries. These problems often stemfrom excessive credit expansions in the past under con-ditions of inadequate prudential supervision. In someemerging market countries, banking sector difficultieslinked to significant exposure to foreign exchange riskhave become more apparent following the reversal ofcapital flows from abroad. Among transition countries,bank loans have often allowed enterprises to delay re-structuring, and as a result many firms have become in-creasingly unable to service their debt. Large portfoliosof nonperforming loans, the erosion of banks’capitalbases, and outright banking crises can affect countries’economic performance by obstructing banks’abilityand willingness to lend, by constraining the operation ofmonetary policy, and because of the budgetary costs ofrescuing and restructuring ailing financial institutions.

A similar warning is reiterated in the October1997 WEO (p. 2), before the attack on the HongKong dollar signaled the general onset of the Asiacrisis. On this occasion, the risk to the sustainabilityof capital flows to emerging markets, the possibilityof contagion, and the problems of weaknesses in fi-nancial systems are all linked together as one of thethree key risks to global growth.

Sustainability of capital flows to emerging market coun-tries. Several factors have contributed to record capitalinflows into many emerging market countries and an as-

sociated compression of yield differentials in recentyears, including the trend toward a more open global fi-nancial system and the increasingly successful economicpolicies pursued in many recipient countries. But theavailability of these flows and their costs are also influ-enced by global cyclical conditions and are vulnerable tohigher interest rates in world financial markets as well asto perceptions that large current account deficits—thecounterpart to capital inflows—may not be sustainablein all cases. The crisis in Mexico late in 1994 and morerecently the financial pressures that have affected Thai-land and a number of countries in Southeast Asia under-score the importance of disciplined macroeconomicpolicies and robust financial sectors. They also havehighlighted the risk and costs of potentially disruptivechanges in market sentiment, including the danger ofvery strong reactions in financial markets and seriousspillovers to other countries when critical policy weak-nesses are not addressed in a timely manner.

The rest of this chapter summarizes the IMFstaff ’s near-term projections and policy assessmentsand identifies some key policy concerns that need tobe addressed in order to strengthen medium-termeconomic prospects in all countries in accordancewith the guidelines set out by the Interim Committeein its September 1996 “Declaration on Partnershipfor Sustainable Global Growth.”1 Other issues dis-cussed include the prospects for EMU and its poten-tial longer-term implications for Europe, lessonsfrom recent exchange market crises and the trend to-ward greater flexibility of exchange rate regimes indeveloping countries, the challenges facing mone-tary policy in the transition countries in safeguardingprogress toward macroeconomic stability, and theneed for so-called second-generation reforms to sus-tain high-quality growth in all regions.

To provide the context and to show the prominencegiven to these warnings, the first three pages of theMay and October 1997 WEOs are attached. Also at-tached are the relevant initial pages from the October1996 WEO, which contain no similar warning. It isapparent that the warning clearly issued in the May1997 WEO and reiterated in the October 1997 WEOwas not just “boilerplate” that is commonly includedin the WEO to protect against later accusations thatsome important risks, subsequently actually realized,were overlooked. The prominence and the clarity ofthe warnings in the May and October 1997 WEOswere upgraded over those in earlier WEOs in order toindicate rising levels of concern about a possibly im-minent problem. Of course, not everyone reads theWEO in the same way. Some will not understand orappreciate a warning even when it comes in clear lan-

108

1See World Economic Outlook, October 1996, p. xii.


guage in a prominent location. Perhaps this is espe-cially so for those who have a vested interest, or atleast a psychological interest, in believing that theywere not adequately warned because they continuedto participate in the massive flow of capital to emerg-ing market countries even after the May 1997 WEOwarning that such flows might prove unsustainable. Isthe unnamed individual quoted in the evaluators’foot-note in this category? We don’t know.

Nevertheless, we do assert that on a dispassionatereading, the May 1997 WEO does provide a clearwarning of important risks to global growth from po-tential slowdowns or reversals of capital flows toemerging markets, possibly magnified by contagion,and from fragilities in their financial systems. If theevaluators disagree, as they clearly do in Box 3.2,then they should continue to assert their conclusions.However, the footnote disparaging the staff responseto this box should have been deleted; it cites oneanonymous source whose objectivity and motivesmay be suspect or who is unaware of the facts. Rather,in their report the evaluators should have quoted theentire text of the two relevant paragraphs providingthe warning in the May 1997 WEO (and noted the re-iteration of this warning in the October 1997 WEO),they should have noted the prominent position ofthese paragraphs in the WEO, and they should haveemphasized the difference between this warning andwhat was said in earlier WEOs. The reader could thenhave been left to judge whether the warning providedwas reasonable and, correspondingly, whether thejudgment of the evaluators is entirely fair.

The above comments relate to publicly availableinformation. However, Box 3.2 also refers to thepreparation by the staff in late August 1997 of “a—broadly upbeat—memorandum to the Board on therisks of contagion in Asia [that] did not even men-tion Korea.” In fact, this reading of the record ig-nores the fact that, in his presentation to the Execu-tive Board on August 27, 1997, the EconomicCounsellor clearly indicated that he saw the risks fora number of countries in Asia as being on the down-side and that a realistic downward revision wouldprobably lower the forecasts for some countries by2–3 percentage points. He recommended againstmaking such an adjustment at the time, in view ofthe concern that such action could exacerbate thecrisis. However, some modest downward revisionswere subsequently made to the forecasts for the pub-lished WEO, along with the inclusion of a moreforceful discussion of the risks as indicated earlier.

Could and should the warning of a possible crisisaffecting capital flows to emerging markets havebeen issued earlier? In the May 1994 WEO, beforethe “tequila crisis,” a warning was given in the initialdiscussion of prospects and possible risks for devel-oping countries (Chapter 1, page 7).

. . . the surges in capital inflows and in stock marketprices give cause for some concern about risk of over-heating and the possibility of sudden changes in marketsentiment. To minimize the risk of speculative bubblesin the emerging stock markets and a reversal of capitalinflows, many of these countries will need to monitordevelopments carefully to avoid the buildup of imbal-ances; some countries may need to take corrective mea-sures relatively soon. A number of countries may alsoneed to strengthen prudential supervision of their finan-cial systems and, in some cases, broaden and deepen fi-nancial market reforms.

This warning was reinforced in a more extensivediscussion of equity flows to emerging markets andtheir risks (in Box 4, pages 26–27). Then, in the Oc-tober 1994 WEO, the risks from a possible sharpslowdown of capital flows to some emerging mar-kets were again noted (in the initial discussion ofprospects and risks for developing countries, onpage 6): “In a few cases, the rise in capital flows ap-pears to have reflected the general enthusiasm foremerging financial markets, rather than well-founded confidence in economic prospects. Forthese countries, the risk of sudden changes in marketsentiment is particularly serious.” The main chapteron developing countries was devoted to “The RecentSurge in Capital Flows to Developing Countries,”and provided a balanced assessment of the benefits,problems, and risks arising from such flows. Todrive home the potential risks, the chapter concludedwith an alternative scenario to the WEO baselineforecast that showed how a “sharp reversal of capitalflows,” together with policy slippages, in developingcountries could have substantial and sustained ad-verse effects. The chapter concluded with the warn-ing, “Despite the generally positive character of thelarge capital inflows, there are a number of countrieswhere the confidence of foreign investors may notbe warranted on a sustained basis.”

Granted, these warnings were not quite as promi-nent as those presented in the May 1997 and October1997 WEOs. Granted also, these warnings were nota forecast of the Mexican devaluation of December1994 and of the “tequila crisis” that would follow inits wake. Nevertheless, someone who read the 1994WEOs in any depth, especially someone who wasinterested in capital flows to emerging markets andread Box 4 of the May WEO and, especially, Chap-ter IV of the October WEO, would have recognizedthat there were significant risks that recent surges ofcapital flows were not sustainable for at least someemerging market countries.

With the onset of the tequila crisis in late 1994,capital flows to Latin America fell off sharply, withflow reversals experienced by Mexico and Ar-gentina. Interest rate spreads shot up for Latin Amer-

109

ANNEX I EXCERPTS FROM COMMENTS FROM THE RESEARCH DEPARTMENT

ican borrowers. Other emerging markets generallyfelt only quite brief adverse spillovers, and net pri-vate capital flows to all emerging markets picked upin 1995 over 1994 levels, with substantial gainsmainly for Asia. At this time, with an actual crisis af-fecting several emerging market countries, and withmarkets sensitized to risks, there seemed to be littlepoint in stressing the possibility of an even deeperand more widespread crisis. Indeed, the perceptionin the Fund was that financial markets had over-reacted to the true underlying weaknesses in Mexicoand, especially, Argentina; and the effort of the inter-national community was to help rebuild confidencein order to promote recovery in the countries directlyaffected and limit risks of a spread of the crisis.

Nevertheless, the October 1995 WEO in ChapterIV, “Increasing Openness of Developing Countries—Opportunities and Risks,” devoted considerable atten-tion to the risks of capital flow reversals and empha-sized both weaknesses in domestic banking systemsand high volumes of short-term capital inflows as fac-tors likely to aggravate such risks seriously. The mes-sage was largely one of important lessons to belearned from the tequila crisis, rather than a clearwarning of imminent risks. In the concluding sectionof the chapter on “Implications of a Reversal of Capi-tal Inflows,” an alternative scenario was again used,as in the October 1994 WEO, to dramatize the ad-verse consequences of a possible future substantialslowdown of capital flows to emerging markets.

The revival of capital flows to Latin America, thecontinuing increase in gross private capital flows toemerging markets generally, and the narrowing ofspreads in late 1995 and the first half of 1996 (see Fig-ure 1) were generally regarded as welcome andhealthy developments. By July 1996, when the Octo-ber WEO was being prepared, the sharply rising vol-ume of, and declining spreads on, capital flows toemerging market countries were pointed out, with adegree of concern about possible overexuberance, tothe Executive Board in the WEMD session. At thattime, however, there did not yet appear to be sufficientreason from global financial developments to expressa clear warning about imminent risks concerning thesustainability of financial flows to emerging markets.The evaluators disagree but have not made clear thebasis for their disagreement (other than hindsight).

By late 1996 (in the WEMD session of late No-vember) and surely by early 1997 (in the WEMDsession of mid-January), concerns about the healthi-ness and sustainability of capital flows to emergingmarkets began to rise as volumes of new gross pri-vate financing continued to increase, as spreads foremerging market borrowers continued to decline,and as more and more emerging market entitiesgained relatively favorable access to global financialmarkets. The Economic Counsellor raised these con-

cerns with the G-7 deputies and with Working PartyThree of the OECD in late 1996 and early 1997.Supporting views were expressed, particularly bycentral bank representatives. Thus, at the time whenthe May 1997 WEO was being prepared for Execu-tive Board consideration (at end-March), there was agrowing official consensus that there were importantrisks to capital flows to emerging markets. This con-cern was expressed publicly, and quite deliberately,in the WEO released to the press in late April andformally published in May.

Could and should the warning of a possible crisishave been more explicit? Until it actually happens,there is no way to be certain whether a financial crisiswill occur or which countries it may affect. (For ex-ample, recent WEOs have continued to warn aboutthe risks from a possible reversal of large gains in in-dustrial country stock markets that has not material-ized.) When the WEO was released to the press inApril 1997, the Czech koruna was under pressure, butabandonment of the peg and depreciation of the cur-rency did not occur until May. Slovakia also had largecurrent account and budget deficits, and there werefears that its currency would also be pulled down. Ifso, other Central and Eastern Europe countries mightcome under pressure, especially if financial marketsbegan to take a more skeptical view of the sustainabil-ity of Russia’s public finances (see the May 1998WEO, pp. 100–101). After reducing the federal fundsrate by 75 basis points when the U.S. economyslowed during 1995 (partly due to spillover effectsfrom the Mexican crisis), the Federal Reserve took aninitial 25-basis- point step toward tightening in March1997 on concerns that the U.S. economy might bemoving toward overheating. In the spring and sum-mer, financial markets were anticipating possible fur-ther tightenings. Although effects from the Asian cri-sis later helped to deter such moves, a relevantconcern in May 1997 was that further Federal Re-serve tightenings might, as in past such episodes, putpressure on capital flows to emerging markets, partic-ularly on flows to Latin America. In such an event,Brazil was seen as vulnerable, and problems in Brazilwere likely to affect Argentina. In Asia, the Fund hadseen problems developing in Thailand at least sincethe second half of 1996. Malaysia, which also had alarge current account deficit financed by buoyant pri-vate capital inflows, was also seen as vulnerable.Other Asian emerging market countries, most impor-tantly Korea, were not perceived to be immune; but asthe evaluators indicate, the risks for these countrieswere not rated as particularly great. Thus, the warningin the May 1997 WEO was a general warning of pos-sible financing problems for a number of emergingmarket countries and of sufficient importance to be arisk to the baseline scenario for global economicgrowth, including through the phenomenon of conta-

110


gion. The location of such financing problems, shouldthey develop, was not—and arguably could not be—identified. A financial crisis of the scale and virulencethat struck several Asian emerging market economiesafter the attack on the Hong Kong dollar in mid-Octo-ber 1997 was not foreseen or warned about in theMay 1997 WEO. But a clear warning about risks ofglobal significance arising from disruption of capitalflows to emerging markets and related to problems intheir financial sectors was issued.

Were the risks potentially arising from weaknessin emerging market financial systems given ade-quate emphasis in the WEO? Here, the evaluatorshave a more valid point, although overstated. TheMay and October 1997 WEOs do clearly emphasizethat weaknesses in emerging market financial sec-tors, especially in an environment of potential dis-ruptions of capital flows, were an important risk.The significance of this problem for some individualemerging market countries, however, was notbroadly recognized in the Fund (outside of somestaff in the Research Department’s capital marketsgroup), and the risks in this area for specific coun-tries were not noted in the WEO. Korea, which is thefocus of attention in Box 3.2, and Indonesia standout in this regard.

What about contagion? In their paper on “The Roleof the Fund” written for Executive Board considera-tion in the aftermath of the tequila crisis, Masson andMussa emphasized the “monsoonal character” of dis-ruptions of capital flows to emerging markets. In thedebt crisis of the 1980s and in the tequila crisis of

1995, several countries concentrated in the same re-gion were affected by the crisis. The caution in theMay 1997 WEO that “the global availability of theseflows [of capital to emerging market countries] andtheir costs are vulnerable to higher global interestrates and to adverse developments affecting systemi-cally important capital-importing countries” conveysthe notion of risks to global growth that would arisefrom something more than isolated instances of dis-ruptions of capital flows to individual emerging mar-ket economies. The May 1999 WEO was clearly notthe first instance in the recent crisis when serious at-tention was paid to potential regional and interna-tional contagion of currency crises. Beyond the warn-ings given in the May and October 1997 WEOs, thedecision to produce the Interim WEO on the AsianCrisis (which was discussed at the annual meetings inHong Kong in early October and made available im-mediately after the attack on the Hong Kong dollar inmid-October) was an emphatic recognition that conta-gion was not merely a potential concern; it was a factfor Asian emerging market economies and a clear riskfor others. By late 1997, the financial media were al-ready talking about “the Asian contagion.” The mainpurpose of the discussion of contagion in May 1999was not to emphasize risks of contagion that had al-ready occurred on a global scale. It was to analyze themechanisms through which contagion operated. TheICMR published in October 1998 and the InterimWEO/ICMR of December 1998 provide further de-tailed information on how contagion operated throughglobal financial markets.

111

EMBI spread(left scale)

Gross privatefinancing

(right scale)

May99

2

4

6

8

10

12

14

16

0

5

10

15

20

25

30

35

40

989796959493921991

Percent Billions of U.S. dollars

Figure 1. EMBI Spread and Gross Private Capital Flows

ATTACHMENT I (World Economic Outlook, May 1997, pp. 1–3)

World economic growth quickened during 1996following widespread deceleration of activity in

1995 (Chart 1). Economic and financial conditions aregenerally propitious for the global expansion to con-tinue in 1997 and the medium term at rates at leastmatching those seen in the past three years (Chart 2).There are few signs of the tensions and imbalancesthat usually foreshadow significant downturns in thebusiness cycle: global inflation remains subdued, andcommitments to reasonable price stability are perhapsstronger than at any other time in the postwar era; fis-cal imbalances are being reduced with increasing de-termination in many countries, which should help con-tain real long-term interest rates and foster higherinvestment; and exchange rates among the major cur-rencies appear to be generally consistent with broaderpolicy objectives.

In many countries, structural reforms are enhanc-ing the role of market forces and thereby strengthen-ing the basis for sustained, robust growth. Theprocess of trade integration continues to deepen andis being supported by growing liberalization of ex-ternal payments. Also, changes in the role of thestate through privatization and deregulation are rais-ing efficiency and spurring private sector activity ina growing number of successfully managedeconomies in all regions.

The favorable global economic conditions are un-derscored by the continued robust growth perfor-mance with low inflation in the United States andthe United Kingdom, the pickup in growth in Japanin 1996, and improved prospects for a strengtheningof the recoveries in continental Europe and Canada.In many of the dynamic emerging market countries,there was a desirable moderation of growth and in-flation in 1996, which should allow their expansionsto be sustained in the period ahead. Growth haspicked up in those developing countries in the West-ern Hemisphere that were particularly affected bythe financial crisis in Mexico in 1995. Activity hasalso strengthened in the Middle East and Africa,while the transition countries, as a group, are ex-pected to register positive growth in 1997 for thefirst time since the collapse of central planning.

Nevertheless, despite these grounds for optimism, itis important to recognize that contrasts in economicperformance across countries have become starker inrecent years. There are also a number of risks to thecentral scenario. First, in much of the European Union

112

IGlobal Economic Prospects and Policies

1

Chart 1. World Industrial Pr oduction1

(Percent change from a year earlier)

1Data are for output in manufacturing in 30 advanced and emergingmarket economies representing 75 percent of world output; three-month centered moving average.

–5

0

5

10

15

20Emerging market countries

World

Advanced economies

1991 92 93 94 95 96 Jan.97

Following a marked slowdown in 1995, the pace of world industrialactivity quickened during 1996.

ATTACHMENT I (World Economic Outlook, May 1997)

(EU), unemployment has risen further to new postwarpeaks, and neither prospective growth nor the progressmade with labor market reforms gives reason to expectany significant decline in joblessness in the near fu-ture. High unemployment and weak growth couldmake it difficult for EU members to fully meet the fis-cal deficit targets associated with the plan for mone-tary union, affect expectations about the likelihood ofthe project going ahead on time, and lead to turbulencein financial markets.

Second, stock markets. The strength of equityprices in the United States and many other countriesin the period up to early March was a reflection of in-vestors’positive assessment of the business outlook.But recent declines in equity prices have underscoredthe risk of a more significant correction, especially ifearnings expectations were to be downgraded or areemergence of inflationary pressures were to requirea marked rise in interest rates. The potential for a mar-ket correction large enough to contribute to a cyclicaldownturn depends partly on the extent to which therise in stock prices has been an element in a broaderbuildup of demand pressures. In contrast to the run-upin asset prices in the late 1980s, especially in Japanbut also in the United States and several other coun-tries, a generalized overvaluation of asset prices,leveraged by increased indebtedness, does not appearto be present in most countries with strong stock mar-kets. Nevertheless, a significant decline in stockprices could undermine confidence in some countries.

Third, capital flows to emerging market countries.The surge in such flows in recent years reflects boththe growing shift to a more open global financialsystem and the successful economic policies ofmany recipient countries. But caution is warrantedsince both the global availability of these flows andtheir cost are vulnerable to higher global interestrates and to adverse developments affecting systemi-cally important capital-importing countries. Whilethe aggregate global flows do not seem excessive,the reliance on capital inflows by some countries,and the associated narrowing of their interest ratespreads, may not be sustainable.

Finally, fragile banking systems are of concern ina broad spectrum of countries. These problemsoften stem from excessive credit expansion in thepast under conditions of inadequate prudential su-pervision. In some emerging market countries,banking sector difficulties linked to significant ex-posure to foreign exchange risk have become moreapparent following the reversal of capital flowsfrom abroad. Among transition countries, bankloans have often allowed enterprises to delay re-structuring, and as a result many firms have becomeincreasingly unable to service their debt. Largeportfolios of nonperforming loans, the erosion ofbanks’capital bases, and outright banking crises canaffect countries’economic performance by ob-structing banks’ability and willingness to lend,

1132

Chart 2. World Indicators 1

(In percent a year)

1Shaded areas indicate IMF staff projections.2Volume of goods and services.3GDP-weighted average of ten-year (or nearest maturity) govern-

ment bond yields less inflation rates for the United States, Japan, Ger-many, France, Italy, United Kingdom, and Canada. Excluding Italyprior to 1972.

0

5

10

15

20

–6

–4

–2

0

2

4

6

8

–4

0

4

8

12

–4

0

4

8

12

Average, 1970–96

Developing countries(median)

Advanced economies

1970 75 80 85 90 95 2000

1970 75 80 85 90 95 2000 1970 75 80 85 90 95 2000

1970 75 80 85 90 95 2000

Average, 1970–96

Growth of World Real GDP Growth of World Trade2

Inflation (consumer prices) World Real Long-TermInterestRate3

The global expansion is expected to continue with the growth of worldoutput and trade above trend, while inflation should remain containedin the advanced economies and slow further in thedeveloping countries.

ATTACHMENT I (World Economic Outlook, May 1997)

by constraining the operation of monetary policy, andbecause of the budgetary costs of rescuing and re-structuring ailing financial institutions.

* * *It is becoming increasingly clear that the benefits

of a favorable global economic environment do notaccrue automatically to any country. In fact, re-markable differences persist in the degrees of suc-cess that countries have had in taking advantage ofthe opportunities for strengthening their economicperformance.

• Among the advanced economies, develop-ments have been mixed and cyclical positionsdiffer widely. Prospects for recovery have im-proved in continental western Europe follow-ing disappointing performances in 1995 andmuch of 1996. But unemployment is expectedto remain at or near record levels in France,Germany, Italy, and several other countries. InJapan, growth was stronger than expected in1996, and there is upside potential for activityin 1997 although there remains uncertainty infinancial markets, in particular, as to whetherthe momentum of Japan’s recovery will bemaintained in the period ahead. The uncertainprospects and lack of confidence characteristicof these economies in recent years contrastwith the favorable performance of the UnitedStates and the United Kingdom as well as anumber of smaller countries including Aus-tralia, Denmark, Ireland, New Zealand, and theNetherlands. These contrasts reflect both cycli-cal and structural factors, including policies.

• An increasing number of developing countries inall regions are reaping the benefits of the stead-fast pursuit of sound financial policies and out-ward-oriented, market-based structural reforms.This is reflected in large inflows of foreign directinvestment, rapid expansion of both exports andimports, and solid growth prospects. But somecountries have experienced setbacks and othersare vulnerable to changes in investor sentiment.While economic conditions have clearly been im-proving in a growing number of low-incomecountries, many of the poorest countries havecontinued to fall behind, facing the risk of mar-ginalization from the mainstream of global eco-nomic progress.

• Among the transition countries, the contrasts inperformance have also widened between some ofthe early, relatively successful reformers andcountries that have started adjustment and reformlater and with less determination and consistency.Between these two extremes, which, to be sure,also reflect widely different starting conditions,there are wide ranges of policy effort and eco-nomic success.

Motivated in part by these contrasts, the InterimCommittee in its September 1996 “Declaration onPartnership for Sustainable Global Growth” set out arange of broad policy principles to promote the fullparticipation of all economies in the global econ-omy. These principles stress the need to implementsound macroeconomic policies that consolidate suc-cess in bringing inflation down, strengthen fiscaldiscipline, enhance budgetary transparency, and im-prove the quality of fiscal adjustment; to foster fi-nancial and exchange rate stability and avoid cur-rency misalignments; to maintain the impetustoward trade liberalization and current account con-vertibility; to tackle labor and product market re-forms more boldly; and to ensure the soundness ofbanking systems and promote good governance inall its aspects. The complementary and mutually re-inforcing roles of macroeconomic and structuralpolicies were given particular emphasis.1

The uneven performance across countries and un-even distribution of rewards within them are fre-quently linked to the phenomenon of globalization—the rapid integration of economies worldwide throughtrade, financial flows, technology spillovers, informa-tion networks, and cross-cultural currents. There is nodoubt that globalization is contributing enormously toglobal prosperity. At the same time, however, publicdebate often focuses on perceived negative aspects ofglobalization, including the effects on employmentand real wages, especially of the low skilled, in the ad-vanced economies. Globalization, like any form oftechnological or structural change, may adversely af-fect the living standards of some in the short run.However, it does not seem to be the principal force be-hind the unfavorable developments in employmentand income distribution observed in some advancedeconomies.

Another widespread perception is that globalizationmay, at some cost, limit the autonomy of policymakersat the national level. It is argued in this report thatwhile it does appear that globalization increases thecosts of economic distortions and imbalances, policyrelated or otherwise, it clearly enhances the rewards ofsound policies. In this way, globalization may be con-tributing to the apparent polarization between success-ful countries and those that are falling behind in rela-tive, and sometimes even absolute, per capita incomepositions. Globalization is not, however, a zero-sumgame with some economies winning at the expense ofliving standards and employment elsewhere. If poli-cies are adapted to meet the requirements of integratedand competitive world markets, then all countriesshould be better able to develop their comparative ad-vantages, enhance their long-run growth potential, andshare in an increasingly prosperous world economy.

114

1See World Economic Outlook,October 1996, p. xii.

3

ATTACHMENT II (World Economic Outlook, October 1997, pp. 1–3)

W ith world output expected to expand by some4!/4 percent in both 1997 and 1998, the

strongest pace in a decade, the global economy is en-joying the fourth episode of relatively rapid growthsince the early 1970s (Figure 1). The expansion is un-derpinned by continued solid growth with low infla-tion in the United States and the United Kingdom; astrengthening recovery in Canada; a broadening of re-covery across continental western Europe, notwith-standing persistent weakness in domestic demand insome of the largest countries; robust growth trends inmost of the developing world, particularly in Chinaand much of the rest of Asia even though some coun-tries are likely to experience a setback associated withrecent turmoil in financial markets in Southeast Asia;and evidence of an end to the decline in output, andperhaps a beginning of growth, in Russia and in thetransition countries as a group. It is worth recalling,however, that each of the three previous episodes ofrelatively rapid growth was followed by widespreadslowdown and even recession in many countries. Tak-ing account of this earlier experience, is there a dan-ger that the present expansion may soon run out ofsteam and give way to a new global downturn?

Although a moderation of world growth is indeedlikely to occur at some point, there are reasons to believe that the current expansion can be sustained,possibly into the next decade. First, there are rela-tively few signs of the tensions and imbalances thathave usually presaged significant downturns in thebusiness cycle: global inflation remains subdued and commitments to safeguard progress towardprice stability are perhaps stronger than at any othertime in the postwar era; fiscal imbalances are beingreduced with increasing determination in manycountries, which is helping to contain inflation ex-pectations and real interest rates; and exchange ratesamong the major currencies, taking account of rela-tive cyclical conditions, are generally within rangesthat appear to be consistent with medium-term fun-damentals. Second, cyclical divergences have re-mained sizable among the advanced economies, andthere are still considerable margins of slack to betaken up in Japan and continental Europe. Strongergrowth during the period ahead in these countriesshould help support global demand and activity asgrowth slows to a more sustainable pace in thosecountries that have already reached a mature stage in their expansions, especially the United States,

115

IGlobal Economic Prospects and Policies

1

Figure 1. World Output and Inflation 1

(Annual percent change)

1Shaded areas indicate IMF staff projections. Aggregates are com-puted on the basis of purchasing-power-parity weights unless otherwiseindicated.

1970 75 80 85 90 95 20000

5

10

15

20

1970 75 80 85 90 95 20000

2

4

6

8

Average, 1970–96

Developing countries(Consumer prices, median)

Advanced economies(GDP deflator)

Growth of World Real GDP

Inflation

The expansion of world output is expected to continue above trend,while inflation should remain contained.

ATTACHMENT II (World Economic Outlook, October 1997)

the United Kingdom, and several of the smaller ad-vanced economies. Third, the recovery that is inprogress in the transition countries seems likely tocontinue to strengthen at the same time as the grow-ing number of successful economies in the develop-ing world are also providing both new markets andincreased production capacities; these developmentsare stimulating trade and growth worldwide whilehelping to dampen price pressures. Taking into ac-count the combination of the strong catch-up poten-tial of the developing and transition countries andthe beneficial effects on productivity of technologi-cal advances and increasing globalization, the sus-tainable rate of world output growth may now in factbe somewhat stronger than in the quarter centurysince the first oil shock. This view is embodied inthe IMF staff ’s medium-term scenario, which pointsto a trend growth rate of world GDPof about 4!/2percent between 1996 and 2002 compared with anaverage rate of expansion of 3#/4 percent since 1970.

This generally positive assessment of the globaloutlook should not lead to complacency becausethere is a wide range of risks and fragilities that con-front individual countries and may affect regionaland world economic and financial conditions. Themain areas of concern relating to prospects over theshort to medium term include the following:

• Risks of overheating. Although world inflationhas subsided to the lowest rates seen since theearly 1960s, inflationary pressures couldreemerge, especially in countries that havereached high levels of resource use. Effective pol-icy to prevent inflation rising requires vigilancenot only against overheating in product and labormarkets but also in asset markets, and it requirespreemptive action when warning signs appear.Problems stemming from large swings in assetprices emerged in the late 1980s and the early1990s in a number of countries, most notably inJapan but also in Australia, Sweden, the UnitedKingdom, and the United States, with repercus-sions on the soundness of financial systems insome cases. More recently, several emerging mar-ket countries, especially in Southeast Asia, haveexperienced similar difficulties in their real estatesectors. Despite some correction in August, thereis also reason for concern about the strength ofworld stock prices, which may to some extent bebased on unrealistic expectations about prospectsfor future profit growth and low interest rates. Amore substantial correction in stock prices, wereit to occur, could adversely affect confidence andeconomic activity.

• Uncertainties about the Economic and MonetaryUnion (EMU) in Europe. The marked conver-gence of interest rates among the prospectivemembers of the monetary union seems to suggestthat financial markets expect the project to go

ahead in accordance with the agreed timetable,which calls for the new currency, the euro, to bein place by January 1999. Investor sentiment maystill change, however, if the feasibility of thetimetable was perceived to be threatened. In thatcase, interest risk premiums might again widenfor some countries, while the currencies of othersmight be subject to unwelcome upward pressure.Also, should growth prove insufficient to permitprogress in reducing record levels of unemploy-ment in much of Europe, confidence would re-main weak; in some cases there might be a risk ofresort to counterproductive fiscal policies incom-patible with the requirements of EMU.

• Sustainability of capital flows to emerging marketcountries. Several factors have contributed torecord capital inflows into many emerging marketcountries and an associated compression of yielddifferentials in recent years, including the trendtoward a more open global financial system andthe increasingly successful economic policiespursued in many recipient countries. Buttheavailability of these flows and their costs arealso influenced by global cyclical conditionsandare vulnerable to higher interest rates in worldfinancial markets as well as to perceptions thatlarge current account deficits—the counterpart tocapital inflows—may not be sustainable inallcases. The crisis in Mexico late in 1994 andmorerecently the financial pressures that have affectedThailand and a number of countries in SoutheastAsia underscore the importance of disciplinedmacroeconomic policies and robust financialsectors. They also have highlighted the risk andcosts of potentially disruptive changes inmarketsentiment, including the danger of very strongreactions in financial markets and serious spill-overs to other countries when critical policyweaknesses are not addressed in a timely manner.

The rest of this chapter summarizes the IMF staff ’snear-term projections and policy assessments andidentifies some key policy concerns that need to beaddressed in order to strengthen medium-term eco-nomic prospects in all countries in accordance withthe guidelines set out by the Interim Committee in itsSeptember 1996 “Declaration on Partnership for Sus-tainable Global Growth.”1 Other issues discussed in-clude the prospects for EMU and its potential longer-term implications for Europe and the world economy,the critical need for labor market reforms in Europe,lessons from recent exchange market crises and thetrend toward greater flexibility of exchange rateregimes in developing countries, the challenges fac-ing monetary policy in the transition countries in

116

1See World Economic Outlook, October 1996, p. xii.

2

ATTACHMENT II (World Economic Outlook, October 1997)

safeguarding progress toward macroeconomic stabil-ity, and the need for so-called second-generation re-forms to sustain high quality growth in all regions.

Advanced Economies

The high degree of price stability remains an im-pressive achievement shared by almost all of the ad-vanced economies. In 1996, the rate of consumer priceinflation averaged 2!/2 percent, and only four countriesexperienced inflation above 5 percent; measured byGDPdeflators, a broader measure of the price level,average inflation was just 2 percent. In terms of outputand employment, the picture is much more mixed asunderscored by sharp divergences in labor marketperformance in recent years. Whereas a number ofeconomies including the United States, the UnitedKingdom, and several of the smaller advanced econ-omies are operating at relatively high levels ofresource use, the three major continental Europeancountries have suffered protracted economic weaknessthat has been accompanied by a dramatic rise in unem-ployment to postwar record levels. Conditions for re-covery have gradually improved, but few forecastersexpect the upswing to make more than a modest dentin unemployment. In Japan, the recovery has alsoproven quite hesitant, as discussed below and in Chap-ter II in greater detail.

The unsatisfactory economic performance of thethree major economies of Germany, France, andItaly cannot be blamed on the external environ-ment. In fact, external markets have been expand-ing strongly and exports have been the main sourceof stimulus in recent years. Thus, between 1992and 1997, the real foreign balance is estimated tohave improved by 1!/4 percent of GDPin Germany(most of this occurring in 1996 and 1997), 3!/2 per-cent of GDPin France, and 5!/2 percent of GDPinItaly. This clearly indicates that the sources ofweakness have been internal, and in fact domesticdemand has expanded by less than 1 percent a yearin these three countries combined over the past fiveyears.

There are at least four sets of factors that need to beconsidered in explaining this exceptional sluggishness.

(1) Fiscal consolidation. Since 1992, there hasbeen a substantial effort in many countries to reducefiscal imbalances that had reached unsustainable lev-els; although beneficial for growth in the longer run,those efforts have tended to weaken aggregate de-mand in the short run notwithstanding offsetting ef-fects from lower interest rates and exchange rates. Incontinental Europe as a whole, however, fiscal policy(measured by changes in cyclically adjusted balances)has not been substantially tighter than in the UnitedStates or the United Kingdom. Differences in fiscalstance therefore clearly cannot by themselves explainthe differences in growth performance.

(2) Labor market rigidities. The lack of flexibilityof continental European labor markets has undoubtedlyexacerbated the weakness of economic activity at thesame time as product market rigidities may have im-peded the private sector’s adjustment to the withdrawalof fiscal stimulus. Some labor market measures, suchas work sharing and early retirement, which were in-tended to reduce open unemployment, may actuallyhave served to dampen growth by reducing the supplyof skilled labor and increasing tax burdens and laborcosts.

(3) Confidence factors. Although such influencesare difficult to assess in isolation from other forces,delays in addressing the root causes of structural un-employment and fiscal imbalances may well haveaffected business confidence, while labor sheddingin response to high labor costs has increased job in-security and undermined consumer confidence. Ex-cessive reliance on revenue increases to reduce fis-cal deficits rather than reform-based reductions inexpenditures may also have discouraged both invest-ment and consumption. Recurrent uncertaintiesabout the feasibility of the timetable for EMU haveprobably added to hesitation in business investment.

(4) Monetary policy. The progressive easing ofmonetary conditions in Germany and the rest of Eu-rope during 1993 and early 1994, and generally de-clining risk premiums in long-term interest rates,played a significant role in the first phase of recoveryin 1994. As this initial pickup failed to turn into a self-sustained expansion, owing in part to the effects of anoverly strong deutsche mark in early 1995, official in-terest rates were reduced further during 1995 andearly 1996. However, while the stance of monetarypolicy since the latter part of 1993 has supported de-mand, a somewhat faster and ultimately more pro-nounced easing of monetary conditions would havehelped put the recovery on a stronger footing withoutjeopardizing price performance. The timing of sucheasing was constrained by the rise in long-term inter-est rates in 1994, but an easier monetary stance wasjustified subsequently by the absence of inflationarypressures, the prevalence of significant cyclical un-employment, the large withdrawals of fiscal stimulus,and depressed levels of consumer and business confi-dence. As of August 1997, with the further weakeningof the currencies participating in the European ex-change rate mechanism (ERM) vis-à-vis the U.S. dol-lar and the pound sterling providing additional stimu-lus, monetary conditions in continental Europe appearto be sufficiently supportive of the emergingrecoveries.

Movements of major currency exchange rates sincethe spring of 1995 have corrected earlier misalign-ments, and the present configuration is generally help-ful and appropriate in view of relative cyclical posi-tions. Specifically, the present relatively strong valuesof the currencies of the United Kingdom and the United States in comparison with the currencies of

1173

ATTACHMENT III (International Capital Markets, August 1995, pp. 13–14)

flexible than other policy instruments. Most capital-importing countries have not explicitly employedfiscal consolidation as a response to inflows, but aspart of medium-term adjustment programs. Moreproblematically, fiscal consolidation may also en-courage capital inflows by easing concerns aboutpossible future liquidity problems.

It is, therefore, not surprising to find that in addi-tion to altering monetary, fiscal, and exchange ratepolicies in response to large swings in internationalcapital flows, many counties have employed mea-sures that discourage capital inflows or seek to influ-ence their character. These measures are often gener-ically referred to as “capital controls.” In fact, suchmeasures range from prudential controls on thebanking system, to market-based measures, all theway to quantitative controls on inflows and outflows(Box 1). In particular, these measures have includedimposing or tightening prudential limits on banks’offshore borrowing and foreign exchange transac-tions (Indonesia, Malaysia, and the Philippines), aswell as taxing some types of inflows by requiringnon-interest-bearing reserve deposits against foreigncurrency borrowing by firms (Brazil, Chile, andColombia). For example, in Chile, the measureshave taken the form of non-interest-bearing 30 per-cent reserve deposits placed at the Central Bank for a period of one year on direct foreign currencyborrowing by firms.

In some instances, measures have taken the formof quantitative restrictions. For example, Colombiarestricts foreigners from investing in the domes-tic bond maket. Malaysia responded to the inflow ofspeculative short-term bank deposits with the impo-sition of several quantitative measures. The mostsuccessful of these measures was the prohibition ondomestic residents selling short-term money-marketinstruments to foreigners. In this case, abandoningthe sterilization of foreign exchange interventionand imposing capital controls appear to have beensuccessful in reducing domestic interest rates andshort-term inflows. A number of countries particu-larly Asian developing countries, have restrictionson foreign borrowing by domestic companies andsome have maintained prudential restrictions on fi-nancial institutions, such as restrictions on the openforeign exchange positions of banks.

It is dangerous to draw general conclusions aboutthe consequences of “capital controls” without refer-ence to the nature of such measures and the circum-stances under which they were employed. On theone hand, comprehensive and detailed restrictionson capital inflows and outflows can have highly dis-torting effects, and such restrictions tend to erodeover time. As the effectiveness of controls becomesweaker, authorities may be tempted to intensifythem, increasing their distortionary effect.

On the other hand, measures to discourage excessshort-term, foreign currency denominated borrow-ing by banks, such as increased reserve require-ments, can be justified on prudential grounds—bank failures can have significant real effects, aswell as fiscal consequences, when deposits are defacto guaranteed. Such measures also tend to have amore permanent effect. Some strong measures, suchas taxes on short-term capital flows and bans on thepurchase of particular types of securities, may bejustified only as temporary measures until domesticfinancial markets and institutions become well es-tablished and resilient, while some other types ofprudential measures and reserve requirements canbe justified as more permanent features of the regu-latory framework.

For example, a review of the Chilean andMalaysian experiences reveals that, in the short run,the volume of inflows was reduced by capital con-trols during episodes of higher exchange rate volatil-ity and little or no sterilization, in 1991 and 1994, re-spectively. Furthermore, capital controls wereundoubtedly less important than sound fundamentalsin explaining the long-run success of several coun-tries cited above in dealing with capital inflows.

In this regard, it should be noted that both HongKong and Singapore have managed large capital in-flows without recourse to capital controls. Therefore,although capital controls may be helpful at times,they are not the distinguishing feature characterizingcountries that have dealt successfully with capital in-flows and outflows. Imposing capital controls onoutflows during a crisis is interpreted as a measure ofdespair and hence is counterproductive. Further-more, market participants tend to view the control ofcapital outflows as a confiscatory measure, whichcan be expected to increase future borrowing costs,

118

Table 4. Reserve Accumulation and Capital Inflows(Changes in reserves as a percent of the balance in the capital account)1

First Second Third Fourth FifthCountry Year Year Year Year Year Average

Argentina (1991) 82 42 46 12 . . . 46Chile (1990) 77 167 86 22 110 92Colombia (1991) 263 13,261 9 4 . . . 3,384Indonesia (1990) 48 63 23 –9 . . . 31Malaysia (1989) 92 126 18 67 121 85Mexico (1990) 43 34 12 21 –159 282

Philippines (1992) 13 9 17 . . . . . . 13Sri Lanka (1991) 17 29 37 59 30 34Thailand (1988) 75 34 39 41 47 47

Sources: International Monetary Fund, International Financial Statistics,and World Economic Outlook.

1The year in parentheses next to each country respresents the first year ofthe surge in inflows. A minus sign indicates reserve losses.

2Does not include 1994.

13

ATTACHMENT III (International Capital Markets, August 1995)

whereas preannounced taxes on short-term inflowsavoid this stigma.25

In sum, shifting international capital flows can rep-resent large shocks to small open economies, occa-

sionally amounting to more than 10 percent of GDPin one year. The policy response to large and volatilecapital flows may require multiple instruments, in-cluding measures that seek to discourage capital in-flows or change their character, and coordination ofpolicies, monetary, fiscal, and exchange rate, to en-sure that recipient countries can derive benefits with-out incurring much of the costs.

119

Box 1. Restrictions on Capital Inflows and Prudential Requirements1

Brazil (1992)

October 1994.A 1 percent tax was imposed on foreigninvestments in the stock market. It was eliminated onMarch 10, 1995.

The tax on Brazilian companies issuing bondsoverseas was raised from 3 percent to 7 percent ofthe total. Eliminated on March 10, 1995.

The tax paid by foreigners on fixed interest invest-ments in Brazil was raised from 5 percent to 9 per-cent, and reduced back to 5 percent on March 10,1995.

The Central Bank raised limits on the amount of dollars that can be bought on foreign exchangemarkets.

Chile (1990)

June 1991.Nonremunerated 20 percent reserve require-ment to be deposited at the Central Bank for a periodof one year on liabilities in foreign currency for di-rect borrowing by firms.

The stamp tax of 1.2 percent a year (previouslypaid on domestic currency credits only) was appliedto foreign loans as well. This requirement applied toall credits during their first year, with the exceptionof trade loans.

May 1992.The reserve requirement on liabilities in for-eign currency for direct borrowing by firms wasraised to 30 percent. Hence, all foreign currency lia-bilities have a common reserve requirement.

Colombia (1991)

June 1991. A 3 percent withholding tax was imposed onforeign exchange receipts from personal servicesrendered abroad and other transfers, which could beclaimed as credit against income tax liability.

February 1992. Banco de la Republica increased itscommission on its cash purchases of foreign ex-change from 1.5 percent to 5 percent.

June 1992. Regulation of the entry of foreign currencyas payment for services.

September 1993. A nonremunerated 47 percent reserverequirement to be deposited at the Banco de la Re-publica on liabilities in foreign currency for directborrowing by firms. The reserve requirement is to bemaintained for the duration of the loan and applies

to all loans with a maturity of 18 months or less, ex-cept for trade credit.

August 1994. Nonremunerated reserve requirement to be deposited at the Banco de la Republica on liabilitesin foreign currency for direct borrowing by firms. Thereserve reserve requirement is to be maintained forthe duration of the loan and applies to all loans with amaturity of five years or less, except for trade creditwith a maturity of four months or less. The percentageof the requirement declines as the maturity lengthens;from 140 percent for funds that are 30 days or less to42.8 percent for five-year funds.

Indonesia (1990)

March 1991.Bank Indonesia adopted measures to dis-courage offshore borrowing. It began to scale downits swap operations by reducing individual banks’limits from 25 percent to 20 percent of capital. Thethree-month swap premium was raised by 5 percent.

October 1991. All state-related offshore commercialborrowing was made subject to prior approval by thegovernment and annual ceilings were set for newcommitments over the next five years.

November 1991. Further measures were taken to dis-courage offshore borrowing. The limits on banks’netopen market foreign exchange positions were tight-ened by placing a separate limit on off-balance-sheetpositions.

Bank Indonesia also announced that future swapoperations (except for “investment swaps” with ma-turities of more than two years) would be undertakenonly at the initiative of Bank Indonesia.

September 1994. Bank Indonesia increased the maxi-mum net open position from 20 percent of capital to25 percent, on an average weekly basis. Individualcurrency limits were no longer applied.

Malaysia (1989)

June 1, 1992.Limits on non-trade-related swap transac-tions were imposed on commercial banks.

January 17, 1994–August 1994. Banks were subject to aceiling on their non-trade- or non-investment-relatedexternal liabilities.

January 24, 1994–August 1994. Residents were prohib-ited from selling short-term monetary instruments tononresidents.

February 2, 1994–August 1994. Commercial banks wererequired to place with Bank Negara the ringgit funds. . .

1The year next to the country name denotes the first year ofthe surge in inflows.

25For further discussion of capital controls, see the backgroundpaper “Controls on Capital Flows: Experience with QuantitativeMeasures and Capital Flow Taxation,” pp. 95–108.

14

ATTACHMENT IV (World Economic Outlook, October 1996, p. xii)

120

The following “Declaration on Partnership for Sustainable Global Growth” was adopted at the con-clusion of the forty-seventh meeting of the Interim Committee of the Board of Governors of the IMF, September 29, 1996.

The Interim Committee has reviewed the “Declaration on Cooperation to Strengthen the Global Expansion,” which it adopted two years ago in Madrid.1 It notes that the strategy set out in the Declaration, which emphasized sound domestic policies, international cooperation, and global integra-tion, remains valid. It reiterates the objective of promoting full participation of all economies, includingthe low-income countries, in the global economy. Favorable developments in, and prospects for, manyindustrial, developing, and transition economies owe much to the implementation of sound policiesconsistent with the common medium-term strategy.

The Interim Committee sees a need to update and broaden the Declaration, in light of the new chal-lenges of a changing global environment, and to strengthen its implementation, in a renewed spirit ofpartnership. It attaches particular importance to the following:

• Stressing that sound monetary, fiscal, and structural policies are complementary and mutually re-inforcing: steady application of consistent policies over the medium term is required to establishthe conditions for sustained noninflationary growth and job creation, which are essential for so-cial cohesion.

• Implementing sound macroeconomic policies and avoiding large imbalances are essential to pro-mote financial and exchange rate stability and avoid significant misalignments among currencies.

• Creating a favorable environment for private savings.• Consolidating the success in bringing inflation down and building on the hard-won credibility of

monetary policy.• Maintaining the impetus of trade liberalization, resisting protectionist pressures, and upholding

the multilateral trading system.• Encouraging current account convertibility and careful progress toward increased freedom of cap-

ital movements through efforts to promote stability and financial soundness.• Achieving budget balance and strengthened fiscal discipline in a multiyear framework. Continued

fiscal imbalances and excessive public indebtedness, and the upward pressures they put on globalreal interest rates, are threats to financial stability and durable growth. It is essential to enhancethe transparency of fiscal policy by persevering with efforts to reduce off-budget transactions andquasi-fiscal deficits.

• Improving the quality and composition of fiscal adjustment, by reducing unproductive spendingwhile ensuring adequate basic investment in infrastructure. Because the sustainability of economicgrowth depends on development of human resources, it is essential to improve education andtraining; to reform public pension and health systems to ensure their long-term viability and en-able the provision of effective health care; and to alleviate poverty and provide well-targeted andaffordable social safety nets.

• Tackling structural reforms more boldly, including through labor and product market reforms, witha view to increasing employment and reducing other distortions that impede the efficient allocationof resources, so as to make our economies more dynamic and resilient to adverse developments.

• Promoting good governance in all its aspects, including by ensuring the rule of law, improving theefficiency and accountability of the public sector, and tackling corruption, as essential elements of a framework within which economies can prosper.

• Ensuring the soundness of banking systems through strong prudential regulation and supervision,improved coordination, better assessment of credit risk, stringent capital requirements, timely dis-closure of banks’financial conditions, action to prevent money laundering, and improved man-agement of banks.

The Committee encourages the Fund to continue to cooperate with other international organizationsin all relevant areas. It welcomes the recent strengthening of Fund surveillance of member countries’policies, which is an integral part of the strategy. It reaffirmed its commitment to strengthen the Fund’s capacity to fultill its mandate. It will keep members’efforts at achieving the common objectives of this strategy under review.

Partnership for Sustainable Global Growth

1See the October 1994 World Economic Outlook, page x.

xii

Comments on Selectivity andPresentational Issues

1. A general comment on the report concerns its se-lectivity in reporting criticisms of the Fund’s ap-proach to surveillance, and at times the substance ofFund policy advice. It is unclear what weight shouldbe given to such criticisms, for example, the refer-ence to “an academic observer’s” Criticism of theFund’s line on currency boards (Chap. III, para.21).The observer is cited as noting that the Fund had“bitterly opposed” a currency board in Indonesiawhile effectively “imposing” one in Bulgaria, with-out adequate explanation in his view. In fact, theFund had explained to the authorities in considerabledetail the reasons for the inappropriateness of a cur-rency board in Indonesia in early 1998 (includingshortage of foreign reserves, likely encouragementof capital flight, and the urgent need for extensivebank and corporate restructuring). It had also beenmade clear to the authorities that some form of fixedexchange rate arrangement could become appropri-ate at a later stage. This could have been acknowl-edged in the report.2. In several instances the report refers to a com-plaint by officials that IMF staff “were seen as com-ing with a preconceived framework” and with a“one-size-fits-all” approach (Chap. III, para. 6; Chap.IV, para. 15). The report notes that the impact ofFund advice was enhanced when missions were ableto “adapt the advice to the particular situation”(Chap. IV, para. 15). It would have been useful forthe report to elaborate on this and provide concreteexamples that describe the circumstances in whichFund staff recommended a set of policies that did nottake into account country-specific situations.3. The report does not delve sufficiently into thetreatment of statistical issues and the role statisticsplay in the context of the surveillance exercise. Thisis relevant in the context of the views expressed bysome small states that“too much time was devotedto data issues rather than policy advice” (Chap. III,para. 46). The report could usefully have providedsome general background information for a balanceddiscussion of statistical issues that have a bearing on

surveillance. This is particularly important given theBoard’s assessment that “the effectiveness of sur-veillance depend[s] critically on the timely availabil-ity of accurate information” and major institutionalinitiatives in this area in light of the crises in Mexicoand the Asian countries.1


4. The paragraph headed “Scope and Coverage”(Chap. III, para. 41) states that, in addition to tradi-tional macroeconomic demand-side topics, surveil-lance has become involved in microeconomic andsupply-side matters such as trade liberalization.This is a misunderstanding of trade liberalizationas addressed in Fund surveillance, which is basedon establishing and sustaining an outward, market-oriented policy environment consistent with an ap-propriate macroeconomic framework. This is fun-damental to the macroeconomic and tradeperformance of the great majority of Fund mem-bers. Article I of the Articles of Agreement sug-gests that this is central to the Fund’s purpose. Infact, the language of Chapter V (para. 29) seems toconfirm this view and contradict the earlier state-ment. The same argument would apply also to thestatements on tax policy and expenditure in thesame paragraph.5. The report’s references to the discussion of mili-tary spending in Fund consultations (Chap. I, para.25) might usefully have taken note of the limitationsthat the Board has placed upon the extent to whichthese issues can be discussed in the context of Arti -cle IV consultations. 6. The discussion of the Fund’s evaluation of equi-librium exchange rates (Chap. I, para. 32) does nottake into account the considerable work done in re-cent years by the Coordinating Group on ExchangeRates (CGER), reported, for example, in the Occa-sional Paper on exchange rate assessment by Isard

Annex II Other Comments

121

1See IMF Annual Report 1998, “Strengthening the Interna-tional Architecture,” p. 35.

and Faruqee.2 While not wishing to oversell the roleof CGER, the Fund’s evaluations of exchange rateshave been used as a basis for specific policy recom-mendations and public statements.3

7. The statement that there is “no attempt by theFund itself to develop a comprehensive set of policyrecommendations” in the context of the WEO is notaccurate. Since 1993, the first chapter of each WEOhas contained a comprehensive overview of policyissues and recommendations akin to the staff reportsfor Article IV consultations.8. Our perspective is that the policy assessmentsprovided through bilateral (Article IV) and multilat-eral (WEO, G-7 notes) surveillance are closely coor-dinated, with the Surveillance Committee chaired bymanagement playing the deciding role when thereare differences in view between area and functionaldepartments (Chap. II, para. 6). Moreover, the Sur-veillance Committee meets regularly to review bothArticle IV briefs and staff reports, but also the mainmultilateral documents, not just on an ad hoc basisas earlier suggested.9. With respect to the proposal that the Fund dis-continue its work in the area of bankruptcy legisla-tion (Chap. III, para. 39), our experience demon-strates that the development of an effectivebankruptcy regime is critical to the Fund’s effortsto strengthen members’financial systems and in-volve the private sector in the resolution of mem-bers’balance of payments difficulties. On the oper-ational level, the need for effective bankruptcylegislation has been recognized in a growing num-ber of Fund-supported programs. On the policylevel, the importance of the Fund’s work in thisarea has been recognized in the communiqués ofthe Interim Committee and in the decision estab-lishing the Fund’s policy on contingent credit lines.

Mission Frequency and the Reduction of Workload

10.We find the report overly optimistic about thepotential for a reduction of the scope of surveil-lance activities in order to relieve the workload onthe staff (Chap. V, para. 70). We are skeptical thatthe recommendation to better focus the surveil-lance process will have any significant effect on theoverall workload. In a world in which many factorsaffect macroeconomic policy, the list of topics thatmight be seen to be “directly relevant” (Chap. V,para. 29) may be a long one in any given case. Al -

though further progress can undoubtedly be madein collaborating with other international institutionsand in more clearly delineating the respective areasof responsibility, this progress is likely to be incre-mental. Thus, the recommendations of the reportrelating to more frequent and/or continuous sur-veillance, more regional surveillance, more discus-sion of capital account issues, more focus on policypriorities and tradeoffs, etc., would likely have sig-nificant implications for staff resources that wouldnot be offset by savings in other aspects of the sur-veillance process. These resource implications ofthe report’s recommendations should have been as-sessed and presented to the Board.

Role of the Executive Board andProposals for Restructuring theExecutive Board

11. On the Board Committee proposal, it is not evi-dent that committees would save (rather than useup) Directors’time (Chap. V, para. 110). In particu-lar, the proposal for committee reports would cer-tainly be labor intensive for someone. There wouldbe an efficiency gain at the end—that is, when thefull Board meets—only if at that stage Directors re-frained from extensive, prepared, presentations.But this idea has been floated before, and evenadopted, but never actually implemented. The two-step briefing process could similarly lead to furtherdemands on an already stretched Board and staff,and, if not strongly managed, it could lead to a pro-liferation of topics for the mission, rather than anarrowing.12.The proposal that the Board approve the set oftopics to be discussed by missions would be a signif-icant departure from the current division of responsi-bility between management and Board (Chap. V,para. 32), and would create an anomalous situationwhere policy issues facing a country would be dis-cussed with the Board before they are discussed withthat country’s authorities.

Legal Issues

13.The report should not have referred to surveil-lance as a form of technical assistance or develop-ment aid (Chap. I, para. 8). Surveillance is distinctfrom the provision of technical assistance, and is nota form of development aid.14. It should also be noted that Fund missions maymeet with nongovernmental representatives in mem-ber countries only with the consent of the relevantauthorities (Chap. V, para. 83).

122

ANNEX II OTHER COMMENTS

2Peter Isard and Hamed Faruqee, eds., Exchange Rate Assess-ment, IMF Occasional Paper No. 167 (Washington: IMF, 1998).

3See the chapter by Kahn and Nord in Isard and Faruqee, eds.

123

Staff Recruitment

Staff Recruitment

15.Coming on the heels of the external evaluationof research in the Fund, we note some interestingdifferences in views on what the Fund’s prioritiesshould be in recruiting staff (Chap. II, para. 47 andChap. V, paras. 86 and 87). Not surprisingly, theevaluation of research stressed the importance ofrecruiting staff with strong interests in research. Onthe other hand, one of the recommendations of thisevaluation is that the Fund should place more em-phasis on recruiting staff with policy experience,and less on academic qualifications, at all levels.This seemingly contradictory advice points to the

importance of the Fund setting priorities and orga-nizing itself against the background of a strategicvision of its role in the international community.Evaluations—both internal and external—thatfocus too closely on one particular activity maylose sight of this. In our recruitment of experiencedeconomists, who account for well over half theFund’s hiring of economists, the emphasis is onpolicy experience already. The suggestion to placemore weight on policy experience in the recruit-ment of Economist Program participants is simplynot realistic: the average age of Economic Programentrants is 29; it is very difficult to have substantialrelevant policy experience by that age.

external evaluation of imf surveillance · 2000. 11. 22. · part 2 external evaluation of...

Documents