STAFF PAPERS

PETER HOLE Chair, Editorial Committee

lAN S. McDoNALD Editor and Deputy Chair

MARINA PRIMORAC Assistant Editor

Editorial Committee

F. Charles Adams Malcolm D. Knight Mario I. Blejer Pau 1 R. Masson David Burton Donald J. Mathieson Daniel A. Citrin Susan M. Schadler David J. Goldsbrough Subhash M. Thakur Peter Isard Howell H. Zee G. Russell Kincaid

Among the responsibilities of the International Monetary Fund, as set fonh in its Anicles of Agreement, is the obligation to "act as a centre for the collection and exchange of information on monetary and financial problems." Staff Papers makes available to a wider audience papers prepared by the members of the Fund staff. The views presented in the papers are those of the authors and are not to be interpreted as necessarily indicating the position of the Executive Board or of the Fund. To facilitate electronic storage and retrieval of bibliographic data. StaffPapers has adopted the subject classification scheme developed by the Journal of Economic Literature.

Subscription: US$54.00a volume or the approximate equivalent in the currencies of most countries. Four numbers constitute a volume. Single copies may be purchased at $18.00. Individual academic rate to full-time professors and students of universities and colleges: $27 a volume. Subscriptions and orders should be sent to:

International Monetary Fund Publication Services 700 19th Street. N.W. Washington, D.C. 20431, U.S.A.

Telephone: (202) 623-7430 Telefax: (202) 623-7201 Internet: [email protected]

©International Monetary Fund. Not for Redistribution INTERNATIONAL MONETARY FUND STAFF PAPERS

Vol. 42 No. 4 DECEMBER 1995

©International Monetary Fund. Not for Redistribution EDITOR'S NOTE

TI1e Editor invites from contributors outside the Fund brief comments (not more than 1 ,000 words) on published articles in StaffPapers. These com­ ments should be addressed to the Editor. who will forward them to the author of the original article for reply. Both the comments and the reply will be considered for publication.

The term "country," as used in this publication, may not refer to a territorial entity that is a state as understood by inter­ national law and practice; the term may also cover some ter­ ritorial entities that are not states but for which statistical data are maintained and provided internationally on a separate and independent basis.

© 1995 by the International Monetary Fund International Standard Serial Number: ISSN 0020-8027

The U.S. Library of Congress has cataloged this serial publication as follows:

International Monetary Fund Staff papers - Imernational Monetary Fund. v. 1- Feb. 1950- [Washington] International Monetary Fund.

v. tables. diagrs. 23 cm.

Threeno. a year, 1950-1977; four no. a year. 1978- lndexes: Vols. 1-27, 1950-80. I v.

ISSN 0020-8027 - Staff papers- lmemationaJ Moncwry Fund. I. Foreign exchange-Periodicals. 2. Commerce-Periodicals. 3. Currency question-Periodicals.

HG3810.15 332.082 53-35483

©International Monetary Fund. Not for Redistribution CONTENTS

Vol. 42 No. 4 December 1995

SPECIAL SECTION Celebrating Fifty Years of the InternationaJ Monetary Fund

Early Contributions of StaffPapers to lntemational Economics

MARIO I. BLEJER, MOHSIN S. KHAN, and PAUL R. MASSON • 707

Fifty Years of Exchange Rate Research and Policy At the lntemational Monetary Fund

JACQUES J. POLAK • 734

The Historical Development of the Principle of Surveillance

HAROLD JAMES • 762

Conditionality: Past, Present, Future

MANUEL GUITIAN • 792

Wage Structures in the Transition of the Czech Economy

ROBERT J. FLANAGAN • 836

Pricing to Market and the Real Exchange Rate

HAMID FARUQEE • 855

Shorter Paper and Comment

Unemployment Hysteresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium

REZA MOGHADAM and CAROLINE VAN RIJCKEGHEM • 882

German Unification and Asymmetry in the ERM: Comment on Gardner and Perraudin

JEROME HENRY and JENS WEIDMANN • 894

IMF Working Papers • 903

Volume 42 Index • 905

iii

©International Monetary Fund. Not for Redistribution Thispage intentionallyleft blank

©International Monetary Fund. Not for Redistribution Foreword

To mark the fiftieth anniversary of the International Monetary Fund, the Editorial Committee of Staff Papers commissioned four articles look­ ing back on the evolution of thought, research, and practice in the Fund in core areas of the institution's responsibilities and concerns. These papers are presented in this issue. The first paper provides a historical assessment of the Fund's research activities by analyzing the contribution of seminal articles published in StaffPapers in the journal's early decades. It was written by Mario Blejer, Mohsin Khan, and Paul Masson, three current staff members ofthe IMF. The second paper, on exchange rate policy and research, was contributed by a former Economic Counsellor of the Fund, Jacques Polak, who spent 33 of the institution's first 50 years as a staff member and remains closely associated with the institution to this day. The third paper, on the histor­ ical development of the principle of surveillance, was written by an outside observer, Harold James, Professor of History at Princeton Uni­ versity, who recently completed a comprehensive study of the man­ agement of the postwar international monetary system, International Monetary Cooperation Since Bretton Woods, published jointly by the IMF and Oxford University Press. In the fourth paper, Manuel Guitian, Director of the IMF's Monetary and Exchange Affairs Department, discusses the history of, present challenges to, and future prospects for lMF conditionality. The papers deal variously with conceptual and operational aspects of the Fund's work. There are, of course, linkages and overlaps between the activities covered in the four papers, and the demarcation between them is, in some respects, necessarily somewhat arbitrary. This reflects the complementarity of these activities as instruments that enable the Fund to fulfil) its assigned purposes, namely, to promote international mon­ etary cooperation, to facilitate the expansion and balanced growth of international trade, to promote exchange stability, to assist in the estab­ lishment of a multilateral system of payments, to give confidence to members by making resources available to them under adequate safe­ guards, and to shorten the duration and lessen the degree of disequi­ librium in the balances of payments of member countries. As the papers bring out, these purposes are still as relevant today as they were when the IMF was established.

V

©International Monetary Fund. Not for Redistribution Thispage intentionallyleft blank

©International Monetary Fund. Not for Redistribution IMF StaffPapers No. 4 Vol. 42,International (December Monetary 1995) Fund (' 1995

Early Contributions of Staff Papers to International Economics

MARIO I. BLEJER, MOHSIN S. KHAN, and PAUL R. MASSON*

Staff Papers has, since its inception in 1950, been an important vehicle for the dissemination of research done by staff of the IM The paper discusses three areas in which articles published in Staff PapersF. up until the 1970s made major contributions to the literature in intemational eco­ nomics. The areas covered are: first, the absorption app roach and the monetary theory of the balance ofpayments; second, the Munde/1-Fieming model; and third, foreign trade modeling. The nature of the contributions and their relationship with further developments in the respective fields are detailed. (JEL B20]

HE INTERNATIONAL MONETARY FUND'S StaffPap ers first appeared in T1950, apparently in response to a suggestion by Dennis Robertson that such a journal would prevent the premature burial of bright young economists in "an anonymous international bureaucracy" (Polak (1995)). Especially in its early days, the journal served as the main outlet for publication of economic research done by the staff. As such, it reflected-and continues to reflect-the preoccupations and operational needs of the Fund. 1 In particular, a large number of articles in the journal

• Mario I. Blejer is Senior Advisor in the Monetary and Exchange Affairs Department. He is a graduate of the Hebrew University of Jerusalem and holds a Ph.D. from the Umversity of Chicago. Mohsin S. Khan is Deputy Director of the Research Department. He is a graduate of Columbia University and the London School of Economics and Political Science. Paul R. Masson is Assistant Director in the Research Department and was educated at McGill University (B.A.) and the London School of Economics (Ph.D.). The authors are grateful to Manuel Guitian, Jacques Polak, Assaf Razin, as well as other colleagues in the IMF for helpful comments. 1 At the same time, the articles in have also closely reflected the prevailing trends in the profession atStaff large.Papers 707

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER, MOHSIN S. KHAN, PAUL R. MASSON 708 and have been devoted to international monetary economics in its various aspects, to the transmission mechanisms of economic policies, and to practical questions relating to macroeconomic, particularly monetary and fiscal, policies in both industrial and developing economies. It has differed from academic journals in that most articles in Staff Papers are generally grounded in the IMF's operational work and reveal the expe­ rience gained by the staff in the course of their work. Despite this focus on the practical aspects of economic policy, Staff Papers articles have also made major empirical and theoretical contributions to the economics literature. One way to assess the impact of Staff Papers articles on the literature is through a numerical tabulation of citations. A recent article in the Economic Journa/2 uses this technique to rank journals, while making adjustments for "inputs'' (e.g., the number of articles in an issue and the number of issues), and finds that Staff Papers is one of the "core jour­ nals." In one ranking, it appears just above the Journal of Economic Theory, and just below the Journal of International Economics. 3 While such exercises should only be taken as indicative, since the conclusions that result are quite sensitive to methodology, these findings do suggest an important role for Staff Papers in the economics profession. In this paper, we concentrate on three areas where Staff Papers articles have, in our view, made especially noteworthy contributions. These areas are first, the absorption approach and the monetary theory of the balance of payments; second, the Mundell-Fleming model of international mon­ etary and fiscal policy transmission under high capital mobility; and third, mode ling of foreign trade relationships. In the first area, seminal articles such as Alexander (1952) and Polak (1957) were followed by a number of others that also appeared in the journal, while the second area is represented by two articles, the Fleming (1962) version of the Mundell­ Fkming mudd, ami a Mumldl (1962) article un the assignment problem. The third area is more diffuse, with a series of important contributions to empirical estimates of trade elasticities (Liu (1954), Junz and Rhom­ berg (1965), and others) and to the theoretical understanding and empir­ ical underpinnings of multilateral trade models (Armington (1969a and 1969b), and others). In singling out these articles, which develop approaches that emerged prior to 1970, we have deliberately avoided making an assessment of the

2 Burton and Phimister (1995). 3 Another recent ranking of journals has been published in the Journal of Economic Literarure (Laband and Piette (1994)). However. this study does not include Staff Papers in its data set.

©International Monetary Fund. Not for Redistribution EARLY CONTR1BUTLONS OF STAFF PAPERS 709

importance of more recent contributions. Without the benefit of hind­ sight, evaluation of the more recent literature is more difficult. However, areas in which Staff Papers articles may well be considered in coming years to have made seminal contributions include exchange rate modeling in the 1970s, third-world debt issues and adjustment policies in the 1980s, and policy credibility in the 1990s.

The Monetary Approach to I. Balance of Payments Determination

At the beginning of the 1970s there was a strong upsurge of interest, both in academic circles and among policymakers, in the examination of monetary relations in an open economy, and in the analysis of the interactions between the behavior of monetary aggregates and the deter­ mination of the balance of payments. This interest led to the formaliza­ tion of what became known as ·'the monetary approach to the balance of payments," an approach that, from the mid-1970s until the mid-1980s, took center stage in the theoretical and empirical debates that char­ acterized the open-economy macroeconomic literature of the period, and in the discussions surrounding policy implementation in a variety of countries. By 1975 the monetary approach was well established as a more realistic alternative both to the "Keynesian" approach-seen as concentrating solely on the trade balance and as excessively concerned with the inter­ relationships between international adjustment and domestic employ­ ment-and to the "elasticities" approach-widely regarded as a partial equilibrium response to questions more of a general equilibrium nature. In the next five years, a vast number of dissertations, journal articles, and working papers were written dealing with the analytical underpinnings, the implications, the empirical validation, and the possible extensions of the approach. Despite the evident intellectual success that the monetary approach was enjoying at that time, a significant amount of skepticism remained in the profession, and the approach stimulated a number of formal theoretical and methodological critiques.4 However, despite the criti­ cisms it generated, the monetary approach has had a lasting and conse­ quential influence on macroeconomic thought. This is so because the most relevant and meaningful elements emphasized by the monetary

for example, Hahn (1977). 4 See,

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER. MOHSIN S. KHAN. PAUL R. MASSON 710 and approach have been widely accepted and. in practice. have become conventional wisdom in today's theoretical and policy-oriented analysis of the balance of payments and of macroeconomic relations in an open economy. As the monetary approach was gaining popularity, a noticeable, albeit amiable, scholarly debate about its origins and '·ownership'' arose. While all those involved in the development of the approach routinely ac­ knowledged that it has a long intellectual history originating with the eighteenth-century contributions ofDavid Hume,5 credit for the modern revival of the monetary approach was simultaneously claimed in two circles of the profession. The academic version of the monetary approach seems to originate with the writings of James Meade in the early 1950s and to continue with the contributions of Harry G. Johnson and Robert A. Mundell in the 1960s.6 It reached its peak of popularity during the early 1970s with the intense work of economists associated with the Workshop in International Economics at the University of Chicago. Years before the standard exposition of the monetary approach arising from these academic contributions became prominent, a number of important analytical and empirical studies were carried out at the IMF, largely under the direction of Jacques Polak. Many of these studies were intended to yield analytical foundations to the Fund's practices and, in many ways, were geared to the IMF's operational work. However, these studies greatly promoted the subsequent development of a rigorous mon­ etary framework for the examination of balance of payments perfor­ mance that became, essentially, the forerunner to the theories that emerged later, in a more refined and robust formulation, in the academic literature. Although the precursory role played by the JMF studies is indeed customarily acknowledged in the classical academic references of the 1970s (e.g., Frenkel and Johnson (1976)). the IMF research was not necessarily considered as setting the base for the subsequent develop­ ments of the monetary approach but rather was characterized as a "short­ Jived burst of interest in the monetary-theoretic aspects of the balance of payments at the International Monetary Fund in the late 1950s ...." Moreover, Harry Johnson believed that the IMF work was disre­ garded because of the dominance of Keynesian views in the 1950s and "the relative impotence and disrepute of the Fund as an international

�Curiously enough, however, Harry Johnson, one of the most prominent exponents of the monetary approach, though never denying its origmal roots, referred to it frequently as 'the new approach: See, for example, Johnson ( 1975). 6See Johnson (1958 and 1972). Meade (1951). and M undell (1968).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 711

monetary institution at that time'' (both points are made in Johnson (1977), p. 261). Despite this dismissal of IMF influence on economic thinking, practi­ cally all the subsequent literature on the monetary approach has tended to relate it to the absorption approach, and in particular to the version of that approach published in StaffPapers by Sidney Alexander in 1952. The monetary approach is sometimes viewed as a rival of the absorption approach, and sometimes as complementary to it. In Alexander's (1952) paper the current account of the balance of payments is viewed as the difference between total output and total expenditures. and therefore the effect of specific policies on the external position of a country can be evaluated by assessing the relative impact of these policies on produc­ tion and spending.7 Although the Alexander formulation basically uti­ lizes a Keynesian formulation, it emphasizes an adjustment mechanism based on the effects of changes in the real value of financial wealth on the rate of accumulation and decumulation of foreign assets through the effects of those changes on absorption. In particular, Alexander analyzes in detail the consequences of a devaluation on the balance of payments through the effects of the devaluation on domestic prices. 8 As prices rise, the real purchasing power of cash and bonds held by the public is re­ duced and so is the propensity to consume (and, therefore, total ab­ sorption), leading to an improvement in the balance of payments. The important points of this model are that the success of an exchange rate adjustment depends on monetary factors and that the analysis utilizes monetary concepts. ln substance, these concepts are analogous to those that became the fundamental blocks of the monetary approach.9 The building up of these fundamental blocks was an essential preoccu­ pation of the Fund staff by the mid-1950s, a preoccupation that was strongly reflected in Staff Papers at the time. The seminal article by

7The connection between balance of payments developments and the flows of income and expenditures was a conceptual analytical tool in the Fund practically since its Lnception. This was already reflected in the first volume of StaffPapers by Tsiang's (1950) article, particularly in the section on causal relationships between mternational transactions and domestic income. 8These concepts were further advanced and elaborated in many of the Fund's operational documents and unpublished papers. A summary of the Fund's views regarding balance of payments adjustment by the mid-1950s is presented in Bernstein (1956) where he also compares the elasticities and the absorption aprroaches. . It is interesting to note that a year before, Alexander published in StaffPapers a pioneering exposition of the optimal tariff argument m which the effects of a devaluation on the trade balance are compared with those of the imposition of trade restrictions. That anal sis, however, does not incorporate explicit monetary principles. See Alexander y ( 1 951 ).

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER, MOHSIN S. KHAN, PAUL R. MASSON 712 and

Polak (1957) is broadly regarded as the starting point of the formal research on the monetary approach in the Fund, but there were a number of articles published in previous StaffPapers issues that covered specific elements of the monetary approach that were later combined into a single analytical framework. These include Polak and White (1955) on the endogeneity of money in an open economy model; Hicks, Dorrance, and Aubanel (1957) on the distinction between internal and external sources of monetary expansion; and Holtrop (1957) on the principle that only domestic credit expansion can be treated as an autonomous variable.

Essence of the Approach

While there are some evident differences between the earlier Fund work and the academic versions of the monetary approach, there is no doubt that both variants deal with a very similar question and use com­ parable methodologies. The essence of the approach is an analytical formulation that emphasizes the interaction between the supply and the demand for money in determining the country's overall balance of pay­ ments position. It could be seen, in fact, as an extension to the open economy of the conventional closed-economy monetary models, which highlight the stability of the money demand function and assess the consequences for the economy of changes in the money supply under different conditions. When the expansion of the money supply is not consistent with an equivalent change in the demand (and vice versa), a stock disequilibrium in the money market arises, which affects the spend­ ing patterns of economic agents. When the money supply grows faster than the demand, the excess flow supply of money so generated gives rise to a corresponding excess demand for goods and nonmonetary financial assets. In a closed economy the disequilibrium in the money market is eliminated by a combination of increases in prices, interest rates, and possibly output. These changes raise the nominal demand for money to a level commensurate with the new money stock, thus restoring monetary balance. Unlike in a closed economy, in an economy open to trade and financial flows, changes in the supply of domestic money can be caused by domes­ tic credit creation and the foreign-exchange activities of the central bank. In these circumstances, the monetary approach to the balance of pay­ ments emphasizes that money market disequilibria are reflected not only in changes in nominal income (prices and output) but also in the country's foreign reserve stocks. Therefore, the approach concentrates on the

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 713

relationships between the supply and demand for money, on the one hand, and prices, output, interest rates, and the balance of payments, on the other. An important implication of this analysis is that, in a regime of fixed exchange rates, the aggregate money supply is beyond the direct control of the monetary authority and is rendered endogenous. The central bank, however, retains control over the volume of credit, one of the sources of monetary expansion. Within the framework of the monetary approach to the balance of payments, the distinction between the monetary base and its domestic-credit component becomes central: tbe monetary authority can control the latter but not the former. For a given expansion of the demand for domestic real balances, an equivalent growth in the money supply can be realized through a suitable increase in domestic credit. However, when the rate of domestic credit creation diverges from the changes in money demand, the difference is made up by equivalent changes in net foreign assets arising from a balance of payments surplus or deficit.

The Contribution of StaffPapers As mentioned above, the preoccupation of staff members with mone­ tary issues in an open economy was reflected in their contributions to Staff Papers in the late 1950s and over the 1960s. Polak's 1957 article, widely regarded as the first general equilibrium model formally designed to analyze balance of payments problems in a monetary setting, assumes exports, capital movements, and domestic credit creation to be deter­ mined exogenously. The values of nominal income, imports, and money are the direct result of the behavior of domestic credit policies and of the path followed by the exogenous elements in the balance of payments (exports and capital flows). While its results are akin in nature to those obtained in the subsequent academic literature, the Polak model em­ phasizes in more detail the dynamics of the adjustment path and there­ fore does not concentrate exclusively on the long-run consequences of monetary policies for the external sector. The largely theoretical contribution of Polak was empirically comple­ mented by the application of the model to 39 countries by Polak and Boissoneault (1960). This paper demonstrates that, using Polak's model, it is feasible to predict the behavior of imports using monetary data and, therefore, that it is possible to calculate reasonably well the effects of various domestic-credit policies on the balance of payments. This theo­ retical point is reinforced in Guitian's (1973) paper, in which an analytical

©International Monetary Fund. Not for Redistribution MARIO 1. BLEJER. MOIISIN S. KHAN. PAUL R. MASSON 714 and framework is developed to demonstrate that domestic credit is the only appropriate monetary instrument for the control of the balance of pay­ ments in a fixed exchange rate system. An empirical investigation similar to Polak and Boissonneault (1960) was published by Fleming and Boissonneault (1961) but it focuses more specifically on the relationship between money and imports. Methodo­ logically, the Fleming-Boissonneault paper studies the Jag structure of the monetary effects in the context of the Polak's model. The relationship between money and imports was further disaggregated by Kanesathasan (1961), who considers the relationship between government imports and tariffs within the context of the monetary model. Polak's framework was further formalized by Prais (1961), who refor­ mulates the analysis in continuous time. Prais's expenditure function includes a term reflecting the deviations of actual from desired monetary holdings. This term provides an explicit linkage between the real and the monetary sectors of the economy and was extensively adopted in the later formulations of the monetary approach .w While the original contribu­ tions to the formulation of the monetary approach were presented in nominal terms, Argy (1969) recast the model in real terms by assuming constant prices. He showed how real income objectives and balance of payments targets could be reached by appropriately setting the two avail­ able policy instruments (government expenditures and money supply). After 1970, writings on subjects related to the monetary approach mushroomed and numerous articles, in StaffPapers and elsewhere, ex­ tended and tested the approach. The publication in 1977 a collection of papers on the monetary approach written in the Fundof after 195711 helped to highlight the contribution of the Fund staff to the development of the model and increased the visibility of articles previously published in Staff Papers.12 The basic models collected in the 1977 volume were the subject of many empirical applications and, also, of a significant number of extensions. 13

m see, for example, Dornbusch (1976a). Prais's article is the only early article that is quoted in the various papers included in the FrenkelStaff and PapersJohnson volume.

1211IMF (1977). It is worth noting that after the publication of the volume, the articles included in the book were mostly quoted as part of the collectionStaff anP aped notrs as ori�inal articles. On the one hand. this, of course, may have reduced the visibility of but on the other, the successful reception of the book may have inStaffcreasedP aper thes, exposure of the papers. 13 Since the intention of this paper is not to provide a review of the monetary approach to the balance of payments but rather to trace the impact of the early contributions. no specific references are provided here. For a survey of the monetary approach literature. see Kreinin and Officer (1978).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF 715 STAFF PAPERS

An important extension was the relaxation of some of the limiting assumptions. For example, if the degree of international capital mobility is low and if the share of nontraded goods is high, then the speed of adjustment to monetary disturbances is reduced. In the short run, there­ fore, monetary imbalances also affect domestic prices, output. and inter­ est rates, and the relative importance of these effects depends on various factors such as the nature of the exchange rate regime, the degree of openness of the economy in both the goods and the capital markets, the degree of resource utilization, and the extent of nominal and real wage rigidities. Many of these elements were modeled within the original monetary framework. as were the effects of exogenous changes in income growth and of external shocks. The basic model was also extended to analyze the consequences of once-and-for-all devaluations and of the abandonment of the fixed ex­ change rate assumption. Monetary research on exchange rate determina­ tion in a flexible exchange rate system was seen as the logical counterpart to the original monetary approach formulation. Such research was, in the second half of the 1970s, carried out simultaneously at the Fund and in academic institutions. As a logical sequel to their previous volume, Frenkel and Johnson (1978) collected a number of Chicago studies on the subject.14 The main theme of the various papers is that the exchange rate, being a relative price of two national monies, is determined primarily by the relative supplies and demands for these monies and that the analysis of exchange rate determination should not be conducted in the partial­ equilibrium framework of foreign trade but rather within a general macroeconomic framework. Despite the appeal of these extended formulations, both the monetary approach to the balance of payments under fixed exchange rates and the monetary approach to the exchange rate lost ground in the mid-1980s. While a detailed analysis of the reasons for these developments is not relevant here, the impact of the debt crisis after 1982 (with the consequent intensification of the phenomenon of currency substitution) and the seemingly exogenous upsurge of capital inflows into emerging markets in the 1990s helped weaken the credibility of some of the central tenets of the approach, particularly as regards the endogeneity or exogeneity of the various monetary aggregates. However, it remains true that the major propositions brought to the fore by the monetary approach (e.g .. the importance of money demand in balance of payments analysis) have been largely incorporated into conventional macroeconomic thinking and that

1" See also Bilson 1978). (

©International Monetary Fund. Not for Redistribution 716 MARIO I. BLEJER, MOHSIN S. KHAN, and PAUL R. MASSON the earlier contributions of the Fund staff, as reflected in publications in Staff Papers , played a central role in this process.

The Mundeii-Fleming Model: Capital Mobility, Exchange 11. Rate Flexibility, and Policy Assignment

A major influence on thinking about international macroeconomic policy issues that appeared, at least in part, in Staff Papers , was what has become known as the Mundell-Fieming model, after Robert Mundell and J. Marcus Fleming, who were then at the IMF (Fleming (1962) and Mundell (1962)).15 These two authors considered the effectiveness of monetary and fiscal policies in the context of capital mobility, and under the alternative assumptions of exchange rate flexibility or fixity. Not only did they isolate strong policy conclusions-for instance, that fiscal policy would be more effective in increasing output under fixed rates than with flexible rates; and that the monetary policy should be assigned to external balance, and fiscal policy to internal balance-Mundell and Fleming also created a streamlined, elegant, and easily mastered model that was quickly adopted by practicing economists and taught to generations of graduate students. Moreover, the model has led to numerous empirical applications and extensions and still serves as the backbone of a host of estimated multicountry models. In a discipline (economics) where often only recent precursors are mentioned in academic articles, the Mundell-Fleming model is exten­ sively cited despite the passage of more than thirty years. Indeed, Jacob Frenkel and Assaf Razin, in their 1987 review article, termed it "still the 'workhorse' of traditional open-economy macroeconomics" (Frenkel and Razin (1987b), p. 568). In the influential 1985 North-Holland vol­ ume, Handbook of International Economics: International Monetary Economics and Finance (Jones and Kenen (1985)), the Mundeii-Fieming model is mentioned prominently in many of the chapters, ranging from the ones on stabilization policy and on exchange rate dynamics, to that on economic interdependence and policy coordination. The explanation for the longevity of the Mundell-Fleming model's contribution may well be the feature Dornbusch highlights when discussing Mundell's work of the early 1960s: "he created simple, forceful models to serve as organiz-

• 15Tbough only the Fleming half of the Mundeii-Fleming model was published m Staff Papers , the Staff Papers article by Mundell cited above is on the assign ­ ment problem in a world of capital mobility, which is closely related. Other articles by Mundell in this broad subject area appeared in various journals (Mundell (1960, I96la. 196lb, 1963, and 1964)).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 717

ing frameworks for thought and policy and as springboards for posing new problems" (Dornbusch (1980)).

The Nature of the Contribution

The Mundell-Fleming model was a major step in the process of turning closed-economy macroeconomics into open-economy macroeconomics. Most earlier treatments of the interdependence between countries had concentrated on trade linkages, and, in particular, on the conditions for a positive effect on the trade balance of devaluation (the Marshall-Lerner condition) or on the size of the foreign trade multiplier. These analyses were often based on partial equilibrium approaches, and neglected mon­ etary factors. Moreover, capital flows typically got little attention; in­ stead, the focus was on policies operating in the context of Umited capital account convertibility. By the early 1960s, Meade's 1951 book had already extensively consid­ ered financial policy issues in an open economy, general equilibrium macroeconomic setting, and the absorption approach discussed above had analyzed the domestic savings-investment requirements for achiev­ ing external balance. However, capital flows do not appear as promi­ nently here as in the later Mundeii-Fieming work, and their implications are not highlighted to the same extent. In Kenen's words: "Meade (1951, eh. 15) was careful to include them [capital movements], but they could be deleted without altering his argument" (Kenen (1985), p. 636). In contrast, in MundelJ-Fieming, they are a central part of the story. No doubt, the greater attention to the subject reflected a decade of enormous expansion of private capital flows, including the move to widespread (current account) convertibility in Europe and the development of the EurodolJar market. Clearly, by 1960, private capital flows were much more important than bad been imagined when the postwar international monetary system was designed, at Bretton Woods in 1944. It is also interesting that an examination of the consequences of flexible exchange rates was undertaken at this time at the IMF, which after all was a pillar of the Bretton Woods system of fixed rates. Mundell's interest in the fixed/flexible rate comparison under highly mobile capital no doubt also reflected his knowledge of the Canadian experience of flexible rates in the 1950s (and Canada's return to a pegged rate in 1962). Another difference from most earlier contributions (but not the Polak work cited above) is the way money market equilibrium is taken into account when modeling the transmission process. In the partial equi­ librium multiplier analysis and the Marshall-Lerner conditions, monetary

©International Monetary Fund. Not for Redistribution 718 MARIO I. BLEJER, MOHSIN S. KHAN. and PAUL R. MASSON factors are given no explicit role. In Meade's general equilibrium ap­ proach, a "neutral" monetary policy is one that leaves interest rates unchanged; thus, there is very little interaction between monetary and fiscal policies. In contrast, the Mundeii-Fleming model highlights very starkly the linkages between money market equilibrium, interest parity, and the exchange rate. If monetary policy keeps the money supply unchanged under flexible rates, then an expansionary fiscal policy tends to increase interest rates and thereby appreciate the exchange rate, and complete crowding out must occur if interest rates are exogenously given by the rest of the world (provided that money demand takes a simple form). Under fixed rates, reserve inflows increase the money supply, and the extent of crowding out is reduced. The way MundeU and Fleming analyze money market equilibrium no doubt reflected the growing interest in the demand for money as a central relationship for analyzing balance of payments and financial program­ ming problems. The Mundell-Fleming model thus has common elements both with the earlier Polak (1957) model and the later academic version of the monetary theory of the balance of payments (see the discussion in Section I above). The work of MundeiJ and Fleming at the Fund also paralleled the revival of interest in money demand and monetary issues elsewhere, in particular at the University of Chicago-most notably associated with Milton Friedman. A third innovation-or rather, refinement of existing practice-was in the method of analysis. Whereas Meade presented his main argument verbally, and relegated the mathematics to a largely impenetrable supple­ ment in which the equations were handwritten, Mundell and Fleming make models a central part of their analysis. Clearly others (Hicks (1939) and Samuelson (1947), not to mention Walras (1954) and Pareto (1909)) had already gone further in sophisticated and elegant mathematical pre­ sentations for the specialist; however, now models were used in a prac­ tical, policy-relevant context for the mainstream economist. For instance, the appendix to Flemiog (1962) clearly sets down equations and compar­ ative static results, so that assumptions and methods needed to prove the results on policy effectiveness under fixed and flexible rates are not subject to dispute or misinterpretation. No doubt, the example of Mundell and Fleming stimulated interest in international problems and their model, because of its clarity, served as a standard for later work. Though graphical analysis was still in much use (including, for instance, in Mundell (1962 and 1963)), by the next decade new contributions to international macroeconomic analysis relied almost exclusively on mathematical techniques.

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 719

Later Extensions of the Mundell-Fleming Model A strength but also a weakness of the model is its simplicity, which made it important to check whether the main conclusions of Mundell and Fleming's work-in particular, concerning the effectiveness of monetary and fiscal policies-were robust. Later extensions of the model read like an inventory of the major advances in international macroeconomics of the past few decades: modeling of capital flows as a stock adjustment (Branson (1970)); an explicit integration of sticky (not rigid) domestic prices and the inclusion of exchange rate expectations (Dornbusch (1976b)); asset stock dynamics and portfolio balance effects on the un­ covered interest parity relationship (Branson (1976); Kouri (1976); Dornbusch and Fischer (1980)); allowing money demand to depend on the exchange rate, through either the effect of import prices on the absorption deflator or exchange rate valuation effects on wealth (Bran­ son and Buiter (1983)); and incorporation of intertemporal budget constraints for the government and for households, including modeling consumption as the result of intertemporal optimization (Bianchard (1985); Frenkel and Razin (l987a)). Generally speaking, these extensions have mad.- the Mundeii-Fleming model richt::r. thus allowing it to be applied to more inle1esting yues­ tions-for instance, the effects of temporary versus permanent changes in financial policies or in the terms of trade-without overturning its central insights at least concerning short-run effects. For instance, the ineffectiveness of monetary policy under fixed rates was shown to depend on perfect capital mobility and on the inability to sterilize reserve flows, which empirical work has to some extent contradicted. However, the limiting case studied by Mundell and Fleming is still a useful benchmark. Likewise, the ineffectiveness of fiscal policy under floating rates no longer strictly holds if the exchange rate affects money market equi­ librium; for example, an exchange rate appreciation may lower the de­ mand for money if the price of domestic absorption (rather than the GDP deflator) appears in that demand function, so that for a given supply of money, output can rise (Branson and Buiter (1983)). Nevertheless, an important insight of the Mundeii-Fleming model remains, namely, that under flexible rates fiscal expansion can be expected in normal cases to lead to crowding out through exchange rate appreciation.16 16Fieming (1962) �c�no�ledges �hat some might consider t�e possibility of exchange rate apprec1at10n m these c1rcumstanccs as an "academ1c curioswn ··but goes on to point out that Rhomberg's model of the Canadian economy gives just that result. Today, wit_h the experience of _the R_eagan fiscal expansion and German umficat10n_ behmd us, few would cons1der th1s case an academic curiosity.

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER. MOHSIN S. KHAN, PAUL R. MASSON 720 and

Empirical Applications

Just as the extended Mundell-Fleming model remains the "workhorse" model for theorizing, it also serves as the intellectual underpinning for the empirical models most widely used to study international economic policy questions. Indeed, the current crop of multicountry macroeco­ nomic models, those that incorporate forward-looking expectations and typically model nonaccommodating monetary policy in terms of keeping a monetary aggregate close to its target value (Gagnon (1991); Masson, Meredith, and Symansky (1990) ; McKibbin and Sachs (1991); Taylor (1993)) are closer to the original Mundell-Fieming model than the earlier vintages of multicountry models with backward-looking expectations and interest rates, rather than the money supply, as exogenous monetary policy variables. In the latter models, a fiscal expansion produces little in the way of crowding out, because with an accommodating monetary policy there is little or no increase in short-term rates, and in any case neither the exchange rate nor the long-term interest rate can jump on impact in anticipation of future effects. 17 Indeed, in some models with backward-looking expectations and a relative neglect of capital flows as determinants of exchange rate move­ ments, fiscal expansion normally produces exchange rate depreciation, not appreciation (Helliwell and Padmore (1985), p. 1121, cite simulations of the Japanese Economic Planning Agency model). As a result, fiscal policy multipliers under both fixed and flexible exchange rates tend to be positive and large in these models and their effects tend to die out slowly. In contrast, simulations of the forward-looking models with flexible exchange rates give results that are much closer to the original Mundell­ Fieming result. These models embody the assumption of high capital mobility, often assuming that uncovered interest parity holds, and ex­ change rate expectations correctly anticipate the effects in the model of present and future values of the exogenous variables. As a result of either a temporary or a permanent fiscal expansion, the exchange rate appreci­ ates on impact, which tends to limit (though not crowd out completely) the expansionary effect on output. The dynamics of the model then imply declining multipliers and a reversal of the appreciation (and, ultimately, a negative effect on output and a long-run depreciation from permanently higher government consumption spending if effects on interest rates and the capital stock are taken into account). A return to the basics of the Mundeli-Fleming model can also be

17 See simulation results reported in Bryant and others (1988).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF 721 STAFF PAPERS explained by a desire to understand the nature of the international linkages-something that had been lost in the large-scale international models, which typically linked together models with quite different intel­ lectual underpinnings (the best-known example of this is Project LINK). 18 This has been especially important in recent work on interna­ tional economic policy coordination, where the nature of the transmis­ sion mechanism is crucial to the gains from policy coordination. Whether monetary expansion is positively or negatively transmitted (under flex­ ible exchange rates) is the key question for evaluating whether the unco­ ordinated equilibrium is too contractionary or too expansionary. The Mundell-Fleming framework is general enough to allow both possibili­ ties, and to relate them to intuitively important structural parameters that can be estimated (see, for instance, Ghosh and Masson (1994)). Thus, empirical studies of the gains from policy coordination have relied on what is basically the Mundeii-Fleming framework, with a few of the extensions mentioned above. lt is likely that future policy questions will continue to use this framework as the benchmark for an intuitive under­ standing of the likely direction of effects, and that much empirical work in international economics will continue to estimate variants of the model.

Foreign Trade Relationships ID. Papers on a variety of empirical aspects of import and export relation­ ships have figured prominently in Staff Papers .19 This is certainly not surprising, since after all foreign trade is central to the business of the Fund. How the demand and supply of imports and exports are deter­ mined has a bearing on a number of important macroeconomic policy issues, including, among others, the effects of both expenditure-reducing (monetary and fiscal) and expenditure-switching (exchange rate and tariff)policies on a country's external balance: projections of world trade and payments; and the international transmission of changes in economic activity and prices. Thus. in this particular area IMF staff research has

18The chapter by Frankel the book on multicountry simulations edited by Bryant and others (1988) inattempts to understand the model simulations in at thetencies. context of the Mundeli-Fleming model and expresses puzzlement inconsis­ 19Between 1950 and 1990, there were about 100 empirical papers on foreign trade published in Staff Papers . This track record probably Papers far ahead of any other journal on this particular subject. For a detailedputs Swff survey of the empirical foreign trade literature, see Goldstein and Khan ( 1985).

©International Monetary Fund. Not for Redistribution MARJO I. BLEJER. MOHSIN S. KHAN. PAUL R. MASSON 722 and meshed closely with the institution's policy interests, and the outcome is evident in StaffPapers. The papers on foreign trade that have appeared in Staff Papers can be grouped into two broad categories. First, there are papers that specify and estimate import and export equations at the individual country level. And second, there are the papers that develop ·'world'' trade models that deal with groups of countries. This section will take up each category in turn.

Individual-Country Trade Models

Virtually all models of imports and exports are variants of what has been termed by Goldstein and Khan (1985) the ''imperfect substitutes·· model. The key underlying assumption of this general model, as the name suggests, is that neither imports nor exports are perfect substitutes for domestic goods.20 The main characteristics of the imperfect substitutes model are as follows. On the demand side, in accordance with conven­ tional demand theory, the consumer maximizes utility subject to a budget constraint. The resulting demand equations for imports and exports relate the quantities to the level of income, the own price, and the price of domestic substitutes. The specification of the supply function is equally straightforward-the producer is assumed to maximize profits subject to a cost constraint. The resulting function relates quantities to productive capacity, the own price. and the price of inputs into the production process. The bulk of the empirical trade literature has focused on demand functions, and that concentration is also apparent in the work done in the IMF. The empirical counterpart to the imperfect substitutes demand model, for example, makes the volume (or real value) of imports a function of current domestic real income and the relative price of imports to the domestic price level. 21 Symmetrically, the demand for exports relates the volume (or real value) of exports to foreign real income and the relative price of exports to the foreign price level. The differences that exist among the studies essentially arise from differences in definitions

211 If they were perfect substitutes then one would see either the domestic or the foreign good taking up the entire market and each country would be either an importer or an exporter of a traded good but not both. Nenher of these predic­ support. tions21 has empirical countri e�, For developing . foreig� exchange con�traints have also someti�es been introduced as an add1t1onal vanable. or m lace of the pnce and income variables. See, for example, Hemphilleven p relattvc ( 1974).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 723

of various variables involved, the samples for which the studies are done. and the types of estimation procedures utilized. Aside from the fact that understanding and explaining foreign trade flows is an essential part of the IMF's mandate, the initial work on the subject was largely motivated by a desire to counter the general belief in "elasticity pessimism'' prevailing in the postwar period. As Machlup (1950) pointed out, most studies in the prewar and early postwar periods seemed to show that price elasticities of imports and exports were quite low. This observation was later confirmed in the comprehensive survey in Staff Papers by Cheng (1959), which compiled estimates of elasticities from 42 books and articles published in the period 1937 to 1957. Clearly, the evidence was worrisome, since it implied that the Marshali-Lerner conditions were apparently not met in a number of countries, thereby casting doubt on the effectiveness of devaluation-a policy being fre­ quently recommended by the Fund-to correct external imbalances. The challenge for the IMF staff was to ascertain whether this was indeed the case or whether it was the result of faulty or inappropriate empirical analysis. Ta king up this challenge, Guy Orcutt. a staff member of the Fund at that time, produced a paper that has probably had the most signifi­ cant and long-lasting influence on empirical trade analysis (Orcutt (1950)). While it was not published in Staff Papers, it nevertheless has become closely associated with the IMF. 22 The debt to Orcutt is clearly acknowledged by Houthakker and Magee (1969) in their seminal paper on estimates of import and export equations. Furthermore, the standard textbook on quantitative international trade by Learnerand Stern (1970) uses the Orcutt paper as its organizing framework. Orcutt gave a list of reasons why existing estimates of price elasticities of the demand for imports and exports may be biased downward. These included non linear effects of relative price changes on quantities; simul­ taneous-equation bias; aggregation bias; and lack of accounting for lags.23 This list provides a convenient way to group and cover a representative

22The paper was discussed by the IMF's Executive Board. There are several stories, all unconfirmed, as to why this paper was not published in Staff Papers. One is that it was considered "too technical." Another story is that the senior staff of the Research Department did not like the paper's implication that it was really difficult to properly estimate trade elasticities. A third story is that Orcutt himself preferred to publish in a well-established journal rather than in one just starting out. While the real reason why it was not published in Staff Pa pers has been forgotten, the fact that it was not means that the journal received fewer citations in the subsequent empirical trade literature than it otherwise would have. 23 A fifth source of bias-errors in observation-is not taken up here.

©International Monetary Fund. Not for Redistribution 724 MARIO 1. BLEJER. MOHSIN S. KHAN. PAUL R. MASSON and

sample of papers on individual-country trade models that have since appeared in StaffPapers . Nonlinear Effects of Relative Price Changes Most econometric studies of the demand for imports, including the more recent ones, assume that the elasticities of import demand with respect to relative prices are constant for all values of the explanatory variable.24 It can, however, be argued that the price elasticity itself will vary directly with the size of the price change, since the price change must be large enough to overcome buyer inertia and the costs associated with switching suppliers. This proposition was tested first by Liu (1954) for the case of U .S. imports. In brief, Liu included a squared relative-price term in a standard import demand equation to capture the hypothesized non­ linearity in behavior, and found that the hypothesis had empirical sup­ port. This finding was also confirmed by a more elaborate set of tests devised by Goldstein and Khan (1976). However, despite the obvious importance of the issue of large versus small price changes for determin­ ing the impact of exchange rate changes-which usually tend to lead to larger than normal changes in relative prices-on trade flows, these are the only two studies in Staff Papers that look at this particular issue. Simultaneity As is well-known, price elasticities in trade relationships can be seri­ ously biased by simultaneity between quantities and prices. Thus, single­ equation estimates of the price elasticities of demand and supply can be weighted averages of the "true" demand and supply elasticities and consequently can be biased downward. The basic conditions under which one can estimate a demand equation that would be free of such bias are either that the price elasticity of supply is infinite, or that the demand function is stable while the supply function shifts. For the case of imports, simultaneity is not that serious a problem. It is quite plausible to assume that for a small country the supply of imports from the rest of the world is infinitely elastic, and most studies have made this assumption. For countries like the United States, however, this assumption has to be tested, as is done, for example by Liu (1954) and Abluwalia an.d Hermindez-Cata (1975).25 On the export side there is a greater presumption of simultaneous-equation bias, and techniques such as two-stage least squares or instrumental variables have been employed

2•This assumption is. of course, inherent in the popular use of a log-linear functional form for the import equation. 25 See also Liu (1955).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF STAFF PAPERS 725 by, among others, Rhomberg and Boissonneault (1964) and Junz and Rhomberg (1965). 26 Aggregation Typically, trade models focus on determining aggregate imports or exports. If the effect of the explanatory variables-relative prices and income-is exactly the same for aggregated and disaggregated groups, or if the relationship between the components and aggregate explanatory variables is a stable one, then one can be indifferent between the aggre­ gate and disaggregated equations. If these basic preconditions are not satisfied, however, the estimates obtained directly from the aggregate relationships are likely to be biased. Therefore, in aggregate trade equa­ tions, if goods with relatively low price elasticities display the largest variation in prices and exert a dominant effect on the estimated aggregate price elasticity, the estimate will be biased downward. To account for the possibility of aggregation bias, estimates need to be obtained at the disaggregated level. For example, Deppler and Ripley (1978) show that the import price elasticity of manufactures is about three times the size of the price elasticity of foods and beverages. This result is typical of that found in the literature-see Goldstein and Khan (1985). Lags The theoretical formulations of import and export models assume that importers and exporters are always on their long-run demand and supply schedules. In reality, of course, the presence of adjustment costs and incomplete information implies that the adjustment of imports and ex­ ports to changes in relative prices will not be instantaneous. This delayed response of trade can be due to recognition Jags, decision lags, delivery lags, replacement Jags, and production lags. Gauging the pattern and length of such time lags is important for calculating the short-term and long-term price elasticities and thus for determining the proper effects of exchange rate and tariff policies. Even though Orcutt (1950) discussed this issue in some detail, early papers in StaffPap ers, such as Liu (1954), Fleming and Tsiang (1956), and Romanis (1961), did not incorporate the notion of lagged adjustment of imports and exports. The first paper to do so, by Junz and Rhomberg (1965), produced extraordinarily long lags. 27 Numerous later studies pub­ lished in StaffPapers showed that the average Jag in adjustment is quite

26 The paper by Goldstein and Khan ( 1978) deals explicitly with this issue of simultaneity in the determination of exports. 27 It is possible that this result is simply a consequence of using pooled annual time-series cross-section data for the estimation.

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER. MOHSIN S. KHAN. PAUL R. MASSON 726 and short-about a year for both imports and exports for a variety of coun­ tries. However, as it is not zero, the distinction between short-run and long-run elasticities has to be made in assessing the effects of relative price changes on trade flows. It is apparent from the preceding discussion that advances made over the years by researchers both in the IMF and elsewhere to overcome the problems outlined by Orcutt (1950) have substantially changed the pro­ fession's view about the sizes of import and export elasticities. Indeed, one can say that the elasticity pessimism of the 1950s has now given way to elasticity optimism, with Marshall-Lerner conditions widely believed to be satisfied in most countries. Thus, it is no longer a matter of debate whether devaluation. other things equal, can be successful in improving the trade balance.

World Trade Models The IMF's mandate to exercise surveillance over the international monetary system and monitor developments at the global level has led since the late 1960s to substantial interest in research on world trade and payments. The main focus of the papers published in Staff Papers in this area has been on projecting world trade flows and studying the policy and economic interactions among countries. The first paper to appear on the subject in StaffPapers was by Taplin (1967), which surveyed existing models that analyzed the structure of world trade and those that traced the transmission of short-run fluctua­ tions between countries or groups of countries. The second issue was taken further by Rhomberg (1970), who considered a variety of ap­ proaches and techniques for constructing a "world" economic model by linking-directly and indirectly-existing national economic models in such a way as to achieve sufficient detail of the various relationships between domestic economic variables, economic policies, trade and pay­ ments flows, and coordinated international action. Both of these papers essentially described the existing state of the art, and proposed alterna­ tive strategies as to how one might go about constructing world trade models. The breakthrough in the design of world trade models was made by Armington (1969a and 1969b). who developed a theoretical approach that became the centerpiece of subsequent work on this subject.28 In brief, the Armington model took the notion of imperfect substitutes

2RSee. for example, Branson (1972), Hickman and Lau (1973), Samuelson ( 1973), and De ·plcr and Ripley ( 1978) . It also provided the basic structure for the Multilaterar Exchange Rate Model (MERM)-Artus and Rhomberg (1973).

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF 727 STAFF PAPERS

a step further by distinguishing commodities both by kind-as done in the imperfect substitutes model-and by place of production. Types of com­ modities ("goods") correspond to broad commodity classifications, such as nontradables, manufactures, food, raw materials, etc. Goods pro­ ,. duced by different countries are called "products. For example, in the Armington framework, French and German manufactures are the same good but are different products, and these products are assumed to be imperfect substitutes for one another. This distinction, along with certain other assumptions relating to the separability of import demand between goods and products. and that the elasticities of substitution between all pairs of products in the same goods family are identical and constant in any market, sharply reduced the amount of information needed to derive the direct and cross-price elasticities of demand for imports of any product into a country from another country.29 Thus, the model's main appeal lies in the fact that it provides an extremely economical and consistent method for estimating all the bilateral and multilateral direct and cross-price effects of a single or simultaneous set of traded goods price changes, such as that brought about by an exchange rate change. Even though the Armington model has had an enormous influence on the papers on world trade models, applying it to actual data is fraught with difficulties. One is choosing the right level of aggregation for the goods categories. If these are defined too narrowly, the separability assumption is likely to be violated; if they are very broadly defined, the assumptions about equal elasticities of substitution are likely to be vio­ lated. A second problem is that the estimates of direct and cross-price elasticities will only be as good as the estimates of the underlying elastic­ ities of substitution and own price elasticities of goods. While there are estimates of the former, much less is known about the latter. The model was subsequently empirically implemented by Armington (1970). Incorporating trade data for 1966-68, together with the requisite assumptions regarding elasticities, the model was used to calculate the effects on trade, as well as other domestic variables such as prices, incomes, and expenditures, of exchange rate changes. The results showed that the full effects of an exchange rate change-that is, taking full account of international feedbacks-were quite different from those obtained when considering a country in isolation. Further work on world trade models that appeared in Staff Papers was more in the nature of refinements and adjustments of the model maintained in the Research Department of the Fund.

29This particular feature made the model very useful to the development of open-economy Computable General Equilibrium (CGE) models. See. for exam­ ple, Dervis, de Melo, and Robinson ( 1982), and the survey by Robinson (1989}.

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER, MOHSIN S. KHAN. PAUL R. MASSON 728 and

Concluding Remarks IV. Our focus in this paper has been on early contributions to international economics by Staff Papers , and we have singled out three of them: the theory of the balance of payments determination, the Mundeii-Fieming model, and empirical foreign trade relationships. Other areas in which major contributions have appeared in StaffPapers from the very first, and continue to do so, are international liquidity, including such aspects of the question as the demand for reserves and the role of Eurocurrency markets; the demand for money; stabilization and adjustment policies in developing economies; and the macroeconomic dimensions of public finance and budgetary policies. However. it is worth noting that a good part of the significant work done at the IMF in these areas did not get published in Staff Papers. For instance, Triffin's discussion of the dollar shortage and dollar glut only appears in Staff Papers in the form of a rebuttal to criticisms by Oscar Altman (Altman (1961); Triffin (1961)). And a large part of the work done on international liquidity in the context of the Special Drawing Right appeared in a book (IMF (1970)), rather than the journal. More generally, Fund economists have frequently pub­ lished in outside journals, and the IMF has published a series of books on fiscal and monetary issues. In recent years StaffPapers has also been in the forefront of work on exchange rate modeling, developing-country debt issues, and policy cred­ ibility. The journal made its mark early and has maintained its high standards throughout its 45 years of existence. It is still therefore true that no economist who specializes in international monetary issues can dis­ pense with keeping reasonably up to date with what is being written in Staff Papers .

REFERENCES

Ahluwalia, lsher J., and Ernesto Hernandez-Cata, "An Econometric Model of U .S. Merchandise Imports Under Fixed and Fluctuating Exchange Rates, 1959-73," Staff Papers, International Monetary Fund, Vol. 22 (November 1975), pp. 791-824. Alexander, Sidney S., "Devaluation Ve rsus Import Restriction as an Instrument for Improving Fo reign Trade Balance," Staff Papers, International Mone­ tary Fund, Vol. 1 (April 1951), pp. 379-96.

--, ·'Effects of a Devaluation on a Trade Balance,'' Staff Papers, Interna­ tional Monetary Fund, Vol. 2 (April 1952), pp. 263-78. Altman, Oscar L., "Professor Triffin on International Liquidity and the Role of the Fund," Staff Papers , International Monetary Fund, Vol. 8 (May 1961 ), pp. 151-91.

©International Monetary Fund. Not for Redistribution CONTRIBUTIONS OF 729 EARLY STAFF PAPERS

Argy, Victor, "Monetary Variables and the Balance of Payments." Staff Papers . International Monetary Fund, Vol. 16 (July 1969), pp. 267-86. Armington, Paul S. (l969a), "A Theory of Demand for Products Distinguished by Place of Production," Staff Papers. International Monetary Fund, Vol . 16 (March 1969), pp. 159-76.

-- (1969b), "The Geographic Pattern of Trade and the Effects of Price Changes," Staff Papers , International Monetary Fund, Vol. 16 (July 1969), pp. 179-99.

--, "Adjustment of Trade Balances: Some Experiments with a Model of Trade Among Many Countries,·· Staff Papers . International Monetary Fund, Vol. 17 (November 1970), pp. 488-523. Artus, Jacques R., and Rudolf R. Rhomberg, "A Multilateral Exchange Rate Model," Staff Papers , International Monetary Fund, Vol. 20 (November 1973), pp. 591-611. Bernstein, E.M., "Strategic Factors in Balance of Payments Adjustment," Staff Papers , International Monetary Fund, Vol. 5 (August 1956), pp. 151-69. Bilson, John F.O., ''The Monetary Approach to the Exchange Rate: Some Empirical Evidence;· Staff Papers . International Monetary Fund, Vol. 25 (March 1978), pp. 48-75. Blanchard, Olivier Jean, ·'Debt, Deficits, and Finite Horizons," Journal of Political Econonzy , Vol. 93 (April 1985), pp. 223-47. Branson, William H., "Monetary Policy and the New View of International Cap Brookings Papers on Economic Activity: ital Movements," 2 (1970), pp. 235-62. --, "The Trade Effects of the 1971 Currency Realignments." Brookings Papers on Economic Activity: I (1972), pp. 15-69.

--, "The Dual R.oles of the Government Budget and the Balance of Pay­ ments in the Movement from Short-Run to Long-Run Equilibrium," Quar­ terly Journal of Economics, Vol. 90 (August 1976), pp. 345-67.

--, and Will em H. Buiter, "Monetary and Fiscal Policy with Flexible Ex­ change Rates," in Economic Interdependence and Flexible Exchange Rates, ed. by Jagdeep S. Bhandari and Bluford H. Putnam (Cambridge, Massachu­ setts: MIT Press, 1983), pp. 251-85. Bryant, Ralph C., and others, eds., Empirical Macroeconomics for Interdepen­ dent Economies (Washington: Brookings Institution, 1988). Burton, M. P., and E. Phimister, "Core Journals: A Reappraisal of the Diamond List," Economic Journal, Vol. 105 (March 1995), pp. 361-73. Cheng, Hang Sbeng, "Statistical Estimates of Elasticities and Propensities in International Tr ade: A Survey of Published Studies," Staff Papers , Interna­ tional Monetary Fund, Vol. 7 (April 1959), pp. 107-58. Depplet, Michael C., and Duncan M. Ripley, '·The World Trade Model: Mer­ chandise Trade;· Staff Papers , International Monetary Fund, Vol. 25 (March 1978), pp. 147-206. Dervis, Kemal, Jaime de Melo, and Sherman Robinson, General Equilibrium Models for Development Policy (Cambridge: Cambridge University Press, 1982).

©International Monetary Fund. Not for Redistribution MARIO I. BLEJER, MOHSIN S. KHAN, PAUL R. MASSON 730 and

Dornbusch, Rudiger (1976a), "Devaluation, Money, and Non-Traded Goods.'' in The Monetary Approach to the Balance of Payments , ed. by Jacob A. Frenkel and Harry G. Johnson (Toronto: University ofToronto Press, 1976), pp. 168-86.

-- (1976b). "Expectations and Exchange Rate Dynamics,'' Journal of Polit­ ical Economy. Vol . 84 (December 1976). pp. 1161-76.

---, Open Economy Macroeconomics (New York: Basic Books. 1980). Dornbusch, Rudiger. and Stanley Fischer, "Exchange Rates and the Current Account." American Economic Review, Vol. 70 (December 1980), pp. 960-71. Fleming, J. M arcus, "Domestic Financial Policies Under Fixed and Under Float­ ing Exchange Rates," Staff Papers , International Monetary Fund, Vol. 9 (November 1962), pp. 369-79.

--, and Lorette Boissonneault, "Money Supply and Imports," Staff Papers, International Monetary Fund, Vol . 8 (May 1961), pp. 227-40. Flerning, J. Marcus, and S.C. Tsiang, "Changes in Competitive Strength and Export Shares of Major Industrial Countries," Staff Papers . International Monetary Fund, Vo l. 5 (August 1956), pp. 218-48. Frenkel, Jacob A., and Harry G. Johnson, eds., The Monetary Approach to the Balance of Paymems (Toronto: University of Toronto Press, 1976). -- (eds.), The Economics of Exchange Rates (Reading, Massachusetts: Addison-Wesley, 1978). Frenkel, Jacob A., and Assaf Razin (1987a), Fiscal Policies and the World Economy: An Imertemporal Approach (Cambridge. Massachusetts: MIT Press, 1987).

-- (1987b), "The Mundeii-Fleming Model A Quarter Century Later: A Unified Exposition," Staff Papers, International Monetary Fund, Vol. 34 (December 1987), pp. 567-620. Gagnon, Joseph, "Forward-Looking Multicountry Model for Policy Analysis: MX3," Economic and Financial Computing, Vol . 1 (Winter 1991), pp. 311-61. Ghosh, Atish R., and Paul R. Masson, Economic Cooperation in an Uncertain World (Oxford: Basil Blackwell, 1994). Goldstein, Morris, and Mohsin S. Khan, "Large Versus Small Price Changes and the Demand for Imports," StaffPapers . International Monetary Fund, Vol. 23 (March 1976), pp. 200-25. --, "The Supply and Demand for Exports: A Simultaneous Approach.'' Review of Economics and Statistics, Vol. 60 (May 1978), pp. 275-86.

--, "Income and Price Effects in Foreign Trade," in Handbook of Interna­ tional Economics, Vol. 2, ed. by Ronald W. Jones and Peter B. Kenen (Amsterdam: North-Holland, 1985), pp. 1041-105. Guitian, Manuel, ··credit Ve rsus Money as an Instrument of Control," Staff Papers , International Monetary Fund, Vol. 20 (November 1973). pp. 785-800. Hahn, Frank H., "The Monetary Approach to the Balance of Payments,,. Journal of International Economics, Vo l. 7 (August 1977), pp. 231-49. Helliwell, John F., and Tim Pad more, "Empirical Studies of Macroeconomic

©International Monetary Fund. Not for Redistribution EARLY CONTRIBUTIONS OF 731 STAFF PA PERS

Interdependence." in Handbook of International Economics. Vol. 2. ed. by Ronald W. Jones and Peter B. Kenen (Amsterdam: North-Holland, 1985), pp. 1107-5 1. llemphill, William L., "The Effect of Foreign Exchange Receipts on Imports of Less Developed Countries," Staff Papers . International Monetary Fund, Vol. 21 (November 1974), pp. 637-77. Hickman Bert G., and Lawrence J. Lau, "Elasticities of Substitution and Export Demands in a World Trade Model," European Economic Review, Vol. 4 (December 1973), pp. 347-80. Hicks, Earl, Graeme S. Dorrance, and Gerald R. Aubanel. "Monetary Analy­ ses," Staff Papers . International Monetary Fund. Vol. 5 (February 1957), pp. 342-433. Hicks, Sir John Richard, Va lue and Capital (Oxford: Oxford University Press. 1939). lloltrop, M. W., "Method of Monetary Analysis Used by De Nederlandsche Bank," Staff Papers , International Monetary Fund, Vo l. 5 (February 1957), pp. 303-16. Houthakker, Hendrik S., and Stephen P. Magee, "Income and Price Elasticities in World Trade," Review of Economics and Statistics , Vol. 51 (May 1969), pp. 111-25. International Monetary Fund, "International Reserves-Needs and Avail­ ability," Paper presented at an IMF seminar. June l-3. 1970 (Washington: IMF, 1970).

--. The Monetary Approach to the Balance of Payments (Washington: IMF. 1977). Johnson , Harry G., "Towards a General Theory of the Balance of Payments.'' in International Tra de and Economic Growth , by Harry G. Johnson (London: George Alien & Unwin, 1958), pp. !53-68. --, "The Monetary Approach to Balance-of-Payments Theory," Journalof Financial and Quantitative Analysis , Vol. 7 (March 1972), pp. 1555-71. --, "The Monetary Approach to Balance-of-Payments Theory: A Dia­ grammatic Analysis,·· Manchester School of Economic and Social Studies. Vol. 43 (September 1975), pp. 220-74.

--, "The Monetary Approach to the Balance of Payments." Journal uf lmemational Economics , Vol. 7 (August 1977). pp. 251-68. Jones. Ronald W., and Peter B. Kenen, eds., Handbook of lmemational Eco­ nomics, Vol. 2 (Amsterdam: North-Holland, 1985). Junz, Helen B., and Rudolf R. Rhomberg, "Prices and Export Performance of Industrial Countries, 1953-63,'' StaffPap ers, international Monetary Fund, Vo l. 12 (July 1965), pp. 224-69. Kanesathasan, S., "Government Imports and Import Taxes in Monetary Analysis of Income and Imports," Staff Papers , International Monetary Fund, Vol. 8 (December 1961). pp. 412-26. Kenen, Peter B .. "Macroeconomic Theory and Policy: How the Closed Economy Was Opened," in Handbook of /memational EcoiiOnucs , Vol. 2. ed. by Ronald W. Jones and Peter B. Kenen (AmsterdaJn: North-Holland. 1985), pp. 625-77.

©International Monetary Fund. Not for Redistribution 7J2 R .tnd \1 \RIO I 111 1.11 \1011\1' .., "11 \� 1'\l l R \1 \'o..,(l\,

Knurl. Penttl J .K . "The 1�.\�.:hange R,lte the 13alPrinceton University. Department of Economics. International Finance Section. 1978). Laband. David N .. and Michacl J. Piette. "The Relative Impacts of Economics Journals: 1970-1990." Jo11mal of Economic Literawre . Vol. 32 (June 1994). pp. 640-66. Leamer. Edward E .. and Robert M. Stern. Q11amiwtive lmernational Economics (Boston: Allyn and Bacon, 1970). Liu, Ta-Chung, "The Elasticity of U.S. lmport Demand: A Theoretical and Empirical Reappraisal ,'' Swff Papers, International Monetary Fund. Vol. 3 (February 1954). pp. 416-41. , ''A Simple Fo recasting Model for the U.S. Economy." StaffPapers . --International Monetary Fund, Vol . 4 (August 1955). pp. 434-66. Machlup. Fritz. ·'Elasticity Pessimism in International Trade, Economia lmer­ nazionale, Vol. 3 (1950), pp. 118-41. .. Masson. Paul R .. Guy Meredith. and Steven Symansky, 1/: A Revised and Excended Model. TMF Occasional PaperMU No.LT/MOD 70 (Washington: Mark International Monetary Fund. 1990). McKibbin. Warwick, and Jeffrey Sachs, Global Linkages: Macroeconomic Inter­ dependence and Cooperation in the World Economy (Washington: Brookings Institution. 1991). Meade. James. The Balance of Payments , Vol. 1 of The Theory of lmernational Economic Policy . and Mathematical Supplement (Oxford: Oxford Univer­ sity Press, 1951). Mundett, Robert A., "The Monetary Dynamics of International Adjustment Under Fixed and Flexible Rates . Quarterly Journal of Economics. Vol. 74 (May 1960). pp. 227-57. .. -- (196la). ·'The International Disequilibrium System.'' Ky klos , Vol. 14 (Fasc. 2, 1961), pp. 153-70. (196lb). "Flexible Exchange Rates and Employment Policy." Canadian --- mal of Economics and Policical Srience. Vol. 27 (November 1961 ). pp. ]OH 509-17. . "The Appropriate Use of Monetary and Fiscal Policy for Internal and ---External Stability," Staff Papers . International Monetary Fund, Vol . 9 (March 1962). pp. 70-77. ---. ··capital Mobility and Stabilization Policy Under Fixed and Flexible Exchange Rates,'' Canadian Jo11mal of Economics and Policical Science. Vol . 29 (November 1963). pp. 475-85. ---. "A Reply: Capital Mobility an<.l Size." Canadian Journal of Economics and Political Science. Vo l. 30 (August 1964 . 421-31. ) pp.

©International Monetary Fund. Not for Redistribution I \RI) < !l'\ IIWH' II

. ln1crnmwnal f. ctiiiOIIIIL\ ( lacnullan. 1\.t:\\ 'I 011-.. :'-. llJM.:). .. 111 On:utt . liu\· . Mt:a�un.:mcnt of Pncc Lla�tiCitiC� I nlt:rnational lrad.: :· Ucncw of Lcononucs mu/ Suai.�tic.1. Vl . llJ50). 117-32. li 12 (:'-.-la� pp. Pareto. V . . Manuel d'Eco!IOIIIU' Poli1ique (Pari�: Gwrd et Bricre. 190':1). Polak. Jacque� J .. "Monetary Analysi� of Income Formation and Payments Swff Pap ers. Problems:· International Monetary Fund. Vol. o (November 1957). pp. 1-50. . ---, . The Internationalization of Economics: The Contribution of the Inter· national Monetary Fund." Paper presented at Duke University. Durham. North Carolina. April 7-9. 1995. . ---, and Lorette Boissonneault.. . Monetary Analy�i),of Income and Imports and lts Statistical Application . . StaffPa pers . InternatiOnal MonetaryFund. Vo l. 7 (April 1960), pp. 349-415. Polak. Jacques L and William H. White. ··The Effect of Income Expansion on the Quantity of Money." Staff Papers . International Monetar� Fund. Vol. 4 (August 1955), pp. 398-433. Prais, S. J.. ··some Mathematical Notes on the Quantity of Money in an Open Staff Pap ers . 1961). Economy." International Monetary Fund. Vol. S (May pp. 212-26. Rhombcrg, Rudolf R .. "Possible Approaches to a Model of World Trade and Payments."' Swff Papers . International Monetary Fund. Vol . 17 (March 1970), pp. 1-26. ---, and Lorette Boissonncault. "Effects of Income and Price Change� on the U.S. Balance of Payments," Staff Papers . Imernational Monetary Fund. Vo l. 11 (March 1964). pp. 59-112. Robinson, Sherman. "Multisectoral Models," in Handbook of DePelopmem Economics . Vol. 2, ed. by Hollis Chenery and T.N. Srinivasan (Amsterdam: North·Holland, 1989). pp. 885-947. Romanis, Anne, "Relative Growth of Exporb of Manufacture� of United States and Other Industrial Countries:· Staff Papers . International Monetary Fund, Vo l. 8 (May 1961). pp. 241-73. Samuelson, Lee, "A New Model of World Trade ." OECD Occasional Swdies (Paris: OECD. 1973). Samuelson. PauL Fo tmdmions of Economic Analysis (Cambridge . Massachu­ setts: Harvard University Press. 19-17). Taplin. Grant B.. "Models of World Trade." Smff Papers . International Mone­ tary Fund, Vol. 14 (November 1967). pp. 433-53. Taylor. John B., Macroeconomic Policy in a Wo rld Economy: From £conome1ric Design 10 Prac1ical Opermion (New York: :--lorton. 1993). .. Triffin. Robert. ··A Brief for the Defense, Swff Pap ers . International Monetary Fund, Vo l. 8 (May 1961), pp. 192-94. Tsiang, S.C.. "Balance of Payments and Domestic Flow of Income and Expen­ ditures," S!aff Papers. International Monetary Fund. Vol . 1 (September 1950), pp. 2.54-88. Walras. Leon. £/emenl:; of Pure Economics , Translated by William Jafft� (Home­ wood . Illinois: Richard D. Irwin. 1954).

©International Monetary Fund. Not for Redistribution Papers lMF Staff Vol. 42.International No. 4 {December Monetary 1995) Fund (I 1995

Fifty Years of Exchange Rate Research and Policy at the International Monetary Fund JACQUES J. POLAK*

Tlzis paper reviews the development over the past fiftyyears in the IMF of research on, and policy attitudes toward, exchange rare matters. Three suc­ cessive phases are noted. In the first ten to fifteen years, within the confines of the par value system, research was mainly focused on achieving a better understanding of the working of exchange rates. Fr om the early 1960s to the mid-1980s, the focus shifted toward discovering and, if possible, enforc­ ing "correct" exchange rates. More recently, che limitations of that ap­ proach have led to increased attention to the merits of alternativeexchange rare regimes, both among the industrial countries and for developing countries. (JEL F31, F33]

The International Monetary Fund will begin exchange transactions on March 1, 1947. The transactions of the Fund will be at the initial par values which have been determined in the manner laid down in the Fund Agree­ ment ....This is the first time that a large number of nations have submitted their exchange rates to consideration by an international organization and thus a new phase of international monetary cooperation has begun. The major significance of the present step is not in the particular rates of exchange which are announced, but in the fact that the participating nations have now fully established a regime wherein they are pledged to promote exchange stability, to make no changes in the par values currencies except in accordance with the Fund agreement, and to assist each other in attaining the general objectives of the Fund. (IMF Press Release, December 18, 1946) The [Interim] Committee considered that recent exchange rate movements for some major currencies had gone farther than warranted by fundamentals and agreed that orderly reversal of these movements is desirable. ln this context the Committee agreed that stronger efforts were needed to reduce internal and external imbalances. (Interim Committee Communique, April 26, 1995)

• Jacques J. Polak, who holds a doctorate in economics from the University of Amsterdam (1937). served on the staff of the IMF's Research Department from 1947 to 1980. He became its Director in 1958, and Economic Counsellor in 1966. From 1981 to 1986 he was a member of the Fund's Executive Board. He is now the President of the Per Jacobsson Foundation. The author is grateful for advice received from Bijan Aghevli, Jacques Artus, James Boughton. Owcn Evans, Peter Isard, G.G. Johnson, and Susan Schadler. 734

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 735

N MANY PLACES in the world, especially universities and research in­ Istitutes, the economics of exchange rates are the object of inten­ sive research; and in many other places-governments and central banks, policy councils and newspaper offices-exchange rate policies are matters of active and often heated discussion. But nowhere in the world is the interaction between the scientific and the policy aspects of exchange rates-between "how do they work?" and "what should be done about them?"-as close and intense as in the International Monetary Fund. Over its entire 50 years, from the proud announcement setting the initial par values in 1946 to the almost pained admission of a misalignment (since reversed) of the major currencies in the spring of 1995, the IMF has been at the very center of thought and action-research and policymaking­ about exchange rates. This paper focuses primarily on research and the interaction between research and policy. The broad developments of the international mon­ etary system, from a par value regime to floating, first of certain major currencies and more generally thereafter, are too well known to need description; and the details, insofar as they are still of general relevance, can be found in the hundreds of pages that successive IMF histories have devoted to exchange rate questions. 1 The linkage between the research activities on exchange rates in the IMF and the policy outcomes is in no sense simple or direct. If there was any straight Line in research, its moving force was the search for improved understanding on how exchange rates and exchange rate regimes worked. Policies were no doubt in part influenced by the outcome of this research, but to a large extent they followed from the situations in which individual countries, or groups of countries, found themselves. Often, moreover, the task of the economist in the lMF was to find the best specification for, and sometimes merely the best rationalization of, the policy that had become inevitable in the circumstances. From the earliest days-including its prenatal period at Bretton Woods-until the present, the IMF has been dominated by economists at all levels of its hierarchy: from ministers of finance to senior treasury and central bank officials, to executive directors, to the overwhelming majority of the professional staff. Two of its Managing Directors, Per Jacobsson (1958-63) and Johannes Witteveen (1973-77), came to the Fund with strong reputations as professional economists. In one sense, the high degree of communality in professional background facilitated

1 On the first 20 years of the Fund, see in particular Margaret de Yries's extensive treatment of exchange rate questions in the second (''analytical'') volume of the first Fund history (M. de Vries (1969)).

©International Monetary Fund. Not for Redistribution 736 JACQUES J. POLAK the blossoming of economics in the institution and the absorption of any new scientific findings into policy prescription. The success of the Bretton Woods conference is probably the best example of fruitful coopera­ tion among economists who shared certain fundamental ideals even though motivated by differing national objectives. Another example was the technical development of the SDR mechanism, as distinguished from the laborious political process of reaching agreement on whether, how, and with what safeguards SDRs should be created. In numerous in­ stances, moreover. research in the Fund benefited directly from the fact that the policy issues that rocked the institution rubbed economists' noses into real world problems and sometimes helped them find serendipitous answers to these problems. But in another sense the symbiosis of economic science and economic policy (or perhaps sometimes plain politics) under one roof carried the risk of contamination of the former by the latter. Up to a point, this could be prevented by the erection of firewalls, protecting the independent work of the staff from too much direct influence of the Executive Board; indeed, at times an important part of the job of the Director of Research seemed to consist in acting as such a firewall. But, to cite the most important instance in which closeness to the action interfered with rather than promoted research: in the crumbling of the par value system be­ tween 1971 and 1973, it would have been impossible for management and staff of the Fund to have labored simultaneously at shoring up the old system and at preparing the transition to a new world of floating rates. Thus, while some work on floating exchange rates had been done in the Research Department (see below), the staff-I believe inevitably, but also wholeheartedly-opted for the former task, which was also what the overwhelming majority of the membership wanted them to do. These introductory remarks should make it clear why there was not necessarily a close connection between the work of the staff on the operation of exchange rates in the international monetary system and the exchange rate policies pursued by individual member countries or indeed by the Fund. Most of the analysis in the remainder of this paper will be devoted to the work of the staff, mostly but not exclusively in the Re­ search Department, on the economics of exchange rates, but attention will also be paid to the policy questions that stimulated the research undertaken and the policy positions that sometimes (and sometimes not) flowed from the findings of that research. The wide variety of research activities concerned with exchange rates conducted by many economists over a period of 50 years cannot, of course, be neatly straightjacketed into a few distinct categories. But the following trichotomy attempts to assign a substantial proportion of

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 737

total research on this subject into three reasonably distinct subject cate­ gories, which moreover broadly succeeded one another in time as leading research themes over the fifty-year period. 1. In the first ten or fifteen years of the IMF, when the par vaJue system constituted the unquestioned regime for the great majority of the Fund's members, and when the influence of the young institution on the partic­ ular par value adopted by any member was modest at best, the staff's main research effort with respect to exchange rates was devoted to seeking improved understanding of how exchange rates worked. 2. In the next phase, which lasted from the early 1960s until the middle 1980s-well beyond the breakdown of the par value system-the research activities of the staff became increasingly focused on, and (it appeared at least initially) successful in, finding the ''correct" exchange rate and inducing member countries to adopt that rate. 3. In the most recent phase-say in the years since 1980-the staff appears to have become less confident of its15 ability to discern the correct exchange rate, especially for the major currencies, and increasingly con­ cerned with the question of the optimum exchange rate regime for these currencies. Widespread dissatisfaction with the volatility of the rates among these currencies and instances of extreme misalignment of some of them have kept the discussion about the best regime for these curren­ cies boiling, in many forums including the IMF. At the same time, with an ever-growing number of member countries linked to the Fund by a variety of financial arrangements, questions of exchange rate regimes also play an increasing role in the conditionality that the IMF attaches to the use of its resources. This third phase was aJso characterized by the emergence in the Fund of a much sharper distinction than had existed before between the industrial countries, which no longer used the Fund's resources, and the rest of the membership, among which use of Fund resources became more widespread. For these reasons the regime ques­ tions that are relevant to these two groups of members are treated separately in the last two sections of this paper.

I. Learning About Exchange Rates

By the second haJf of the 1940s, when the Fund began its activities, there was a broad common understanding of the economic principles on which the postwar world was to be based. Briefly, these principles envis­ aged the international application of the main tenets of Keynesian eco­ nomics in a spirit of cooperation among sovereign countries. This ap­ proach had indeed, as noted at the time, become part of ''the general

©International Monetary Fund. Not for Redistribution 738 JACQUES J. POLAK trend of social-economic thought in the United Nations today" (League of Nations (1943), p. 8).2 But the broad consensus on international financial policy still left many questions of application unsolved. More­ over, the IMF had, by its Articles of Agreement, been charged with a number of specific obligations concerning exchange rates whose execu­ tion was by no means self-evident. Before it could begin its financial operations, the Fund would have to agree to "initial par values" for the currencies of its members. Changes in these par values would only be permissible in the case of "fundamental disequilibrium." Members would have to avoid "competitive exchange depreciation''-but there was no comparable injunction against a country maintaining an overval­ ued currency, even though major instances of that form of misalignment in the interwar period counseled otherwise.3 It soon became clear to the newly assembled research staff that much economic work, both theoret­ ical and applied, would be required on such issues as: how could the theoretical concepts of tpe Articles be given operational content; by what mechanisms do exchange rates affect the balance of payments and em­ ployment; and what other variables play a role in determining balance of payments outcomes. The question of initial par values was solved rather easily by accepting for most member countries the fixed rates they were maintaining at the time, on the ground that, although these rates were probably not suitable as medium-term equilibrium rates, they did not seem to impede the modest flow of exports that these countries were able to spare from their domestic economies, while widespread scarcities would in any event make it necessary to restrain imports by controls rather than by the workings of the price mechanism. Having finished the task of agreeing initial par values in December 1946, the research staff, in addition to its work on individual countries, set out to produce a wide range of applied­ theoretical papers touching on many questions related to exchange rates. Once the Fund had started to publish Staff Papers, this journal provided

2 An important part in the educational process that led to this broad policy consensus had been played by the work of a group of economists at the League of Nations (De Marcht (1991)), whose publications included Nurkse's justly famous (1944) International Currency Experience (League of Nations (1944)) as well as two less well-known League of Nations committee reports (League of Nations (1943 and 1945)). 3 " •.•countries whose currencies have for whatever reason become grossly overvalued may have to be urged or persuaded to devalue their currencies in the general interest. The initiative in international consultation regarding possi­ ble exchange adjustments should therefore not be confined to individual coun­ tries desiring a change in their own currency values" (League of Nations (1944), p. 142).

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 739 a suitable outlet for the papers that appeared to be of laSting interest. A few for which this expectation has proved particularly accurate have been singled out in one of the other contributions to this issue (Ble jer, Khan, and Masson (1995)). Other papers provided building blocks for a growing body of Fund expertise on the working of exchange rates. In the first decade of Staff Papers these included, with a thumbnail indication of their contribution: Alexander (1951)-the welfare costs comparison of de­ valuation versus import restriction; Alexander (1952)-development of the absorption theory of balance of payments adjustment; Bernstein (1956)-a synthesis of ten years' work on exchange rates in the Research Department; B. de Vries (1950)-effects of devaluation on raw material prices; Fleming (1958)-effects on domestic price level; Gardner and Tsiang (1952)-competitive depreciation; Polak and Chang (1950)­ effects on export price level; Polak (1951)-payments effects of the 1949 devaluations; and Wyczalkowski (1950)-the economic meaning of the ruble exchange rate. A number of papers dealing with the few avail­ able cases of fluctuating exchange rates may also be noted: Rhom­ berg (1960)-Canada; Tsiang (1957)-Peru; and Tsiang (1959)-some European countries. Although the Articles leave the formal initiative for par value changes with the member country, the Fund in its early years was not squeamish about making suggestions for change, either confidentially or (leaving little doubt as to the countries considered ripe for such action) in its publications. As early as 1947, the Managing Director (Camille Gutt) advised the French government that the exchange rate for the French franc would need to be changed. The third Annual Report , that for 1948, spoke out clearly against any notion that the Fund "would regard any changes from the agreed par value ...as highly abnormal and to be sanctioned only reluctantly and in the most unusual circumstances" (IMF (1948), p. 21). It observed that there were indications "that in some countries [an obvious reference at the time to the United Kingdom and other countries in Western Europe] the exchange rate is becoming a restraining factor on exports and that it is adding to the difficulties of earning convertible currencies" (p. 23). And the Report reminded the reader of the Fund's "obligation to keep the exchange rate situation constantly under review" and of the propriety of expressing its views in its informaJ consultations with members (p. 24). When, a Little over a year after the publication of this Report , the devaluation of sterling (followed by that of many other currencies) was actually proposed, the Fund may have been only mildly surprised as to its precise timing, but it had had no inkling of the magnitude of the devaluation proposed (30.5 percent). Indeed, there is no indication that

©International Monetary Fund. Not for Redistribution 740 JACQUES J. POLAK the Fund had formed an opinion as to how large a devaluation would appear appropriate, which it could have communicated to the U.K. authorities. Thus, while the Fund was more than sympathetic to the need for par value adjustments, it had not yet moved into a clear quantitative mode on this subject. It recognized a wrong exchange rate but was not yet ready to put forward its own views on the correct rate.

ll. The Search for the Correct Exchange Rate

A number of currencies were devalued in the decade of the 1950s, but the first par value change of systemic importance after 1949 occurred in March 1961, when the deutsche mark was appreciated by 5 percent, followed two days later by the Netherlands guilder. On this occasion the staff had a view: that the devaluation was too small, and that 15 or at least 10 percent would have been more appropriate. 4 Over the following years, the staff sharpened its tools to enable it to get a clearer view of what exchange rates were or were not reasonably close to equilibrium levels, and how large a change might be appropriate for rates that it believed to be (to use a term that came into use only later) "misaligned." At the same time, official opinion in the major countries became increasingly reluctant to envisage any changes in the par values of their currencies. When in October 1963 the Ministers and Governors of the Group of Ten (G-10) asked their deputies to review the working of the international monetary system, they instructed them to accept two char­ acteristics as given: "fixed exchange rates and the established price of gold." The report of the deputies to their masters in June 1964 indicated how rigidly they interpreted their instruction with respect to exchange rates. In discussing the process of adjustment, they listed six instruments of economic policy that countries might need to use to counteract a tendency toward a sustained deficit or surplus in their balances of pay­ ments: fiscal policies, income policies, monetary policies, capital con­ trols, commercial policies, and selective policies directed to particular sectors of the economy-but, conspicuously, not exchange rate policies. Since it was not possible to wholly ignore the obvious, a cauda was

• Erin Jacobsson (1979), p. 336. The Managing Director (Per Jacobsson) had written to the President of the Bundesbank m September 1960 strongly coun­ seling against revaluation (Emminger (1986), p. 114). He was in any event opposed to a small revaluation, which "would only disturb exchange rela­ tionships'' and he showed himself furious with Otmar Emminger, who came to Washington to inform the IMF Executive Board of Germany's decision (Emminger (1986), p. 126).

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 741 attached to the list of six recognized policies: "Such instruments must be employed with proper regard for obligations in the field of international trade and for the IMF obligation to maintain stable exchange parities which are subject to change only in cases of fundamental disequilibrium." (G-10 (1964), pp. 4 and 5). The driving force behind this new attitude on exchange rates, which was sharply at variance with the ideas expressed in the early Annual Reports of the Fund, was Robert Roosa, the U.S. Under Secretary of the Treasury for Monetary Affairs and the first chairman of the G-10 Deputies. His fear was that any hint of flexibility of the parities for the main currencies might ultimately undermine the other postulate of the Deputies' study, the established dollar price of gold. Three years later the same fear inspired a last-minute U .S. Treasury attempt to save sterling from an inevitable devaluation by offering a large credit package (Solomon (1982), p. 94). In contrast to the doctrine espoused by the G-10 of near-absolute fixity of par values for the major currencies, the staff continued to hold the view that changes in par values should remain a realistic option, and it was successful in having a balanced description of countries' adjustment options included in the next Annual Report:

Deficit countries ...will need to choose whether to seek a solution to their problem in a degree of deflation or at least a slowing down of the rate of growth, in the adoption of measures to restrain payments for visible or invisible imports or capital transfers, in an adjustment in their exchange rate, or in some combination of these various broad policy approaches. Surplus countries may similarly find themselves forced to choose between inflation­ ary pressure, the adoption of steps to restrain the inflow and promote an outflow of capital, and revaluation of their currencies (lMF (1965), p. 11). In the course of its statistical work, including model building, in the area of international trade, the staff gradually acquired an ability to derive quantitative views on appropriate changes in par values. But before I turn to the staff's increasingly sophisticated models to determine the appropriateness of exchange rates I must refer briefly to what is sometimes seen as-and sometimes is -the most naive indicator of the extent to which the prevailing exchange rate deviates from the "correct" rate. That indicator is purchasing power parity, usually known by its initials (PPP). It has been studied in the Fund from the earliest days, with essential agreement that its relevance can be summed up in three propo­ sitions: (1) Relative (that is, foreign over domestic) prices have a major impact on international trade and hence on a country's balance of pay­ ments; (2) For any attempt to measure this impact, the choice of indices of (home and foreign) prices is crucial: prices for fully identical traded commodities do not diverge significantly, and prices for commodities that

©International Monetary Fund. Not for Redistribution 742 JACOUES J. POLAK are entirely untradable have no direct bearing on international trade; and (3) Factors other than relative prices also have important effects on international trade: equilibrium exchange rates are a function of more than PPP alone. The most interesting work on PPP done in the Fund relates to the second of these propositions. lt is published monthly, with some expla­ nation of the rationale for the choices made, in International Financial Statistics , in the form of "real effective exchange rate indices·· (see, for example, issue of July 1995, pp. 60-1). For 17 industrial countries, price index comparisons, adjusted for exchange rate changes, are made on the basis of 5 different price indices at home and in the other industrial countries: unit tabor costs, normalized unit labor costs, value-added deflators, wholesale prices, and export unit values. Weighing systems reflect both bilateral trade and trade in third markets. For the same countries and some 120 developing countries, similar real effective ex­ change rate indices are calculated on the basis of consumer price indices. 5 In the process of learning to appraise the economic validity of exchange rates, the staff's experience with the 1967 devaluation of sterling was crucial. This experience was applied a year later in the case of the French franc, and then, with the benefit of substantial model building in the meantime, in that of the U .S. dollar in 1971. ln the following years, the staff further refined its statistical techniques to make them serviceable to the IMF's new task of the surveillance of members' exchange rate policies.

Sterling, 1964-67

Ever since the Labor Government's accession in the United Kingdom in November 1964, the staff had believed that there was a convincing case for a devaluation of the pound; but it had also learned of the domestic and international political arguments against such a step. By the late summer of 1967, however, the staff was made aware that the U.K. authorities themselves were giving serious consideration to devaluation. This led to the assignment to a small group of economists in the Research Department, under the leadership of Marcus Fleming, to use the avail­ able lead time of perhaps two or three months to make the best possible estimate of the required magnitude of a devaluation of sterling. The assignment proved difficult, but feasible. With the help of estimated elasticities of demand and substitution, calculations were made of the

5 For a more extensive discussion of these alternatives, see Clark and others (1994), 6-10. pp.

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE 743 IMF impact of a given devaluation on rather detailed classes of British imports and exports; the counterpart deterioration of the current accounts of competing countries was also calculated and found generally to be small, on the order of percent of these countries' exports. By the time of the 1967 Annual Meetings1 (held in Rio de Janeiro in late September 1967) the staff team had reached the conclusion that a devaluation of sterling on the order of 15 percent would suffice to bring the U.K. balance of payments into equilibrium. It gradually transpired that the U.K. author­ ities had arrived at a similar conclusion: the devaluation that they ulti­ mately proposed to the Fund on November 17 was from $2.80/£ to $2.40, or by 14.3 percent. The staff's thorough preparatory work made it pos­ sible for the IMF to make two contributions to the smooth adjustment of the par values of sterling and a limited number of other currencies that it could not have made on the occasion of the September 1949 devalua­ tions. First, the staff was able to play an active role in the difficult diplomatic process of persuading the United Kingdom's main trading partners that a 15 percent devaluation was sufficient and deserved to be supported by a large stand-by arrangement from the IMF.6 And second, having found that the impact of a change of sterling of that magnitude on the balances of payments of the great majority of other countries would be bearable, the staff was able to mount a massive effort to contain the number of secondary devaluations and thus to avoid a chain reaction such as the one that had occurred in 1949.7

The French Franc, 1968-69

Probably the most bizarre par value change of the Bretton Woods period was that of the French franc in 1968-69: discussed semipublicly­ without a French proposal-at the emergency G-10 conference held in Bonn in November 1968; agreed as to its magnitude by the other coun­ tries present, which offered to support it with $2 billion in central bank

6 In what our other luncheon partners may well have regarded as the ultimate in Dutch boorishness, I spent the whole of a long social function in Rio explaining the findings of the Fund staff to my neighbor. Emile van Lennep, then the chairman of Working Party 3 of the OECD. Initially skeptical but persuaded by the time of the dessert, he then used his influence over the following weeks to bring a number of OECD members. in particular France, around to accepting the move by the United Kingdom without a change in their own par values. 7The second activity is described in the Fund history as being organized in a single business day (M. de Vries (1976), Vol. I, p. 434). It would not have been possible to do this without the staffs fengthy preparatory work.

©International Monetary Fund. Not for Redistribution 744 JACQUES J. POLAK credits; withdrawn from discussion the next day by orders of General de Gaulle (the support credits remaining in effect); and finally announced by the French Government (before being proposed to the IMF) in August of the next year. Of interest in the present context are not the theatrics of the Bonn conference (which also discussed a possible revaluation of the deutsche mark, to which the German Government was adamantly opposed), but the fact that the new par value "agreed'' (though not by France) in Bonn was proposed to the conference by Managing Director Schweitzer and, like the devaluation of sterling the year before, reflected staff calcula­ tions, in this instance hurriedly communicated between Washington and Bonn, that pointed to the need for an adjustment of between 10 and 15 percent. As in the British case, the precise percentage devaluation would have to result from an operation in terms of suitable round numbers. In the case of France, it seemed obvious to the staff that these would have to be round numbers in terms of gold: the franc was valued at precisely 180 milligrams of gold, making 160 milligrams of gold the natural out­ come of a devaluation of an adequate size. That was the content of Mr. Schweitzer's proposal at the conference: it implied a devaluation by 11.11 percent (M. de Yries (1976), p. 454).8

The U.S. Dollar, 1971

In the aftermath of the Bonn Conference, the Executive Directors initiated a two-stage process to re-examine the par value system. It began as a highly informal, but very active, set of meetings in the winter and spring of 1969, the results of which were laid down in the 1969 Annual Report, which in turn led to a spirited discussion at the 1969 Annual Meetings. This was followed by resumed study by the Executive Direc­ tors, an analytically excellent report (IMF (1970)), on which they found it very difficult to agree, and a new round of discussions by the Governors at the 1970 Annual Meetings. These proceedings over a period of nearly two years have been carefully documented in the Fund's history (M. de Yries (1976), Vol. I, pp. 500-16). What is evident from this record, as was

8The first press reports from Paris in August 1968 announced a devaluation of 12.5 percent. That figure perplexed us. Did the French want to devalue more (but only slightly more) than had been agreed in Bonn? Had we misread the magich of round numbers in terms of grams of gold? Our doubts were soon relieved: t e French did want to devalue the franc to 160 milligrams, but in calculating the percentage devaluation they had divided the shaving of 20 miUigrams by 160 tnstead of-as the Fund rules prescribed-by 180.

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 745 recognized by the staff at the time, is the extremely slim support then existing among member governments, comforted by the absence of ex­ change rate turbulence since September 1969, for anything but the smallest modifications (for example, a slight widening of the margins) of the par value system. There was lukewarm support for further study from the Governor for the United States (Treasury Secretary David Kennedy). but that was about it; almost all other Governors, including those for Germany and the United Kingdom, rejected any pursuit of the idea of greater exchange rate flexibility to the point of amending the Articles of Agreement. The world was thus ill prepared for the onset of the 1971 crisis of the dollar, which began when, in May of that year, massive inflows of capital from the United States swept a number of European currencies off their par values. Although the authorities in Germany, the Netherlands, and Belgium denied at the time any intent to revalue their currencies (M. de Vries (1976), p. 523), it rapidly became clear that the stage was set for a major currency realignment, although it remained equally unclear for many months thereafter how this realignment was to be brought about. Fo llowing up on its work on the U.K. devaluation, the staff had constructed its Multilateral Exchange Rate Model (MERM), which en­ abled it for the first time to make consistent estimates of the trade effects of simultaneous changes in the exchange rates for the currencies of all industrial countries.9 Starting with estimates of the required improve­ ment in the current account of the balance of payments of the United States-which the staff estimated at $8 billion-and compensating changes in the balances of payments of the other industrial countries, the staff used this model to derive a consistent set of new parities among these currencies. 10 The exchange rate changes adopted at the Smithsonian conference in December 1971 were not greatly different from those calculated by the Fund staff half a year earlier. Experience over the next year or so suggests, however, that the degree of adjustment agreed at the Smithso­ nian was insufficient. The reason for the failure to achieve an adequate depreciation of the dollar was, of course, in part political: the U .$.

�The model was not published in full detail until 1973 (Artus and Rhomberg (1973)), but an earlier version of it. based on work by Armington and Rhomberg. had become available in 1970. '0 As part of the negotiating process that ultimately led to the Smithsonian agreement, the IMF staff in October 1971 presented a paper estimating desirable cfianges in payments balances on current account for the OECD countries to a G-10 meeting. (There is a reference to this paper in IMF (1984a), 19.) The paper shows in detail the estimating methods used by the f:und staffp.

©International Monetary Fund. Not for Redistribution 746 JACQUES J. POLAK starting point had been that an improvement in the current account of some $13 billion would be needed, but creditor countries could not be induced to accept the correspondingly large adjustments in the opposite direction of their balances of payments (Volcker and Gyohten (1992), p. 81). But hindsight also suggests a technical explanation for the under­ estimation by the Fund staff of the required average depreciation of the dollar: the fault did not lie with the response coefficients of trade to the exchange rate changes but in the staff's underestimation in mid-1971, on the basis of information that did not go much beyond the end of 1970, of the magnitude of the underlying deficit that needed to be corrected.

A Correct Level for Floating Rates?

All in all, the technical work on exchange rates with sterling, the French franc, and the dollar gave the staff confidence that it had, in principle, developed a serviceable method of calculating equilibrium rates, or at least, until certain recognized weaknesses of the approach used would be overcome, "rather vaguely defined equilibrium zones" (Polak (1972] (1994), p. 395).11 In subsequent years, much staff effort was devoted to making technical improvements in the search for equilibrium rates; World Economic Outlook papers in particular became the vehicles to acquaint the Board with the staff's views on the rates for major currencies, often through the use of "scenarios" that showed how the payments positions of certain countries were expected to develop on certain assumptions with respect to exchange rates. The fact that the exchange rate regime gradually became more flexible over the 1970s did not diminish the IMF's interest in equilibrium rates; on the contrary, it led to an intensification of this type of work. 12 This was in line with the development of the IMF's attitude toward exchange rates as reflected in the second Amendment to its Articles of Agreement. Even after the notion of "stable but adjustable rates" that had dominated the early phase of the 1972-74 reform exercise had been

11 It should be noted here that at any moment of time-that is, at a given level of prices in all countries-the calculations could aim at finding an equilibri um nominal rate . However, with differential rates of inflation, it was the correspond­ ing equilibrium real exchange rate that was more likely to be the constant of the model-subject always to two important qualifications: the use of the correct price index and the realization that equilibrium real rates are changeable too (see the discussion on PPP above). 12 IMF (1984a) describes the active consultation with member countries under which this research was conducted in the course of the 1970s.

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 747 given up, the principle that members, whatever the'ir exchange rate regime, should see to it that they had the right exchange rate , remained the law of the Fund; indeed, it would be more accurate to say that the second Amendment made it the law of the Fund. The rule contained in the new Article IV, Section l(iii), that members should "avoid manipu­ lating exchange rates ...in order to prevent effective balance of pay­ ments adjustment or to gain an unfair competitive advantage over other members,'' conveyed, in much stronger terms than the par value regime had ever done, the notion that members should avoid serious deviations of their exchange rates from an equilibrium rate, either upward (over­ valuation) or downward (undervaluation). To ensure a system of correct exchange rates, the Fund, under Section 3(b) of the same Article, was to "exercise firm surveillance over the exchange rate policies of mem­ bers." The guidelines for surveiJlance that the Fund adopted in 1977 after long debates made it clear that both fixed and floating rates were to be appraised on their compatibility with the Articles. The guidelines spelled out the warning signs that might indicate that something was amiss with the rate (IMF (1992), pp. 8-14). Section 2(i)-(iv) of the principles of Fund surveillance, adopted as part of this decision, Lists indicators that might typically apply to countries with pegged rates. such as large-scale intervention, borrowing, or tightening of restrictions. This listing is fol­ lowed by an additional indicator-Section 2(v)-which makes it clear that surveillance applies also to floating rates. It reads: "behavior of the exchange rate that appears to be unrelated to underlying economic and financial conditions including factors affecting competitiveness and long­ term capital movements." The opening sentence of the staff's most com­ prehensive paper on the applied economics of exchange rate surveillance (IMF (1984a) takes its cue from the provision just quoted. By 1984, however, the basis for calculating equilibrium exchange rates had already considerably eroded.13 In a formal sense, such rates must correspond to a demand and supply situation reflecting both internal and external balance (Ciark and others (1994), p. 12). But the traditional criterion of internal balance, whether defined as "full employment" or "NAIRU" (the non-accelerating inflation rate of unemployment), had already ceased to be realistic by the mid-1970s; the staff had found it

13 As early as August 1982, the Managing Director, in his opening statement at an IMF-NBER conference on Exchange Rate Regimes and Policy Inter­ dependence, had commented on the difficulty the Fund had in making pronounce­ ments on the exchange rates of major currencies; the Fund, he noted, appeared to have clearer ideas on what an equilibrium exchan$e rate was forthe currency of a small country than on that of a large country (de Larosiere (1983), p. 1).

©International Monetary Fund. Not for Redistribution 748 JACQUES J. POLAK necessary to replace it by the expected degree of activity for the medium term (IMF (1984a), p. 20). But once it is acknowledged that some govern­ ments may lack the domestic-policy instruments to bring about the de­ sired cyclical position and the use of the exchange rate for this purpose is therefore no longer ruled out, the concept of an objectively deter­ minable equilibrium exchange rate is at risk of evaporating. The well­ orchestrated effort in July-August 1995 to bring about a depreciation of the yen deserves special scrutiny in this context. Although generally believed to be justified by evidence that the yen was overvalued and the dollar undervalued, it appears to have been inspired also by the desire, shared by Japan and the United States, to help jump start the Japanese economy; and press reports extolled the benefits to the U.S. economy that would result from larger exports to a more prosperous Japan. External balance had also been a relatively uncontroversial criterion for the industrial countries in the 1960s and the early 1970s. At that time, capital movements in many countries were still subject to control and, in any event, international capital markets were much less developed than they are now; thus, there was little likelihood of any country changing from one pattern of capital exports (or capital imports) to a radically different one within a few years' horizon. In the 1971 staff paper referred to above (footnote 10), the staff could feel confident estimating "target current account balances" for 1972 for each of the OECD countries. It did so by combining estimates of what, in the light of past experience, countries appeared likely to be able to finance in 1972 through normal net capital flows and government transfers, with figures derived from official balance of payments aims; these figures were then subjected to some scale adjustments. By the early 1980s, the opportunities provided by widening and deepening capital markets had undermined the plausi­ bility of any assumption that each country had an implicit or explicit normal current account position (or target) that was relatively constant over time. Yet, although stressing the difficulty of determining a normal level of current account balances, the staff (in 1984) still went on to estimate this level for each industrial country as equal to the average ratio of capital flows to GNP for the period 1975-82 (IMF (1984a), pp. 24- 25). For the United States (and for Japan) this yielded an estimated normal capital outflowof $5-10 billion, which contrasted with the paper's calculation of an underlying U.S. current account deficit of close to $90 billion (p. 21). With hindsight, the assumption of a small structural surplus for the United States was obviously a poor call. But more is at stake here than an incidental mistaken estimate. A consistent set of current account balance targets is, as Williamson ((1994), p. 91) points out, an essential

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 749 ingredient in the derivation of a set of equilibrium exchange rates. Williamson "solves" this problem by having all "significant countries" (which would include not only the G-7 countries but many others, devel­ oped and developing as well) "select a current account target, which would need to be consistent with the expected saving-investment balance and therefore with medium-run fiscal intentions ....The secretariat [of the IMF) would then have to appraise the realism and mutual consistency of the various targets. If an inconsistency emerged, a formula would have to be developed to reduce target surpluses to a uniform level consistent with the aggregate of target deficits." Many would probably share the view on this approach expressed by Cooper that there is "no basis for establishing such targets in today's world" (Cooper (1994), p. 112). That then leaves as the only possible approach the choice of a formula that would impose on all "significant countries" a current account target that the IMF found reasonable. One such formula, which had considerable support with reference to the industrial countries in the 1960s, was a current account surplus of 1 percent of GNP. That formula was still used recently by a team of Fund economists to test the robustness of equi­ librium exchange rate calculations (Bayoumi and others (1994)14); it is obviously not suitable for calculations that would include developing countries. An alternative formula recently developed in the IMF focuses on the stock of net foreign assets, assuming a target current account balance that would keep this stock at a constant percentage of GNP (Clark and others (1994), p. 15; Faruqee (1995)). But until the realism of formulas of this nature is established, 15 equilibrium exchange rates calculated on the basis of them must inevitably remain only illustrative, and the Fund's surveillance over the major currencies will continue to have to rely on its ability to recognize situations of clear misalignment without the benefit of agreed analytical procedures. At the same time, the positive calculations on balance of payments effects that can be made with MERM or successor models fully maintain their value in assessing

14 The exercise was presented as a test of the Smithsonian rates but was in fact less refined than the staff work in 1971. For example, the model does not include invisibles; the target used is that for a trade balance, not a current account balance, which is set equal to percent of GNP, whereas the staff in 1971 incorporated separate estimates1 for individual countries, including a current account deficit for Canada. 15 It is clear, for example, that, in this form, the net foreign asset approach could not explain the transition of the United States from a net foreign creditor to a net foreign debtor position and would also conflict with a commonly held view on the optimum patternover time of borrowing and debt repayment for countries developmg with the help of foreign capital.

©International Monetary Fund. Not for Redistribution 750 JACQUES J. POLAK the balance of payments effect of alternative policies with respect to exchange rates and other policy variables. In concluding this section on the search for the "right" exchange rate, a brief reference is needed to some of the contributions by the Fund staff to what has, on the whole, been the most disappointing branch of modern exchange rate economics: the attempts to explain, and then to predict, the movement of exchange rates under a regime of floating rates. This is not the place to go in depth into this subject; a rather recent issue of Staff Papers provides a comprehensive survey (MacDonald and Taylor (1992)). The main contributions made in the IMF to this vexed subject relate to the role played by nonmonetary factors in the determination of exchange rates. Dooley and lsard (1983) drew particular attention to the element of country risk (as distinguished from the narrower element of exchange risk), which made itself clearly felt in the large real deprecia­ tions of the countries suffering from the 1982 debt crisis; in a somewhat similar approach, Boughton (1988) stressed the importance of the long­ term real interest rate in the determination of the exchange rates among major currencies, given that financial markets for long maturities are much less internationally integrated than those for short maturities.

The Choice of an Exchange Rate Regime Ill. for the Major Currencies

The difficulties encountered in the search for an objective measure­ ment of equilibrium rates largely shifted the IMF's research away from this task-at least for the time being-and stimulated interest in the relative merits of alternative exchange rate regimes. After the demise of the par value system, that subject had been slow to get off the ground. The Committee of Twenty (1972-74) never discussed the future exchange rate regime in any depth (T. de Vries (1976), p. 587). A high-level conference held in the Fund on "the new international monetary system" focused mainly on the contribution that monetary policy coordination could make toward exchange-rate stability (Mundell and Polak (1977)). The first paper in the Fund to address the regime question head-on appeared only in 1979 (Artus and Young (1979)). Appraising the experi­ ence of the 1970s, that paper finds that flexible exchange rates do not (as some had claimed) guarantee policy independence, but also that their drawbacks have not proved as damaging as others had feared. The paper contains prophetic comments on the difficulties of pegged rate systems and concludes that if countries want to marry floating with reasonable exchange rate stability, the most valuable policy indicator would proba-

©International Monetary Fund. Not for Redistribution EXCHANGE RATERESEARCH AND POLICY AT THE IMF 751

bly be the exchange rate itself rather than some monetary aggregate or interest rate differential. Most of the writing in the Fund on regime questions of the last decade takes its cue from a 1984 paper by Morris Goldstein (IMF (1984b)), written as a response to "a resurgence of calls for a re-examination, or perhaps even a reform, of the international monetary system" that had become evident in the preceding few years (p. 1). After a lengthy evalu­ ation of the present exchange rate system, the paper draws six lessons, which can be highlighted as follows: (1) Do not overestimate the capacity of the exchange rate system per se to do good or evil. (2) Do not underestimate the importance of disciplined and coordinated macroeco­ nomic policies for the successful operation of floating rates. (3) The present wide diversity of exchange arrangements reflects the fact that the optimal degree of exchange rate flexibility differs across countries. (4) It is important to distinguish the effects of floating rates from the array of economic troubles that characterize the current period of floating rates. (5) The present exchange rate system has shown considerable strengths : it has promoted payments adjustment and maintained a market mecha­ nism of conflict resolution that has avoided the suspension of currency convertibility and large-scale restrictions of trade and capital movements. (6) But the present system has also manifested serious problems, the most critical of which have been instances of major misalignment, which in turn have led to inefficient resource allocation, adjustment costs, and a wide range of restrictive measures (IMF (1984b ), pp. 47-49, italics in original). In the subsequent decade, calls for the reform of the international monetary system did not abate and the Research Department responded with a series of further papers on the exchange rate system, includ­ ing Frenkel and Goldstein (1986), Crockett and Goldstein (1987), Frenkel, Goldstein, and Masson (1991), Goldstein and others (1992), and Mussa, Goldstein, and others (1994).16 Although the arguments pre­ sented in these papers covered broadly the same ground as the 1984 Goldstein paper, the conclusions became increasingly precise. Most specifically, while Frenkel and Goldstein in their 1986 "Guide to Target Zones" had gone out of their way not to express their view on the desirability of target zones (p. 633), the most recent (1994) paper explic­ itly rejects this proposal. And it bases this negative conclusion not, as is often done, on the defeatist argument that the major countries are simply unwilling to take account of international considerations in setting their

16 A book by Goldstein written while at the Institute for InternationalEconom­ ics Goldstein (1995)) can perhaps be considered the conclusion of this series. (

©International Monetary Fund. Not for Redistribution 752 JACQUES J. POLAK monetary policies. Rather, the paper invokes, at least implicitly, the proposition that the world is too large to be treated as an optimum currency area; hence "for [the three largest] countries, sacrificing domes­ tic stability to pursue a narrow concept of exchange rate stability would probably be harmful from a worldwide, as well as from a domestic, economic perspective" (p. 32). The growing consensus among the staff on the principal issues relating to the most desirable exchange rate regime for the major countries was not reflective of a simnar consensus in the Executive Board. The issuers of the three main reserve currencies-the United States, Germany, and Japan-have made it clear that they share the staff's negative attitude toward target zones, however attractive that concept may appear to a wider audienceY But some of the other industriaJ countries are less outspoken on the subject and among the rest of the Fund membership there remains an understandable hankering after the blessings of a regime that would again deserve the hallowed name of "Bretton Woods."

Exchange Rate Policies in the Context IV. of Fund Financial Arrangements

In the learning phase of the IMF's work on exchange rates, the ex­ perience gained from the observation of developing countries-where the Fund staff had at the time a strong comparative advantage compared to academic economists-was of particular importance. The devel­ opment in the IMF of the absorption theory of balance of payments adjustment, for example, was a direct consequence of the soul-searching that went on in the Research Department in response to the Mexican exchange crisis of 1948.18 In those early years, transactions were still rare in the IMF and its policies on conditionality had not yet been fully developed. But over time, the situations in which the IMF exercised the greatest influence on the exchange rate policies of member countries became those in which

17 The point was made decisively in the interventions of Lawrence H. Summers (United States) , Gert Haller (Germany ), and Kosuke Nakahari (Japan) in the July 1944 discussion of the Report of the Bretton Woods Commission. That report had endorsed "a more formal system of coordination, involving credible commitments ...," which was understood to refer to a target zone system-but only as a second step, after the major industrial countries had achieved more macroeconomic convergence (Bretton Woods Commission (1994)). 18See the first sentence of Polak [1948] (1991): ·'This pal?er has been written in connection with Mexico's proposal for exchange deprec1ation."

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 753

members sought the IMF's financial assistance. Thus the history of the policies pursued by the Fund in the context of its financial arrangements sheds much light on the development of thinking on exchange rates in the institution.19 Over the years, these policies were increasingly directed at ensuring that countries that borrowed from the Fund adopt economically justifiable exchange rates. In the 1950s and 1960s, most stand-by arrangements were granted to countries that faced only moderate balance-of-payments problems, not compounded by structural distortions (Johnson (1985), p. 6); in these circumstances a change in the exchange rate was frequently not needed as part of the country's adjustment program. But the greater distortions of the 1970s and the even more serious disequilibria of the 1980s did typically require action on the exchange rate as well as on other policy levers. Thus, while exchange rate action was included in only 32 percent of the Fund programs agreed in 1963 to 1972, the percentage rose to 59 percent in the period from 1973 to 1980, to 82 percent from 1981 to 1983, and nearly lOO percent in the later l980s.20 Mostprograms of the 1980s called, moreover, for adjustments during the program period to sustain the real depreciation initially achieved or to "prevent a loss of competi­ tiveness"; 18 of the 25 arrangements concluded in 1983 that began with exchange rate actions contained follow-up provisions of this general nature (Johnson (1985), Table 3).21 Even so, few Fund-supported adjustment programs in the early 1980s provided for the early elimination of restrictions, and exchange rate policy continued to rely on the government setting new rates rather than on permitting market forces to do so (Johnson (1985), pp. 3, 18, 19), leading to adjustment results that were often unsatisfactory. This experi-

19 I have covered a few of the observations in this section in an earlier paper (Polak (1991), pp. 36-8). 20 All figures refer to arrangements with countries that did not belong to a currency union. 21 In the same spirit, the staff started in 1982 to circulate to the Board monthly tabulations of real effective exchange rate indices for all members. Moreover. "information notices" were sent to the Executive Board whenever the index for any country had moved by more than 10 percent since the last Board discussion of the country concerned. These statistical activities seemed to imply a degree of confidence in PPP that was not shared by the staff as a whole: some theoretical papers (e.g., Frenkel and Goldstein (1986), Clark and others (1994)) go to considerable lengths to restate the well-known limitations attaching to PPP. Because of concerns that the approach was too mechanical and with the increas­ ing awareness of the perils of a real exchan�e rate rule (see below), the interest in these routine PPP calculations has declined and it was decided in 1995 to discontinue the information notices.

©International Monetary Fund. Not for Redistribution 754 JACQUES J. POLAK ence led to a further shift of Fund thinking on exchange rate policy for developing countries in the direction of "flexibility." Free floating in conjunction with a serious liberalization of trade and payments began to be seen as the only way to induce certain countries in extreme payments difficulties to accept an economically correct exchange rate. The adop­ tion of a free float gave the authorities in these countries an opportunity to liberalize exchange and trade restrictions and "to shed political respon­ sibility for the adjustment of the exchange rate'· (Quirk and others (1987), p. 4). In more recent years, however, the pendulum has to some extent swung back again to a more balanced appraisal of the merits and demerits of floating rates. Even if floating is efficient in inducing countries to aban­ don economically unrealistic exchange rates, does it provide the best exchange rate regime for those developing countries for which the avoid­ ance of inflation is also an important policy objective? In addressing this question, the paper that probably most clearly signals the turning point in Fund staff thinking starts out with the dual role performed by the exchange rate in small open economies: "Its movements can achieve and maintain international competitiveness and so ensure a viable balance of payments. At the same time, a stable exchange rate can anchor domestic prices" (Aghevli, Khan, and Montiel (1991), p. 1). In this formulation, the reference to the exchange rate as an "anchor" for price stability is relatively new in the Fund, but the concept is by no means new: in the 1970 Report of the Executive Directors cited above, the term "discipline" was used to convey the same notion. The Aghevli, Khan, and Montiel paper attributes its new insights to "recent theoretical developments" in economic theory that focus on a policy conflict of many developing countries, in particular those that have recourse to the Fund's resources, between the balance of payments constraint to which they are subject and the difficulties they encounter in controlling inflationary tendencies. The balance of payments requires a sufficiently competitive exchange rate, while inflation in excess of that abroad makes any fixed rate soon overvalued. In the 1980s, the standard Fund remedy for this problem had been a "real exchange rate rule,'' involving automatic depreciation at the country's differential inflation rate. But certain objections to this approach had become evident. If applied mechanically, the rule takes no account of possible changes in the equilibrium exchange rate since the real rate was first set. Such changes, or an initial choice of an excessively low real rate. might result in persis­ tent but futile attempts at undervaluation, with disastrous inflationary effects without any benefit to the balance of payments (Adams and Gros

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 755

(1986); Calvo and others (1994)). Even if application of the real exchange rate rule steered clear of these pitfalls, it would leave the country without any anti-inflationary anchor. Even a country that is too weak to adopt a fixed rate as an anchor, should-it is stressed-nevertheless meet at least some of the pressure brought about by domestic inflation through restric­ tive financial policies rather than offsetting all of it painlessly by auto­ matic adjustments in the nominal exchange rate. The paper concludes by suggesting that countries in this situation could to some degree combine the elements of discipline and competitiveness in their exchange rate policies by adopting a nominal crawling peg with a pre-set rate of crawl (p. 21). These new theoretical insights occurred at a time when certain devel­ opments in the real world made IMF opinion more receptive to them. One such development was the transformation of the European Mone­ tary System (EMS) in the second half of the 1980s. Before 1985, exchange rates in the EMS had been adjusted frequently, almost as soon as differ­ ential rates of inflation had brought about any significant change in real exchange rates. After 1985, by contrast, the EMS became a disciplinary regime in which the greater fixity of exchange rates became an anchor to force down other countries· inflation rates to the German norm. The success of this regime (until the emergence of some major question marks in 1992/93) strengthened the position, in both the Executive Board and the staff, of those who emphasized the positive policy aspects of a fixed exchange rate. Equally relevant were the experiments with heterodox stabilization plans-the first of which by Israel in 1985-in which the exchange rate, wage rates, and certain prices were frozen for a certain period to cut through what had begun to be called "inertial inflation." Although most of the heterodox plans of the next few years failed after a short period for lack of sufficiently strong fiscal support, the notion did gain currency that in certain extreme inflationary situations pure orthodox stabilization attempts, not backed by a suitable anchor, were bound to fail also. By the late 1980s, many countries with high inflation rates-such countries constituted as many as half the countries receiving Fund assistance at the time-saw a radical reduction in the rate of inflation as a key objective of their programs with the Fund. In almost all of the transition economies too, the mastering of inflation surfaced as a central problem. These considerations with respect to the optimal exchange rate regime for developing countries are reflected in the staffs most recent appraisal of the actual exchange rate policies of 36 countries that had stand-by or extended arrangements in the period 1988-91 (Schadler and others

©International Monetary Fund. Not for Redistribution 756 JACQUES J. POLAK

(1995)).22 Of these, 21 did not adopt an exchange rate anchor; instead, their exchange rate policies were designed with primary emphasis on maintaining, or sometimes improving, competitiveness. In that they suc­ ceeded, but in the course of their programs, these countries' inflation rates typically remained high or even increased, leading the staff to conclude that a nominal anchor is virtually a necessary (though not a sufficient) condition for a successful disinflation program. The 15 countries in the study that did adopt exchange rate anchors were, curiously, those with the highest and the lowest inflation rates of the sample, the former choosing the anchor in the hope of breaking inflationary expectations and introducing discipline into policies, the latter to maintain discipline, sometimes in the face of expected price increases. In general, the countries that adopted anchors23 improved fiscal discipline more than the other countries and were able to reduce inflation-but not enough to avoid a gradual loss of competitiveness, which forced many of them over time to move either to a more depreci­ ated rate or to a more flexible regime. The paper concludes that the results of the study do not provide a clear choice between the containment of inflation and the quest for competi­ tiveness as the primary objective for exchange rate policy. Rather, the lesson is drawn that a short-term trade-off between these two objectives has to be considered, taking into account such factors as the strength of the country's financial policies and the adequacy of its external reserves (Mecagni (1995), p. 70-1). Two other conclusions drawn by the staff are likely to remain of key interest in the future consideration of exchange rate policy issues. First, the countries that did not feel strong enough to trust their fate to an exchange rate anchor would not have been better off if they had tried to do so anyway;they might indeed have been worse off as they had to contend with repeated exchange crises. Second, coun­ tries adopting the exchange rate as an anchor need both the determina­ tion to stick to a sufficiently restrained fiscal policy and the flexibility to adopt a timely exit when, as is probable, the anchor becomes an excessive drag on the economy. Taking the past fifty years as a whole, the exchange rate problems that the Fund has had to deal with in connection with requests for the use of

22This extensive study is limited to countries with stand-by or extended ar­ rangements approved between mid-1988 and mid-1991; there is unfortunatel no dy comparable treatment of the exchan�e rate policies that the Fund supporte for the low-income countrieswith which 1t had arrangements under the structural and enhanced structural adjustment facilities (SAF and ESAF). 23These included not only the countries that adopted a fixed rate but also a few that followed a pre-announced crawl.

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 757

its resources, or more generally in the surveillance of the policies of the great majority of its members, have been problems of overvaJuation. In policy advice handed out by the Fund, "exchange rate flexibility" has almost always served as an only slightly veiled euphemism for devaluation or depreciation. In the most recent years, however, this is no longer necessarily so. As an increasing number of developing countries are faced with large inflows of capital, some real appreciation of their currencies can usually not be avoided. Although strong fiscal action (as practiced by Thailand in recent years, for example) is probably the best policy re­ sponse, the staff has also advocated nominal appreciation to relieve the pressure on the economy, noting that this has the side benefit of contain­ ing inflation (Schadler and others (1993), p. 30). A few developing coun­ tries (such as Chile and Colombia) have in recent years taken steps toward some revaluation of their currencies. These steps have, however, been quite modest, providing confirmation of an asymmetry in countries' exchange rate policy that can be observed over the entire period covered by this paper: countries often fail to take action needed to improve com­ petitiveness, but they hesitate to take any action that would reduce it.

REFERENCES

Adams, Charles, and Daniel Gros, "The Consequences of Real Exchange Rate Rules for lnflatioo: Some Illustrative Examples," Staff Papers , International Monetary Fund, Vol. 33 (September 1986), pp. 439-476. Aghevli, Bijan B., Mohsin S. Khan, and Peter J. Montiel, Exchange Rate Policy in Developing Countries: Some Analytical issues , IMF Occasional Paper No. 78 (Washington: International Monetary Fund, 1991). Alexander, Sidney S., "Devaluation Versus Import Restriction as an Instrument for Improving Foreign Trade Balance," Staff Papers, International Mone­ tary Fund, Vol. 1 (April 1951), pp. 379-96.

--, "Effects of a Devaluation on a Trade Balance," Staff Papers , Interna­ tional Monetary Fund, Vol. 2 (April 1952), pp. 263-78. Artus, Jacques R., and Rudolf R. Rhomberg, "A Multilateral Exchange Rate Model," Staff Papers , International Monetary Fu nd, Vo l. 20 (November (1973), pp. 591-61 1. Artus, Jacques R., and John H. Young, "Fixed and Flexible Exchange Rates: A Renewal of the Debate," Staff Papers , International Monetary Fund, Vol . 26 (December 1979), pp. 654-98. Bayoumi, Ta mim, and others, "The Robustness of Equilibrium Exchange Rate Calculations to AlternativeAssumptions and Methodologies,·· in Estimating Equilibrium Exchange Rates , ed. by John Williamson (Washington: Insti­ tute for International Economics, 1994), pp. 19-60.

©International Monetary Fund. Not for Redistribution 758 JACQUES J. POLAK

Bernstein, E.M., "Strategic Factors in Balance of Payments Adjustment,·• Staff Papers , International Monetary Fund, Vo l. 5 (August 1956). pp. 151-69. Blejer, Mario I., Mohsin S. Khan, and Paul R. Masson "Early Contributions of Staff Papers to International Economics," Staff Papers , InternationalMon­ etary Fund, Vol. 42 (December 1995), pp. 707-733. Boughton, James M., "Exchange Rates and the Term Structure of Interest Rates," Staff Papers , International Monetary Fund. Vol. 35 (March 1988), pp. 36-62. Bretton Woods Commission, Bretto11 Woods: Looking to the Future, Commission Report and Conference Proceedings , 2 vols. (Washington: Bretton Woods Committee, 1994). Calvo, Guillermo A., Carmen M. Reinhart, and Carlos A. Vegh, Targeting the Real Exchange Rate: Theory and Evidence, IMF Working Paper 94/22 (Washington: International Monetary Fund, Fe bruary 1994). Clark, Peter, and others, Excha11ge Rates and Eco11omic Fu11damentals : A Frame­ work for Analysis, IMF Occasional Paper No. 115 (Washington: Interna­ tional Monetary Fund, 1994). Cooper, Richard N., "Comment" in Managing the World Economy : Fifty Years After Bretton Woods , ed. by Peter B. Kenen (Washington: Institute for International Economics, 1994), pp. 112-16. Crockett, Andrew, and Morris Goldstein, Strengthening the International Mone­ tary System: Exchange Rates, Surveillance, and Objective IndicatOrs , IMF Occasional Paper No. 50 (Washington: International Monetary Fund, 1987). de Larosiere. Jacques, "Opening Statement," Conference on Exchange Rate Regimes and Policy Interdependence, Staff Papers , International Monetary Fund, Vol. 30 (March 1983), pp. 1-2. De Marchi, Neil, "League of Nations Economists and the Ideal of Peaceful Change in the Decade of the 'Thirties'.·· Economics and National Securiry: A History of Their Irueraction, ed. by Craufurd D. Goodwin (Durham. North Carolina: Duke University Press, 1991 ), pp. 143-78. de Vries, Barend A., "Immediate Effects of Devaluation on Prices of Raw Materials," Staff Papers, International Monetary Fund, Vol. 1 (September 1950), pp. 238-53. de Vries, Margaret Garritsen , "Exchange Rates" in The International Monetary Fund, 1945-1965, Vo l. IT, ed. by J. Keith Horsefield (Washington: IMF, 1969), pp. 39-173. 1966-1971 --. The InternationalMonetary Fund, (Washington: IMF. 1976.) de Vries, Tom, "Jamaica, or the Non-Reform of the International Monetary System," Foreign Affairs, Vol. 54 (April 1976), pp. 577-605. Dooley, Michael P., and Peter lsard, "The Portfolio-Balance Model ofExchange . Rates and Some Structural Estimates of the Risk Premium, . Staff Papers , International Monetary Fund. Vol. 30 (December 1983). pp. 683-702. Emminger, Otmar, D-Mark, Dollar, Wiilmmgskrisen: Erinnerungen eines ehemaligen Bundesbank Priisidenten (Stuttgart: Deutsche Verlags-Anstalt, 1986).

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 759

Faruqee, Hamid, "Long-Run Determinants of the Real Exchange Rate: A Stock­ Flow Perspective," Staff Papers, International Monetary Fund, Vol. 42 (March 1995), pp. 80-107. Fleroing, J. Marcus, ''Exchange Depreciation, Financial Policy, and the Domes­ tic Price Level," Staff Papers . International Monetary Fund, Vol. 6 (April 1958), pp. 289-322. Frenkel, Jacob A., and Morris Goldstein, "A Guide to Target Zones," Staff Papers , International Monetary Fund, Vol. 33 (December 1986), pp. 633-73. , and Paul R. Masson, Characteristics of a Successful Exchange Race ---Syscem , TMF Occasional Paper No. 82 (Washington, International Monetary Fund, 1991). Gardner, W.R., and S.C. Tsiang, "Competitive Depreciation," StaffPapers , International Monetary Fund. Vol. 2 (November 1952), pp. 399-406. Goldstein, Morris, The Exchange Rare System and the JMF: A Modest Agenda (Washington: Institute for International Economics, 1995). , and others, Policy Issues in the Evolving International Monetary Sys­ ---tem , TMF Occasional Paper No. 96 (Washington: International Monetary Fund, 1992). Group of Ten, Ministerial Statement of the Group of Te n, and Annex Prepared by Deputies Paris: G-10, 1964). ( International Monetary Fund, Annual Report of the Executive Directors for the Financial Year Ended April 30 (Washington: IMF, 1948). , Annual Report of the Executive Directors for the Financial Year Ended ---April 30 (Washington: International Monetary Fund, 1965). , The Role of Exchange Rates in the Adjustment of lmernationalPayments: ---A Report by the Executive Directors (Washington: International Monetary Fund, 1970). --- (1984a), Issues in the Assessment of the Exchange Rates of industrial Countries, IMF Occasional Paper No. 29 (Washington: International Mon­ etary Fund, 1984). (1984b), The Exchange Rate System: Lessons of the Past and Options for --the Future, IMF Occasional Paper No. 30 (Washington International Mon­ etary Fund, 1984). : ---,Selected Decisions and Selected Documents of 1he International Monetary Fund (Washington: International Monetary Fund, 17th issue, May 31, 1992). Jacobsson, Erin E . A Life for Sound Money: Per Jacobsson, His Biography (Oxford: Clarendon. Press, 1979). Johnson, G. G., and others, Formulmion of Exchange Race Policies in Adjustment Programs, IMF Occasional Paper No. 36 (Washington: International Mon­ etary Fund, 1985). Report of the Delegation on Economic Depression. Pare The League of Nations, 1: Transition from War to Peace Economy (Geneva: League of Nations, 1943). ---, lncernacional Currency Experience (Princeton, New Jersey: Princeton University Press, 1944).

©International Monetary Fund. Not for Redistribution 760 JACQUES J. POLAK

---, Report of the Delegation 011 Eco11omic Depression. Part 11: Economic Stability in the Post-War World: The Conditions of Prosperity After the Transition from War to Peace (Geneva, League of Nations, 1945). MacDonald, Ronald, and Mark P. Taylor, "Exchange Rate Economics: A Survey," StaffPapers , International Monetary Fund, Vol. 39 (March 1992), pp. 1-57. Mecagni, Mauro, "Experience with Nominal Anchors,'· in IMF Conditionality: Experience Under Stand-By and Extended Arrangemems, Part Back­ I/: ground Papers , ed. by Susan Schadler, IMF Occasional Paper No. 129 (Washington: International Monetary Fund, 1995), pp. 65-106. Mundell, Robert A., and Jacques J. Polak, The New International Monetary System (New Yo rk: Columbia University Press, 1977). Mussa, Michael, Morris Goldstein, and others, Improving the InternationalMon­ etary System: Constraiflts and Possibilities , IMF Occasional Paper No. 116 (Washington: International Monetary Fund, 1994). Polak, Jacques J. [1948], "Depreciation to Meet a Situation of Overinvestment," in Jmernational Financial Policy: Essays in Honor of Jacques J. Polak , ed. by Jacob A. Frenkel and Morris Goldstein (Washington: lnternational Monetary Fund and De Nederlandscbe Bank. 1991), pp. 46-57.

---, "Contribution of the September 1949 Devaluations to the Solution of Europe's Dollar Problem, •· Staff Papers, International Monetary Fund, Vol. 2 (September 1951), pp. 1-32.

-- [1972], "Reform of the International Monetary System: A Sketch of Its Scope and Content," in his Economic Theory and Financial Policy: The Selected Essays of Jacques J. Polak (Aldershot. England: Edward Elgar, 1994), Vol. Ll, pp. 392-404. -- (1991), The Changing Nature of IMF Conditionality , Princeton Essays in International Finance No. l84 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, 1991). ---, Economic Theory and Financial Policy: The Selected Essays of Jacques J. Polak (Aldershot, England: Edward Elgar, 1994). ---, and T.C. Chang, "Effect of Exchange Depreciation on a Country's Export Price Level," Staff Papers. International Monetary Fund, Vol. I (February 1950), pp. 49-70. Quirk, Peter J., and others, Floating Exchange Rates in Developing Countries: Experience with Auction and lnterbank Markets, IMF Occasional Paper No. 53 (Washington: International Monetary Fund. 1987). Rhomberg, Rudolf R., "Canada's Foreign Exchange Market: A Quarterly Model," StaffPapers , International Monetary Fund. Vol. 7 (April 1960), pp. 439-56. Schadler, Susan, and others. Recent Experiences with Surges in Capita/ Inflows , IMF Occasional Paper No. 108 (Washington: InternationalMonetary Fund, 1993).

---, /MF Conditiona!ity: Experience Under Stand-By and Extended Arrange­ Part Key Issues and Findings, ments, T: IMF Occasional Paper No. 128 (Washington: International Monetary Fund. 1995).

©International Monetary Fund. Not for Redistribution EXCHANGE RATE RESEARCH AND POLICY AT THE IMF 761

Solomon, Robert, The International Monetary System, 1945-1981 (New York: Harper and Row, 2nd ed., 1982). Tsiang, S.C., "An Experiment wi th a Flexible Exchange Rate: The Case of Peru, 1950-54," Staff Papers, International Monetary Fund, Vol . 5 (February 1957), pp. 449-76.

--, "Auctuating Exchange Rates in Countries with Relatively Stable , Economies: Some European Experiences After World War I,. Staff Papers , International Monetary Fund, Vol. 7 (October 1959), pp. 244-73. Volcker, Paul A., and Toyoo Gyohten, Changing Fo rtunes: The World's Money and the Threat to American Leadership (New York: Times Books, 1992). Williamson, John, and C. Randall Henning, "Managing the Monetary System," in Managing the World Economy: Fifty Years After Bretton Woo ds, ed. by Peter 8. Kenen (Washington: Institute for International Economics, 1994), pp. 83-111. Wyczalkowski, Marcin R., "The Soviet Price System and the Ruble Exchange Rate," Staff Papers , International Monetary Fund, Vo l. 1 (September 1950), pp. 203-23.

©International Monetary Fund. Not for Redistribution lMf Stoff Popus Vol. No. (December 1995) 42. 4 10 1995 International Monetary Fund

The Historical Development of the Principle of Surveillance

HAROLD JAMES*

The article discusses the evolution of surveillance from the rules-based Bretton Woods regime to the mulrilateral surveillance of the JM the G-5 and G-7 Finance Ministers, and the G-7 summit. The creation ofF, a mech­ anism for collecting and analyzing data and providing forecasts through the World Economic Outlook exercise allowed a form ulation of a policy response to the economic shocks of the 1970s and 1980s. lames argues that the supply of information came to play a central role in guiding choices on economic policy; and that publicly available information is critical if market panics and crises are to be avoided. [JEL E61, F42, 019)

ANY OF THE ideas and suggestions that have come to the fore in Mdiscussions about institutional development and renewal fifty years after the Bretton Woods conference involve concepts such as ·•multilat­ eral surveillance" and "early warning systems," which have long been a part of debates about the international financial system. Surveillance does not appear by this name in the Bretton Woods agreement, but it is implied by Article I, Section (i) of the IMF's Articles of Agreement, which established a "mechanism for consultation and collaboration on international monetary problems." In general, the IMF has under the Articles of Agreement a responsibility for the overall functioning of the international monetary system. The practice of surveillance has developed out of two parallel consid­ erations. First, the general overview of the international system called for the investigation of national economic policies. Initially, the major re-

* Harold James is Professor of History at Princeton University. He is the author of International Monetary Cooperation Since Bretton Woods (Oxford University Press and IMF), which was commissioned by the IMF to mark its fiftieth anniversary. Mr. James holds a Ph.D. from Cambridge University. 762

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 763

quirement for the establishment of an international monetary order was the adoption of current account convertibility. The Articles of Agree­ ment provided for consultations by the IMF with countries that had not lifted exchange restrictions five years after the IMF had begun operations (Article XIV, Section 4). Article VIII also committed the Fund to act as a central place for the exchange and transmission of information (Article VIII, Section 5(c)), while member countries had the corresponding obli­ gation to provide data on national income, prices, and a variety of international financial transactions. Later, as a response to the liberaliza­ tion of current account transactions and then the emergence of interna­ tional capital flows on a scale not anticipated at Bretton Woods, the assessment of the global economy and of interactions with individual national economies became an increasingly urgent task. The second principle for surveillance emerged out of the conditionality of the use of Fund resources (on which there is of course a substantial literature1), and a consequent need for supervision. From the beginning this was a central issue in debates about the role of the IMF. The details of conditionality were almost inevitably contentious, and at some times it became conventional for outsiders to speak of "harsh conditionality," but the general principle, except in the very early period, commanded a considerable consensus. As a result, one commentator suggested that most people would probably favor IMF-like conditionality in the case of any country but their own.2 There has always been a trade-off between surveillance for the benefit of the functioning of the international order, and the preservation of the prerogatives of national governments (sovereignty). This paper focuses largely on multilateral surveillance as a central element in the manage­ ment of the international monetary system, and as a means of reducing disturbance originating from policy shocks (since inappropriate policies in one country may have a major impact on the rest of the system). It does not deal in detail with the mechanisms and policy implications of bilateral surveillance.

Conditionality and Surveillance I. The substance of surveillance was already adumbrated in the early wartime discussions between John Maynard Keynes and Harry Dexter White about the shape of the postwar currency order. In particular, the

� See particularly Gold {1979), Dell {1981), Guitian (1981 ). - Cooper ( 1983), p. 571.

©International Monetary Fund. Not for Redistribution 764 HAROLD JAMES notion of discretionary management of the international monetary sys­ tem by a supervisory body was introduced by the United States in order to avoid the potentially very high liabilities implied in Keynes's original vision of an almost completely automatically operating clearing union. The United States intended to endow the stabilization fund suggested in its own proposal with powers sufficient to tackle those obstacles that had in the past stood in the way of world economic cooperation. The stabiJization fund would have to be given extensive powers to control matters that had previously been regarded as the purview of national sovereignties. "It is vital to the success of any international stabilization fund designed to play a really useful role that there be a suspension of certain economic elements of national sovereignty in favor of interna­ tional collaboration ....If no government is prepared to sacrifice for the sake of a larger though possibly a less obvious good what it regards as an advantage when that advantage is obtained at the expense of some friendly power, then the world will revert to the barbaric international economic relationships of the 'twenties and 'thirties. "3 Most of the proposed limitations on governments were set out as rules to be imposed by the new institution, including: (1) the abandonment, not later than one year after joining the stabiliza­ tion fund, of any restrictions and controls on foreign exchange trans­ actions that had not been approved by the fund; (2) a prohibition on altering exchange rates without the approval of the fund; and (3) a prohibition on accepting investments and deposits from another country except with that country's consent (the participants in the de­ bates on reform all had vivid memories of the large sums involved in capital flight in the 1930s). However, the proposal also included commitments and limitations that could not be formulated so precisely. For example. members would commit themselves to reduce trade barriers. And countries would not impose any monetary or banking or price measure that might "bring about sooner or later a serious disequilibrium in the balance of payments, if four-fifths of the member votes of the Fund submitted to the country in question their disapproval of the adoption of the measure. "4 This would mean an activist IMF, debating, consulting, warning, and advising-in short, behaving very differently from the inherently auto­ matic Credit Union of Keynes's plans. In order to work effectively, the proposed stabilization fund would need a trained and expert technical

3Horsefield (1969), Vo l. III, p. 62. "Horsefield (1969), Vol. I. p. 68.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 765 staff that would constantly supervise all international transactions in order to be able to advise the voters on the IMF's governing body. "The condition and movements of the international accounts of the member countries would be to the Fund's research staff, what the thermometer, stethoscope, x-ray, and microscope, etc., are to the diagnostician. "5 After Bretton Woods, the next step toward surveillance was the formu­ lation and systematization of a principle of conditional lending in the form of the stand-by arrangement. In 1948, an extended debate began in which it became increasingly evident that policy conditions to be placed on a country would depend on the amount it was drawing from the Fund. In part this move was a consequence of the IMF's April 1948 decision to deny access to IMF resources to countries participating in the European Recovery Program (ERP, or Marshal! Plan). In 1948, shortly after the ERP decision, an internal IMF memorandum suggested that the pay­ ments positions and prospects of members should be evaluated twice yearly and used to determine countries' eligibility to use Fund resources. The decisive breakthrough came when the IMF's second Managing Director, Ivar Rooth, proposed linking increasingly stringent conditions to increasingly higher levels of borrowing. A member's drawing of the first 25 percent of its quota would be automatic, but for larger sums, conditionality and surveillance would be applied. This method of opera­ tion was approved by the Board on February 13, 1952. In October 1952, in association with a provision for approving necessary drawings in ad­ vance, "stand-by" facilities-or, in other words, a way to approve in advance any necessary drawings-were approved and the basis had been laid for increased activity by the Fund as a financial institution. The stand-by arrangement, which made balance of payments support avail­ able on condition that certain policy measures be taken and subjected these measures to a performance test, became the most characteristic form of Fund financial activity. The first of these arrangements was a six-month $50 million program for Belgium, which was used by the Belgian Government to support the operation of the European payments system. After this breakthrough, the granting of waivers under Article V, Section 4 to members allowing them to draw more than 25 percent of quota in the course of one year became commonplace. At the same time a set of rules governing the availability of funds emerged. While the IMF was providing aid to Finland in 1952 it was established that in order to secure the principle of revolving access to Fund resources, drawings by a particular country should not be for more than three years. The technical basis of the Bretton Woods system was thus established. It depended on

5 Horsefield (1969), Vol. Ill, 57. p.

©International Monetary Fund. Not for Redistribution 766 HAROLD JAMES a limited automaticity in regular transactions, and increasing amounts of discretion by the IMF corresponding to the member country's increasing macroeconomic imbalance as revealed by the balance of payments. The newly created financial incentive structure could-and was-used to motivate countries to adopt further degrees of liberalization, and to create in practice the global economy that at Bretton Woods had been glimpsed only as a vision.

11. Surveillance and the International Financial System

Between 1946 and 1958 the major form of IMF surveillance was the consultation process with countries with nonconvertible currencies. held in accordance with Article XIV of the Articles of Agreement. Surveil­ lance took on a new meaning with the transition between 1958 and 1961 of major industrial countries to the current account convertibility pro­ vided for in Article VIII of the IMF Articles of Agreement. By the early 1960s, private capital markets (the "Euro-markets") began a dramatic expansion; the movement of capital imposed a new sort of constraint on national economic policymaking and potentially challenged the positions of the two major reserve centers-the United States and the United Kingdom. It only then became possible to speak of a genuine ··international monetary system": indeed, the phrase only entered the general vocabu­ lary in the context of discussions in the early 1960s about the creation of the General Arrangements to Borrow (GAB) by ten members of the IMF-Belgium, Canada, France. Germany, Italy, Japan, the Nether­ lands, Sweden, the United Kingdom, and the United States, as well as Switzerland (which joined the group, but not the IMF, slightly later).6 The 1960s saw a shift from a world of nation-states with an institution supervising rules of conduct, to a system in which nation-states, whose behavior was affected by those rules, interacted with markets, and in particuJar with the growing currency and capital markets. (Thus, the international financial system was conceptually entirely different from the international political system: in the latter, states alone interact with each other, while in the former there is a large body of other agents, private as well as officiaL continually judging and second-guessing the decisions made by the states.) The creation of additional resources through the GAB produced a new sort of conditionality, intended to calm continental European fears about

"Gold (1984). 22-4. pp.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 767 the possible use of the GAB by the major reserve centers of the time, the United States and the United Kingdom. After discussions with French Finance Minister Wilfrid Baumgartner. the IMF's Managing Director, Per Jacobsson, pledged that before making a call under the GAB, the IMF would obtain the consent of the participating countries. The agree­ ment was recorded in a letter of Baumgartner's to the other participants in the GAB.7 GAB participants would consult with each other and inform the IMF of the amount they were prepared to lend. The French provisions were accepted by the JMF's Executive Board on December 18, 1961. A request for support under the GAB required the consent of the GAB members as well as that of the Executive Board. This "double lock"' represented a major dent in the Fund's claim to universality and its capacity to judge by itself the conditions for assistance in dealing with balance of payments problems. The GAB and its members rapidly aroused the suspicion that a new ideology of cooperation between industrial countries had replaced the universalist aspirations of Bretton Woods. Cooperation between the G-10 members became institutionalized through study meetings of Deputies (senior or permanent civil servants at the heads of Finance Ministries or Central Banks). which from October 1963 were chaired by Robert Roosa, the U.S. Treasury Under Secretary for Monetary Affairs. and after 1965 by Otmar Emminger. Vice President of the German Bundesbank. The group committed itself to "undertake a thorough examination of the outlook for the functioning of the international mon­ etary system and of its probable future needs for liquidity.,. It defined its task as "multilateral surveillance" (the first use of the term in discussions about the international economy), which it interpreted as an appraisal of "the various means of financing surpluses and deficits" in order to develop "a common approach to international monetary matters. "8 Debate about reform of the system shifted for a time to this new forum. In April 1961 the OEEC (later the OECD) Economic Policy Committee created a study group called Working Party Three with the intention that it should analyze the effect on international payments of monetary, fiscal, and other policy measures, and would consult together on policy mea­ sures, both national and international, as they related to international payments equiljbrium. Working Party Three was to be composed of senior officials from Belgium, Canada, France, Germany, Italy, the Netherlands, the United Kingdom, the United States, Switzerland. and

7The text of the December 15, 1961. letter reproduced in Horsefield L969). Vol. ITI, pp . 252-54. is ( 8G-10 (1964), 4 and 9. pp.

©International Monetary Fund. Not for Redistribution 768 HAROLD JAMES a Scandinavian country (in practice Sweden), with the later addition of Japan, thus intentionally replicating the composition of the G-10. In this way, the industrial countries, with a heavy over-representation of Europeans, appeared to be on their way to establishing their own inter­ national financial system. Their actions challenged the vision many had of a more global, inclusive international economy. Working Party Three's essential function was multilateral surveillance of balance of payments issues. In order to accomplish this, it needed to analyze broader macroeconomic developments. Meetings characteristi­ cally began with a rour d'horizon of crucial economic indicators in a number of the large industrial countries. But in the increasingly serious currency crises of the later 1960s, the Working Party played an ever more peripheral role. This was not for lack of foresight. There was an urgent awareness of the gravity of the problem by the mid-1960s, but also, already then, an increasing sense of impotence and ineffectiveness. In August 1966 Working Party Three produced a report entitled "The Balance of Payments Adjustment Process." The report's recommenda­ tion was the creation of an "early warning system" to try to ensure that as soon as evidence of an actual or potential imbalance began to accumu­ late, it was the subject of a collective evaluation of the situation of the country or countries concerned, with a view to facilitating the adoption of appropriate policies. The key objective, spelled out in Paragraph 70, was to entice countries to begin the process of domestic adjustment earlier-before the outbreak of a major crisis. The report largely ignored exchange rate adjustment, giving the subject only two brief references. Within the Working Party, Emminger (who at the time was also the chairman of the G-10 Deputies) devoted himself to the development of a mechanism for such an early warning system. Instead of looking indis­ criminately at a large collection of data, policymakers should pay atten­ tion to a number of key indicators (money supply, bank lending, bank liquidity, short-term interest rates, and use of central bank credit). The function of the surveillance exercise was to focus attention on macroeco­ nomic data that should initiate a policy change or policy response. Em­ minger concluded that due emphasis should be placed on the relationship between the developments in the field of money. credit, and public finance on the one hand, and the balance of payments on the other. The problem that made surveillance less than effective lay largely in the political task of prompting countries to make earlier, and less dra­ matic, responses to potential balance of payments difficulties. In partic­ ular, the failure of the United Kingdom to address the overvaluation of sterling in the five-year period before the belated sterling devaluation in November 1967 demonstrated how surveillance was made ineffective by

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 769 poUticaJ constraints: in this case, the commitment of the U.K. Govern­ ment to full employment strategy, and its refusal to contemplate deval­ uation (which in fact had become known as "the unmentionable"). The practical unwillingness to consider parity changes as a mechanism of adjustment severely limited the room for maneuver under surveillance, especially since fiscal policy remained oriented toward the domestic political priority of maintaining full employment. In addition, there was also the almost insuperable intellectual chal­ lenge of convincing some monetary authorities that there existed such a link between money and the balance of payments. Effective multilateral surveillance becomes practically impossible in the absence of a basic consensus about the interactions involved in the global economy. The parallel institution that occupied itself with multilateral surveil­ lance fared Uttle better. The G-10 had an intrinsically strong position because of their economic, financial, and political strength (including voting power on the Executive Board). Nor did the purse power estab­ lished through its connection with the GAB harm that position. But a combination of political and intellectual resistance blocked the evolution of the G-10 as an effective provider of surveillance. In the mid-1960s, U .S. officials were referring to the G-10 as "more mouse than elephant. "'9 They resented the leverage that its composition inevitably gave to the Europeans, who regularly caucused before meetings, and as a conse­ quence the United States tried to make that body as much of a mouse as it could by restricting the policies that could be considered by the G-10. The influence of the United States, and the U.S. wish not to have the dollar discussed, led to a general refusal to consider parity alterations as a solution to balance of payments issues. In 1964, the G-10 Deputies' Statement included the following Ust of "appropriate instruments of economic policy"': budgetary and fiscal policies, incomes policies, mon­ etary poUcies, measures relating to international capital transactions, commercial policies, and selective measures on housing or hire pur­ chases. The document added that "such instruments must be employed with proper regard for obligations in the field of international trade and for the IMF obligation to maintain stable exchange parities which are subject to change only in cases of fundamental disequilibrium. "10 In other words, exchange rate changes were not to be considered a usable policy instrument. Even the IMF was willing to present the option of altering exchange rates only in the most cautious of terms. When discussing balance of

9 in Kunz (1994), p. 93. 10G-10Cited (1964), p. 5.

©International Monetary Fund. Not for Redistribution 770 HAROLD JAMES

payments adjustment, the IMF's 1964 Annual Report stated: "Adjust­ ments in exchange rates are of course not precluded by the par value system, and are indeed foreseen by the Articles in the event that a country has fallen into fundamental disequilibrium; but such situations should arise less frequently to the extent that the policies described above are followed. "11 Subsequent Annual Reports . however, were silent on the issue of parity changes. After the major sterling crisis of 1967. as a response to the French franc and deutsche mark turmoil of 1968, and in preparation for the meeting of the G-10 Finance Ministers in Bonn in November 1968. the IMF Research Department began to calculate appropriate parity changes to deal with the mounting current account imbalances. These calculations, based on the Multilateral Exchange Rate Model (MERM) might be said to mark the beginning of the IMF's engagement in a more concrete multilateral surveillance. In practice, however, it was politically impossi­ ble to raise these issues until after a general crisis had broken out in the Bretton Woods system following President Nixon's closing of the gold window on August 15, 1971. Even then, any leak about the MERM calculation was politically explosive. The figures provided by the Re­ search Department nevertheless eventually provided a major input in the resetting of exchange rates at the Smithsonian meeting in December 1971. But within two years, the new exchange rates had collapsed, and a general skepticism about pegged or fixed rates developed.

The Jamaica Agreement and the Second Amendment ill. The outcome of the currency chaos of the early 1 970s, often interpreted as reflecting too much discretion in the system, was a wish to return to increased monetary management. This was most easily effected in na­ tional political settings, where monetary control offered precisely a de­ politicized, rules-based approach that would take some of the excessive burden off the political system. Central banks tried after the mid-1970s to impose new rules in the form of monetary targets, but international flows of money and international policy considerations frequently threat­ ened the attainability of these targets. In the European Monetary Sys­ tem. a major group of states attempted to move back in a collective cross-national effort to a more rules-based system, and to limit the range within which currencies could float. In the global international arena, on the other hand, debates about

11 IMF, Annual Report (1964), p. 28.

©International Monetary Fund. Not for Redistribution PRINCIPLE SURVEILLANCE 77L OF monetary reform resulted in the IMF Interim Committee's Jamaica meet­ ing (January 1976) and the formulation of the Second Amendment of the Fund Articles of Agreement. These were less a basis for a move toward rules than a legal foundation for coordination based on discussion, per­ suasion, and the exchange of information. The logical outcome of this approach was less coordinated formulation of policy than the reaching of a common understanding about a stable policy framework. in which markets were exposed to as few surprises and shocks as possible. The new principle of surveillance was universal and all embracing. It included not only countries with IMF programs, but also those without. It recom­ mended structural adjustment in all countries with chronic balance of payments difficulties. whether developed or developing, 12 reflecting the theoretical revolution of the 1970s. In order to secure structural adjust­ ment, advice often had to be combined with the provision of resources during the adjustment period. The new system accepted the principle of non fixed exchange rates and encouraged an active role by the international monetary order's central institution-the International Monetary Fund-in prompting national authorities to accept realistic exchange rates. Such a regime would avoid both the disruptions caused by the "excessive" fluctuations of exchange rates and an undue rigidity of rates, which would hinder the international adjustment process. In this way, the reform might produce a return to "more orderly conditions. ''13 The agreement at Jamaica had been negotiated between the United States and France, in particular by senior finance ministry officials Edwin Yeo Ill, Under Secretary for Monetary Affairs of the U.S. Treasury, and Jacques de Larosiere, the French Director of the Treasury. France had begun these negotiations with a robust statement of its traditional com­ mitment to the principle of par values: Article IV, Section 1 of the new IMF Article of Agreement should refer to "the aim of establishing stable but adjustable exchange rates." The United States would then be drawn into the philosophy that lay behind the post-Bretton Woods pegged rate system that had been established as the "European snake." The IMF would "notify members, as soon as international economic conditions exist, that par values will be reestablished." But this would be backed up by a very intensive mechanism for regular consultation and surveillance: de Larosiere proposed a "collective agreement" between central banks to prevent erratic fluctuations, accompanied by a "secret agreement"

12 "Timely adjustment is as important for developing countries as for developed countries." IMF, Annual Report ( 1985), 13 IMF. Annual Report (1979) , p. 44.Annual Report ( 1978), p. 40; IMF, p. 43.

©International Monetary Fund. Not for Redistribution 772 HAROLD JAMES to determine "when conditions would be disorderly and erratic, and how central banks should intervene. "14 There should be a pattern of regular meetings between central bankers, and between finance ministry officials. The U .S. response to this initiative was somewhat skeptical. Yeo responded to de Larosiere by pointing out that currency intervention was not always a source of stability, and drafted a much more permissive version of Article IV. The experience of the European snake was cited by the United States as an example of the destabilizing effects of fixed exchange rates. In the U .S. draft, the purpose of the Agreement would be "to establish a framework for the promotion of exchange stability. the maintenance of orderly exchange arrangements, and the pursuit of exchange poucies that contribute to adjustment. Each member under­ takes to collaborate with the Fund and with other members toward these ends. ,.15 In lieu of a system of rules, the new Article IV set out a new philosophy of management of the international economy. Section 1 referred to the obligation of IMF members "to assure orderly exchange arrangements and to promote a stable system of international rates" (i.e., nor "a system of stable international rates"). Section 4 in practice indefinitely post­ poned (at least as long as the United States was opposed) the readoption of par values. "The Fund may determine, by an eighty-five percent majority of the totaJ voting power, that international economic condi­ tions permit the introduction of a widespread system of exchange ar­ rangements based on stable but adjustable par values. The Fund shall make the determination on the basis of the underlying stability of the world economy, and for this purpose shall take into account price move­ ments and rates of expansion in the economies of members." If there ever was to be a return to a fixed standard, it could not be gold or any national currency (the anchors of the so-called Bretton Woods system), but had to be the SDR or another common unit agreed by the IMF. In practice, it seemed unlikely that a new stable system would be put in place soon. As a result, the system would be a managed one. rather than one created simply by an automatic rule. While this did not include any specific provision for the surveillance of purely domestic economic policies (which naturally should be set in accordance with national political choices), the Articles of Agreement did provide a basis for IMF surveillance of domestic policies affecting

14 IMF. Annual Reporr (1979), p. 40; IMF, Amwal Report (1978). p. 43. 15 Yeo (1975).

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 773

growth and inflation.16 The interrelationship of domestic policies with surveillance activities was also acknowledged in an Executive Board decision (EBD 5392, April 29, 1977, as amended). In particular, it is noted in that decision that the Fund's appraisal of a member's exchange rate policies "shall be made within the framework of a comprehensive analysis of the generaJ economic situation and economic policy strategy of the member, and shall recognize that domestic as well as external policies can contribute to timely adjustment of the balance of payments.·· At the same time, the exchange rate and its problems inevitably pro­ vided an indicator reflecting the outcome of a large number of choices made about national economic policy: about financial and lhonetary management, the openness to capital movements, and the degree of openness to trade and the extent of flexibility in labor and product markets. In this way, exchange rates provided a guide to assessing all Links of the national to the international economy. The weakness of the Bret­ ton Woods system had been the continued absence of any adequate mechanism for encouraging members of the system to adjust exchange rates promptly to altered circumstances. The new Article IV stated: "The Fund shall exercise firm surveillance over the exchange rate policies of members, and shall adopt specific principles for the guidance of all members with respect to these policies.'' This was intended to avoid the problems that had dogged the par value regime. The 1978 Annual Report set out the new mission in the following way: "One of the task� of Fund surveillance will be to identify exchange rate policies leading to inappro­ priate rates at an early stage, and thus to reduce the economic costs and international Erictions associated with such rates. "17 Surveillance would in addition address (under the provisions of Article VIII, Section 7) the issue of international liquidity, and would have as an objective making the SDR the "principal reserve asset in the international monetary sys­ tem." The principle of surveillance was not new to the IMF. Surveillance had been practiced through the regular consultations held under Article XIV for countries with nonconvertible currencies, and later also through consultations with those member states that had accepted convertibility. The intention of the Jamaica Agreement was simply to strengthen an already existing practice. The specifics of surveillance were set out in an IMF decision of April 1977. The objective was that "a member shall avoid manipulating

16 IV, For instance, it stated in Article Section 1 that '"each melllber shall endeavor to direct itsis economic and financial policies(i) toward the objective of fostering orderly economic growth with reasonable price stability ...' 17 IMF, Annual Report {1978). p. 43.

©International Monetary Fund. Not for Redistribution 774 HAROLD JAMES change rates or the international monetary system in order to prevent effective balance of payments adjustment or to gain an unfair competitive advantage over other members." The Fund would initiate discussions with members outside the framework of the regular Article IV consulta­ tions in the case of deliberate manipulations carried out "for balance of payments purposes": if a country practiced protracted large-scale inter­ vention in foreign exchange markets; if it maintained an unsustainable level of borrowing or excessive lending; if it introduced or intensified or maintained restrictions on current account or capital account transac­ tions; if it pursued monetary or other domestic financial policies provid­ ing abnormal encouragement or discouragement of capital flows; or if the exchange rate seemed to be "unrelated to underlying economic and financial conditions." (The last condition was added, at the insistence of countries committed to the principle of fixed rates. so as to avoid the impression that countries that through a floating rate left their exchange rate policy to the market could automatically avoid the surveillance of the IMF.) This was a very long list of circumstances requiring IMF action and a restraint on members' action, but in practice these IMF-initiated discussions were very rarely applied.18 Although it turned out that the 1980s were replete with circumstances in which exchange rates moved wildly and without relation to "underlying economic conditions," and with balance of payments oriented actions, the IMF intervened only on two occasions to initiate special consulta­ tions. However, Article IV consultations often included formal or infor­ mal discussions of exchange rate arrangements. The two "special" con­ sultations took place in response to complaints about exchange rate policies by other IMF members. Scandinavian countries complained about the extent of the Swedish 15.9 percent devaluation of October 1982; and the United States criticized the large current account surplus and what it believed to be an undervaluation of the Korean won in 1987 (the time a formal IMF proceeding was launched with regard to a surplus country). In this sense the provisions of Article JV with regard to ex­ change rate policy represented more of a pious code filled with a hope of liberalization than a serious attempt to change countries' policies by specific intervention by the IMF. Initially, the principles for surveillance set out in the 1977 decision were regarded as temporary, and there was a provision for regular review. Although many observers both inside and outside the IMF were con­ cerned about the practical effectiveness of surveillance. and wanted a "strengthening," in fact there were few operational changes made during

18IMF (1993), pp. 8--14.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 775

the periodic reviews of the surveillance policy. The discussions provided for under surveillance, it appeared, could never be as "strong" as rules, but then the limitations of rules had been illustrated in 1971-3. There remained a constant danger that large and destabilizing capital flows and exchange rate movements might result from the adoption of mutually inconsistent national policies. Some surveillance was of course taking place in other forums. Central bankers continued to meet in Basle at the Bank for International Settle­ ments (BIS), with the G-10 central bank governors continuing to be a crucial group. However, since there was no longer any provision for the regular meetings of central bankers, finance ministry officials. and min­ isters-as originally suggested by de Larosiere in the discussions with Yeo-surveillance in practice depended in part on the regular IMF con­ sultations (known as Article IV consultations after the Second Amend­ ment, when the original purpose of the consultation process as a discus­ sion of the removal of exchange restrictions was obliterated). In practice, Article IV consultations for the major industrial countries involved an assessment of their macroeconomic stance, and frequently a detailed examination, with recommendations, on fiscal policy, as well as a discus­ sion of monetary policy. But exchange rate policy was in general such a highly politicized topic, and so sensitive to leakages, that it was more or less tacitly excluded at least from the formal part of the consultations procedure (even though it was the rationale for them). However, there were sometimes informal discussions. In 1983, it was agreed that for some members of the Fund these consultations need not be held annually, but should occur at least every two years, and in 1987 a mechanism for "bicyclic" consultations was introduced, though in 1993 the Fund moved back to the principle of the annual cycle. 19 Over time, a demand for increased publicity about the surveillance process developed. From 1990, brief summaries of the indi­ vidual consultations, reports, and related discussion by the Executive Board were included in the Annual Report. Surveillance was not a separate part of the Fund's activities, but rather constituted a prerequisite for effective support operations. The consulta­ tion exercise made the IMF aware of problems that might potentially require financial assistance. As a result, the fMF's financial programs, and the conditionality attached to them, could be regarded as nothing more than an extension of the surveillance procedure. The IMF rapidly reached the conclusion that "the effectiveness of the Fund's role depends

'" IMF, Annual Re ort 1983), IMF. Annual Report ( 1990), 13. See also p p. 64; p. Van Houtven (1994). (

©International Monetary Fund. Not for Redistribution 776 HAROLD JAMES not so much on formal or rigid procedures as on the quality and candor of the dialogue between the Fund and each of its member countries. In implementing surveillance, the Fund should therefore rely as much as possible on persuasion, rather than on prescription. "20 The surveillance process involved a continual exchange of information as a means of persuasion. Since it was a global exercise, it also served as a channel through which members could influence the policies and con­ duct of other states, either through the Fund's bilateral Article IV nego­ tiations, or through discussions and the provision of information about the state of the world economy. The Article IV consultations provided the basis for statements about world economic developments (and the role of national policies in these developments). This was the foundation for the forecasting and diagnostic exercise associated with the World Economic Outlook. This exercise was especially critical since discussion of economic problems focused increasingly on the global scenario; on the large current account imbalances that had emerged in the later 1960s, and their implications for exchange rates; on the impact of commodity price shifts (and particularly of the oil price increase); and, in the later 1970s, on the contribution made by growth in one or two of the large industrial economies toward world economic developments. The first World Economic Outlook, with projections of macroeco­ nomic indicators for the seven largest industrial economies, was prepared in 1969. Initially, many skeptics treated it as little more than "academic discussion,"21 but it became quite crucial in the wake of the major shocks of the 1970s, when it began to serve as a useful compass for an uncertain economic voyage. In 1973 the exercise was extended in order to make the forecasting of country performance compatible and consistent, so that each country's projections were adjusted in account of those for other countries. At the crucial meeting of the Committee of Twenty after the 1973 oil price increase, the IMF's Managing Director used World Eco­ nomic Outlook estimates as a basis for discussion of the appropriate response by the global economy to the price increase.22 The figures showed such large imbalances that it was clear that any attempt at sudden adjustment would produce a dramatic collapse of economic activity throughout the world. Full adjustment thus had to be delayed. Supplying information and ideas-as well as funds-now became the major means by which the IMF would seek to guide the evolution of

20IMF,Annual Report 57. 21 (1980), p. de Vries (1985), Vol. Il, p. 786. 22ThedetenoratJOn of the balance on current account for the industrial coun­ tries other than the United States was estimated at $45 billion, for the United States at $3-4 billion, and for the developing countries at $15 billion.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 777 economic policymaking. Managing Director H. Johannes Witteveen was particularly impressed with the possibilities for the new instrument. From 1974, the timing of the preparation and discussion of World Economic Outlook reports in the Executive Board was adjusted to fit the pattern of Interim Committee and Annual Meetings, and to provide a basis for discussion of trends in world economic development and the effects on the development of payments positions. At the end of 1973, in addition, in line with the increased emphasis placed on discussion of global eco­ nomic developments as a way of guiding decisions in individual countries, the Fund undertook so-called special consultations on international cur­ rency relations, initially with the large industrial countries. The World Economic Outlook mechanism allowed a continual interplay between the ideas of the IMF, which were partly formed through the staffs regular contact with member country officials, and, at the political level, the views of countries expressed through their constituency representatives on the IMF's Executive Board. In this way, the World Economic Outlook provided an institutional channel to mediate between "technocrats" on the one hand and "politicians" on the other. These different pressures became particularly evident later in the 1970s, as international demands that Germany and Japan should pursue more expansive policies in­ creased. The question of how Fund forecasts could be made more useful in policy discussions raised the difficult issue of whether governments should be publicly confronted with alternative forecasts for statistics as politically sensitive as growth rates and inflation levels. The general and advisory aspect of surveillance in fact soon took on a semipublic aspect as, first, the Managing Director's 1978 Note on the World Economic Outlook was given to the press, and then, from May 1980, the World Economic Outlooks themselves were published­ although initially much of the forecasting involved in the preparation of the "scenarios" was omitted. One of the purposes of making data and projections available is to inform and guide the shape of national debate about economic policymaking in the world's major economies (which of course influences the environment for other countries), and the openness of information and its availability for public debate are clearly a major facilitating factor. The discussion of surveillance had become more ur­ gent during 1978. At this time, the rapid depreciation of the dollar, especially against the yen, but also against European currencies, in­ creased fears about the consequences of currency floating for interna­ tional trade. The imperative for greater surveillance that also pushed the Europeans into the European Monetary System in 1978-79 made other countries demand a more effective application of surveillance. The centrality of the World Economic Outlook in the surveillance

©International Monetary Fund. Not for Redistribution 778 HAROLD JAMES

process was expressed not simply in a technique (collecting data) or a mechanism (discussion of the World Economic Outlook report), but also in an intellectual revolution that marked a major divide in the approach of many governments to the international economy. That revolution stands at the watershed between the inflationary 1960s and the orienta­ tion toward stability in the L980s. Debate within the JMF about the appropriate character of monetary policy was conducted largely in terms of a discussion of World Economic Outlook recommendations. The shift in thinking, replacing a short-term by a medium-term strategy, was quite abrupt. In immediate terms, it was a response to the mixture of stagnation and inflation in 1975. In 1975, the primary concern of the IMF Research Department in compiling the World Economic Outlook had been the possibility of a dramatic world wide collapse in demand as a response to the transfer of income to higher-saving and lower-spending oil producers; the recommendation thus reflected the belief in a general need for reflation. "Present policies might not provide sufficient stimulus to assure a strong and well sustained recovery. On this issue, the possibility should be recognized that consumer spending (counted on to lead the recovery in its earlier phase) and business fixed investment (expected to resume expansion after its customary cyclical Jag) may be more subdued than in earlier postwar recoveries ....Policies that were overly cautious could prolong the underutilization of resources, lead to widespread pressures for a rapid shift to expansionary measures, and forgo the beneficial effects of gains in productivity stemming from the resumption of solid economic growth and the absorption of slack." Only just over a year later, the tone of the World Economic Outlook had changed dramati­ cally. "The recent experience indicates that, unless the currently high level of price inflation is brought down and inflationary expectations are eradicated, the effects of policies aimed at stimulating growth and em­ ployment are likely to be short-lived ....It would seem to be in the interest of the entire international community and of the international adjustment process for the industrial countries to pursue policies directed toward the abatement of price inflation and of inflationary expectations." In practical terms, the new advice meant an increasing skepticism about the use of incomes policies as the primary weapon against inflation, and greater use of monetary and fiscal discipline. In intellectual terms, the reports reflected an analysis of shifts in current account position as a consequence of the behavior of investment and savings levels, a type of analysis that subsequently became the academic orthodoxy of the 1980s.23 These new points emerge very clearly in the later 1970s: there

z.•See Sachs (1981 ).

©International Monetary Fund. Not for Redistribution PRJNCIPLE OF SURVEILLANCE 779

was a "deep-seated" problem of lower productivity and potential GNP in industrial countries, attributable to "the relative weakness of capital formation in the 1970s compared with the previous decade.'' Short-term fiscal measures could not resolve this problem. As an IMF document at the time stated, "It is a lesson of experience that decisions to change the stance of fiscal and monetary policies should not be based on the movement of economic indicators over a very short period." The change of thinking within the IMF was an outcome of internal policy debates. In the mid-1970s, Managing Director Witteveen became increasingly frustrated with the shape of policy in major industrial coun­ tries, and believed that there had been too much financing and too little adjustment. He worked very closely in elaborating the new approach with the Research Department, which had been hostile to the new low condi­ tionaJity financing initiatives (the Oil Facilities) established by the IMF. Some of the institution's area departments also pressed in the same direction: in 1978, for instance, demanding a firmer line against inflation and an insistence that making forecasts about likely inflation levels should not run the risk of being interpreted as an endorsement of inflation. Finally, discussion of the World Economic Outlook in the Executive Board and the interventions of Executive Directors from countries with a greater sympathy for stability helped to strengthen the new stance. The full development of this approach still lay in the future. In the second half of the 1970s it appeared to many that the most important forum for the exercise of international surveillance lay not with the multilateral International Monetary Fund, but rather with the device of the "summit," excluding multilateral institutions, excluding developing countries, excluding oil producers, excluding the European Community, and including only "our small group"-as the summit had become known to its participants.

IV. G-5/G-7 Surveillance

For much of the first half of the 1980s, the absence of any consensus about economic policy made the practice of global surveillance problem­ atic. The U .$. administration's insistence that its budget deficit did not represent a major international problem, coupled with the belief of European politicians that it was a major source of disturbance, meant a complete impasse. The dissension divided G-7 summits, G-5 Finance Ministers meetings, and the IMF's Interim Committee. The apparently complete absence of agreement, however, prompted a new search for a mechanism that could make multilateral surveillance

©International Monetary Fund. Not for Redistribution 780 HAROLD lAMES more effective. At the beginning of 1982, some U.S. officials began to reflect on the cost of the high dollar for America's foreign political relations and to ask, "is there something missing from U .S. foreign economic policy?'' What was required internationally was a "conceptual glue," a "common analysis of international economic problems ...to move domestic economic policies in less diverging directions.·· The mis­ take of the past had lain in attempting to use international negotiations immediately to affect policy outcomes (as in 1978 at the highly controver­ sial Bono G-7 summit), rather than to work on discussions that would create an understanding and a consensus about the way policies affected performance.24 This was an approach that in the end proved highly fruitful, but it reached fruition only after the transition in the U.S. Treasury from Donald Regan to James Baker in 1985. At Rambouillet on April 24-25, 1982, in the course of a preparatory meeting before the Versailles G-7 economic summit, the French. Director of the Treasury, Michel Camdessus, and the U.S. Treasury Under Secre­ tary for Monetary Affairs, Beryl Sprinkel, spoke about the need for greater policy coordination.25 The French spokesman reasserted the tra­ ditional preference of his country for fixed exchange rates. In addition, he argued that greater official intervention in foreign exchange markets, and less of a reliance on pure "market forces," would produce greater stability.26 France was worried that she would suffer from the "erratic fluctuations" on currency markets, and the apparent resurgence of "benign neglect" on the U.S. side. Sharp currency movements would place a strain on the European Monetary System. ln addition, they would endanger the domestic reforms and the macroeconomic expansion pro­ gram of the new government, under the Presidency of the socialist party leader Franc;ois Mitterrand. A flight from a depreciating franc was the most likely threat to the new course. There were memories of the 1920s and 1930s when financial chaos had also undermined the ambitions of left-center governments. As a result, Mitterrand wanted to turn to the international system to safeguard his objectives. International coordina­ tion might lead to some measure of currency stabilization, and create a more favorable framework for domestic policymaking. The French Pres­ ident had pointed out that it was not just in France that high levels of unemployment would produce the risk of social explosions that presented "a greater danger than inflation.·· At the 1982 Versailles G-7 summit, Mitterrand went even further and explained that "My country is one of those that has had least success in the struggle against inflation, but we

24$ee 213-16. Nau (1990), pp. 25 Putnam and Bayne (1984), 160-61. 26Boughton (1994). pp.

©International Monetary Fund. Not for Redistribution PRINCIPLE Of SURVEILLANCE 781

have the fastest growth. There is a lot we can learn from each others' experience. "27 The United States, on the other hand, saw the most important task as securing a new disinflationary convergence of monetary policy and price stability. had little sympathy for intervention in exchange rates. or for Mitterrand'sJr reflationary ideas. The United States had an alternative approach to international monetary discussion, and to the building of a new consensus. The U .S. negotiators envisaged a sort of economic report card on country performance, including inflation and monetary perfor­ mance; they also hoped that it would be subtle enough not to judge fiscal stance solely by the size of the deficit-other measures such as the rate of increase of government spending or the full employment budget would be preferable. The outcome of this sometimes acerbic Franco-American discussion was productive in that it led to a more institutionalized mech­ anism for the discussion of economic policy in the major industrial countries, but the creation of a forum for debate corresponded much more closely to the U .S. view than to the more rules-oriented French vision. The proposal first set out by the economic officials at Rambouillet and then elaborated at the political level at Versailles (June 1982) in­ volved a regular meeting of G-5 Finance Ministers. in 4-6,practice twice yearly, with the Managing Director of the TMF attending in a personal capacity rather than as the representative of his institution. This repre­ sented a personalized (and perhaps slightly peculiar) way of implement­ ing the principles on surveillance set out in the Second Amendment of the Fund Articles of Agreement. The IMF's surveillance exercise could now be used directly as a basis for policy discussions by the major industrial countries. President Mittcrrand thought that this agreement would initiate a more general reform of the international monetary system. The summit communique stated: ·'In order to achieve this essen­ tial reduction of real interest rates. we will as a matter of urgency pursue prudent monetary policies and achieve greater control of budgetary deficits." A supplementary note stated: "We attach major importance to the role of the IMF as a monetar) authority and we will give it our full support in its efforts to foster stability ....We are ready to strengthen our cooperation with the IMF in its work on surveillance; and to develop this on a multilateral basis taking into account particularly the currencies constituting the SDR (i.e., the G-5] ....We rule out the use of exchange rates to gain unfair competitive advantages.··

27 Attali (1993), pp. 61. 239.

©International Monetary Fund. Not for Redistribution 782 HAROLD JAMES

The G-5 surveillance process in the new format was centered around a relatively short presentation (at first of only ten minutes) prepared by the Managing Director for G-5 Ministers meetings, and accompanied by a series of charts with explanations of major developments. An analysis of macroeconomic fundamentals would be the best way of producing a policy convergence around the objectives of low inflation and growth. The first G-5 meeting of the new type was held on September 3. 1982, in To ronto, shortly before the Fund/Bank Annual Meetings. The most dramatic action of the G-5, however, left both the rMF and the surveillance principle on the sidelines. The G-5 Plaza meeting in September 1985 was intended to correct the overvaluation of the U .S. dollar, with a mixture of high publicity and coordinated intervention (in fact, such intervention had begun already in January in a successful but more gradual attempt to lower the dollar). But interest rates and mone­ tary policy in general were not part of the Plaza discussion. Over the course of the next two years, they increasingly became the focus of controversy, as the United States saw greater expansion elsewhere as the best solution to its payments problem. In consequence, many felt that it would be desirable to include a more wide-ranging set of economic data in the discussions of the G-5. At the Tokyo economic summit (May 1986), the accompanying G-7 Finance Ministers meeting reached an agreement on the use of indicators in multilateral surveillance in order to encourage discussion of "appropri­ ate remedial measures.'' Initially, there were to have been ten: GNP growth, inflation, interest rates, unemployment, central and general government balances, current account balance, trade balance, monetary growth, reserves, and exchange rates. The number was later reduced to seven at the insistence of Nigel Lawson, the British Chancellor of the Exchequer, with the vaguer "monetary conditions" replacing the trio of interest rates, monetary growth, and reserves.28 In practice, most discus­ sions of monetary conditions concentrated on interest rates. As a result of the wish to strengthen cooperation by providing a more specific data set, the IMF became much more deeply involved than it had been at the time of the Plaza meeting (where public relations had mattered more than macroeconomic data). Underlying the indicators proposal was the belief that debates about international coordination had in the past focused too exclusively on exchange rate issues and that more fundamental corrections, affecting the current account positions of the major industrial countries, would be needed for the balanced development of the world economy. lndicators

��Lawson (1992), p. 547.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 783

were a major policy departure, a significant addition to the arsenal available for international economic cooperation, although logically they were simply an extension of the concept of surveillance expounded in the Second Amendment of the IMF Articles of Agreement. ln telling the story of any innovation, it is important to remember that the motives of the agents involved do not necessarily determine the nature or quality of the outcome. Potentially desirable outcomes can be launched for all manner of reasons, some good, some bad. The United States, which took the major part in launching this initia­ tive, wanted to use it primarily to alter policy in other countries. But it also represented a major change in the U.S. stance. The number of occasions in the postwar world on which the United States had offered to subject itself to external discipline had been few, and the willingness of the United States to entertain the concept had been completely absent in the early 1980s. In this light, other countries with an interest in a cooperative system should have seized the opportunity now provided by a move in part dictated by U.S. self-interest. In particular, the United States wished to address the large current account imbalances that seemed endemic by the mid-1980s. The United States was re-enacting a drama familiar on the international stage since John Maynard Keynes's wartime missions to Washington: the desire of a deficit country to find institutionalized ways of putting pressure on surplus countries. But the U .S. approach contained an obvious danger. Treasury Secretary Baker saw indicators primarily as a lever to obtain rapid political action: tax reductions in Germany or Japan, import liberalization in Japan, and (later) a lowering of German or Japanese interest rates. The process inspired excessive expectations about a quick response. The danger was that if something happened, the surplus countries would feel that they were being unfairly pressured, and if nothing happened the surveillance and coordination mechanism itself would get the blame. Baker had begun a discussion of the possible use of "objective indica­ tors" in April 1986, and had included originally also an eleventh indica­ tor, a commodity price index, which was later removed at the insistence of Secretary of State George Shultz. The notion was presented to the United Kingdom at the time of the spring Interim Committee meeting, and afterwards to France and Japan. The Interim Committee commu­ nique referred to discussion of external imbalances, policy interactions, and exchange rates. "An approach worth exploring further was the formulation of a set of objective indicators related to policy actions and economic performance, having regard to a medium-term framework." The Interim Committee communique also specified a mechanism for applying the indicators approach. The IMF's World Economic Outlook

©International Monetary Fund. Not for Redistribution 784 HAROLD JAMES would be expanded in function "to improve the scope for discussing external imbalances, exchange rate developments, and policy interactions among members. "29 At the Tokyo summit, the word "objective" was dropped from the summit declaration at the insistence of Germany and Japan. Both the Germans and the Japanese were generally skeptical about the approach, and hostile to its policy implications. Lawson recognized that the indica­ tors were "essentially a device to put pressure on the Japanese and Europeans to take 'expansionary measures' and in this way to take the heat off the falling dollar. "30 The Japanese Vice-Minister in the Finance Ministry, Tomomitsu Oba, commented that "the thrust of the whole argument revolved around this one consideration. We're not going to allow indicators to meddle into the domestic politics and sovereignty. ''31 On the other hand, the Japanese Government desperately wanted to keep the yen rate stable, since it believed that it would lose the general election due to be held in July if the yen rose above a level of 160 to the dollar. The other surplus country, Germany, was no more enthusiastic about indicators than was Japan. Of the Germans, Bundesbank President Karl Otto Pohl in particular was contemptuous of the multiplicity of indicators involved, and what he felt to be the intellectual fuzziness of the U.S. approach. He wrote soon after that "those who wish to replace persons and ad hoc decisions by regulatory mechanisms and indicators apparently have no perfectly clear idea about the nature of monetary policy decisions and the difficulties of reaching them ...there are, I think, very few situations in the area of internationalmonetary policy in which a depersonalized, predetermined decisionmaking process could have covered up, much less resolved, such differences of aim. "32 As a result, although the Germans felt that some indicators (fiscal deficits and monetary measures) mattered more than others, they did not attempt to set up a list of priorities. If there had been one, the United States would have preferred to put current account imbalances at the top. In June 1987, the G-7 summit produced the Venice Economic Decla­ ration, which proposed a specific method for the greater use of indicators in surveillance. First, each country would accept a commitment to de­ velop medium-term objectives and perspectives for its economy, and use these to accept mutually consistent objectives and projections. Second, the countries would use performance indicators to review and assess economic trends.

:!9IMF, Annual Report (1986), 111. 30Lawson (1992), p. (1989p) , . 547.13 5. 31 Funabashi p . (1987), 22. 32Pohl pp. 20.

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 785

Indicators attained a more prominent position in national economic policymaking because they were not simply intended as illustrations or demonstrations at high-level political meetings. Parallel to the sessions of G-5 ministers, the deputies would hold discussions on the indicators with the IMP's Economic Counsellor. These civil servants also discussed exchange rate and intervention issues, but without the advice or presence of IMF representatives (although this was clearly a policy area for which the IMF, under the Articles of Agreement, had a responsibility). The main task of the IMF in these meetings was to present indicators or projections from national sources, along with equivalent figures differ­ ently derived, through econometric models of the world economy. In the course of 1987-88, along with the indicators (GNP/GDP growth rates, inflation, fiscal balances, current account and trade positions, monetary conditions and exchange rates), the IMF also provided "composite indi­ cators" (real GDP/GNP, industrial production, consumer and commod­ ity prices, monetary growth, long- and short-term interest rates, employ­ ment growth rates, and unemployment levels).33 As the meetings developed over time, some shifts in emphasis occurred. The financial liberalization of the later 1980s made the relationship between national monetary behavior and output problematic, and the IMF began to pay attention to the "output gap," the difference between actual and poten­ tial production, as a guide to an appropriate monetary stance. In the early 1990s, in addition, it presented data on the constant employment fiscal balance (i.e., a fiscal stance adjusted for the state of the business cycle) in order to point out underlying fiscal shifts: deteriorations masked by the improvement in business conditions in the late 1980s, or improvements concealed by the world recession of the early 1990s. The IMF "indicators" of course did not always correspond with na­ tional estimates, nor with the actual performance of the indicators. (In terms of growth, they missed some of the strength of performance in the late 1980s. and failed to anticipate quite how severe would be the reces­ sion at the beginning of the 1990s).34 But they aimed at and succeeded in giving guidelines on what medium-term set of policies would achieve greater fiscal stabilization and a consistent monetary policy. In principle the indicators approach constituted a bold undertaking. It implied the acceptance by the G-7 of some sort of externally imposed constraints that would help to free the making of economic policy from the apparently irrational impulses of domestic politics. Some saw it as an "international Gramm-Rudman-Hollings. "35 Its most ambitious variant

33 See Dobson (1991 ). 34Barrionuevo (1992). H (1989). 149. Funabashi p.

©International Monetary Fund. Not for Redistribution 786 I-IAROLD JAMES envisaged the creation of "monitoring zones. ,y, If indicators moved outside these zones, they would trigger special meetings of the G-5/G-7, and perhaps even the adoption of remedial measures. The bolder concept was never realized, however, as the United States believed that figures for a medium-term projection of the current account position should be treated as "objectives" while the critics in Germany and Japan did not want to accept anything more binding than ''forecasts.·· Even a modest version of indicators made possible a more rigorous and structured discussion of international economic problems, and strength­ ened the role of the IMF in the surveillance process. One of the architects of the IMF implementation of the indicators approach, Andrew Crock­ ett, later wrote that "it did help to crystallize a consensus about the operation of economic policy ...The watchwords of this consensus are stability and sustainability. "'37 Taken in this sense the spirit of the exercise mattered more than the following of precise guidelines about the de­ velopment of indicators-about which countries almost inevitably hesi­ tated, as economic fundamentals changed. Even this spirit, however, required an acceptance by nation-states of some derogation of national sovereignty.

V. Global Surveillance

The next major impetus for an international reconsideration of the IMF's surveillance role came in the aftermath of two crises that demon­ strated the volatility of capital flows in a period of greatly increased mobility. First, the crises of the European Monetary System (EMS) exchange rate mechanism in 1992-93 had left the IMF apparently im­ potent. There had been some informal warnings about the strain on exchange rates following the shock of German monetary unification in 1990, but in the actual development and unfolding of the crisis the IMF played no role. Second, the Mexican devaluation crisis of December 1994 and the subsequent U.S. Treasury and IMF rescue package raised an analogous issue, but one that clearly involved the IMF directly. It was widely felt that the Mexican crisis did not reflect the longer-term policies of stabiliza­ tion. fiscal adjustment. liberalization, and deregulation pursued from the mid-1980s. On the other hand, it did result from a deterioration of economic policy in 1994, a sharp increase in government dollar-denom­ inated indebtedness, and the cessation and then reversal of capital flows.

36Baker (1988). 37 Crockett (1992), p. 281.

©International Monetary Fund. Not for Redistribution PRINCiPLEOF SURVEILLANCE 787

Such a sharp and sudden deterioration would be too abrupt to be handled in the conventional manner of annual IMF Article IV reports. These crises differed from the older problems of pegged rate systems, such as the Bretton Woods regime, in that the period prior to the out­ break of the crisis involved very large capital flows and judgments about the sustainability of those flows. The victims of the September 1992 EMS crisis-Italy, Spain, and the United Kingdom-had experienced major inflows attracted by higher interest rates and the promise of an exchange rate guarantee. If doubts arose about the commitment to the exchange rate mechanism. the flows would be very quickly reversed. Mexico also attracted major inflows (some $75 billion between 1991 and 1993, mostly of portfolio investment), which financed large current account deficits. The timing of the crisis depended on guesses as to when the inflows would end, and what the consequences would be for Mexico's crawling peg exchange rate regime. In other ways, the crises of 1992-95 were not so new. The discussion of both the EMS and Mexican crises raised the classical problems of external advice in the context of fixed or pegged exchange rate systems. Since changes and news of impending alterations offer the controllers of private funds the possibility of making dramatic gains, many market­ sensitive policy issues became very hard to discuss and analyze. The more substantial the capital flows, the greater the sensitivity and vulnerability. This was especially true of exchange rates, and of central bank interven­ tion on exchange markets-both in the fixed and in the flexible systems, but it is also true of interest rate policy. In addition, strong political pressures and incentives led to an attempt to orchestrate policies in a narrower setting, to create "our small group" (see end of Section IIJ). The most remarkable postwar example of the increased difficulty of practical surveillance is perhaps to be found in the contrast between the alacrity with which parity alterations were discussed by the IMF as a policy tool and a facilitator of adjustment in the late 1940s, and the reluctance of the G-10, the OECD, and the IMF to consider parity alterations for the major currencies in the 1960s. The extreme receptivity of markets to rumor, combined with the political delicacy inevitably associated with issues affecting national prestige, produced what amounted to a taboo on discussion. There was a fear, which grew with the threats to the credibility of the system, that any dent would make impossible the attaining of any new stability. In particular, the U.S. unwillingness after the late 1950s to consider a change in the dollar parity of gold, despite balance of payments deficits, produced near paralysis. It led to an institutional incapacity to deal with the needs of the global economic situation. In these circumstances, the only hope for change lay

©International Monetary Fund. Not for Redistribution 788 HAROLD JAMES

not in additional discussion but in deliberate and forceful behavior to break the impasse. In the circumstances of 1971, U .S. Treasury Secretary John Connally played that part with considerable verve. One of the major tasks of a reformed system, the IMF's Executive Board concluded, would be to establish "criteria and procedures for orderly change which will accord to the United States, as well as to other members, a due measure of initiative in the effective exercise of exchange rate flexibility. "38 This story, from the classical Bretton Woods era, of increasing inability to discuss market-sensitive problems despite the creation of new institutions for multilateral surveillance was repeated (with different institutional actors) in the tale of the EMS.39 In the early years after the creation of the system in 1979, there were few problems in both discussing and undertaking parity alterations. Later, from the mid-1980s, consideration of parities within the EMS became so politically sensitive, within the European Community, but also consequentially within the OECD and IMF contexts, that it was in practice ruled out. This dilemma provides an example of a more general problem, that an institution responsible to member governments finds the discussion of market-sensitive material very hard, as governments may resent the implications of second-guessing the market, and can only be persuaded by arguments about what the market is likely to do after the market has actually done it. For instance, in a different context, it is possible to imagine the outrage if any international institution had given a clear and statistic-laden warning about the extent of bank exposure to middle­ income debtors in the summer of 1982 and thereby touched off a panic flight of funds: it needed to wait for the crisis to be triggered by market sentiment. The same considerations applied in the development of the 1994 Mexican crisis. Similar problems have appeared elsewhere, outside the context of multilateral surveillance. In general. it might be concluded that the major problems for modern policymaking involve information and its availabil­ ity. For instance, in dealing with centrally planned economies in the 1980s, the worst cases of mismanagement, and the failure of surveillance, occurred in the cases (Romania, Yugoslavia) where deliberately decep­ tive information was submitted to the IMF, and the IMF Jacked both the technical capacity to correct it and the political ability to protest against the insufficiency of information. In the case of Romania, for some time there was no country page in Inremational Financial Statistics.

38 August 18, 1972, ·'Reform of the International Monetary System: A Report by the Executive Directors to the Board of Governors,,. reproduced in de Vries (1985), Vol. III, . p 27. 39 Ke nen 1995 ) ( .

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 789

There is a widespread recognition of the centrality of information to economic management. The April 1995 Interim Committee recom­ mended that "timely publication by members of comprehensive data would give greater transparency to their economic policies," and ''re­ quested the Executive Directors to work toward the establishment of standards to guide members in the provision of data to the public. "40 At Halifax in June 1995, the G-7 summit endorsed this view in its conclusions on reform of international institutions. Paragraph 16 of the summit communique called for "fuller disclosure of this information to market participants," and urged the IMF to: -establishbenchmarks for the timely publication of key economic and financial date; -establish a procedure for the regular public identification of coun­ tries which comply with these benchmarks; and -insist on full and timely reporting by member countries of standard sets of data, provide sharper policy advice to all governments, and deliver franker messages to countries that appear to be avoiding necessary actions. The 1994-95 Mexican case in particular has highlighted a difficult problem about the supply of information. The regular release of informa­ tion is a way of depoliticizing its supply and leaving thousands of market participants rather than any one political body to make judgments about the appropriateness of policy. But this should be supplemented by con­ fidential advice on how markets are likely to respond to particular policies or developments. In the era of globalization, the IMF has a function in providing advice backed by resources, or the potential or eventual supply of resources (the traditional effect of IMF conditionality). In terms of size, however, these operations will inevitably be outweighed by market flows. Here the provision of information is a way of involving the world's capital markets in practice in the operation of surveillance. This development involves a logical extension of the Bretton Woods vision. For a long time the practice and effectiveness of surveillance had been stymied by concerns about confidentiality of data provided and of policy recommendations. But greater publicity opens the way for the institution of a much more effective surveillance mechanism. The provi­ sion of better and more extensive information. more regularly, and in a more standardized form, allows the international financial system to make judgments on policy and supply funds-in other words, participate in the exercise of surveillance, which in the Bretton Woods framework was much more exclusively the preserve of the IMF.

40 IMF, Annual Report (1995), p. 210. Interim Committee Communique. April 26, 1995.

©International Monetary Fund. Not for Redistribution 790 HAROLD JAMES

REFERENCES

Verbatim, Vol. Chronique des annees /981-1986 Attali, Jacques, 1: (Paris: Fayard), 1993. Baker, James, Speech to the Council of Foreign Relations. May 20. 1988. repro­ duced in U.S. Treasury News (May 1988). Barrionuevo, Jose M., "The Accuracy of World Economic Outlook Projections , for Industrial and Developing Countries,. Annex li! to World Economic Outlook, by International Monetary Fund, World Economic and Financial Surveys (Washington: IMF, October 1992), pp. 77-8 1. Boughton. James M., "France and the IMFin the Floating-Rate Era: What Role for Surveillance?" Paper prepared for conference on "France and the Bret­ ton Woods Institutions 1944-1994" (Paris: International Monetary Fund, June 30-July 1, 1994). Cooper, Richard N., "Panel Discussion," in IMF Conditionality , ed. by John Williamson (Washington: Institute for International Economics. 1983), pp. 569-77. Crockett, Andrew D., "The International Monetary Fund in the 1990s,·· Govern­ ment and Opposition, Vol . 27 (Summer 1992), pp. 267-82. Dell, Sidney, "On Being Grandmotherly: The Evolution of IMF Conditionality," Princeton Essays in International Finance, No. 144 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, October 1981). de Vries, Margaret Garritsen, The International Monetary Fund 1972-1978: Cooperation on Trial (Washington: IMF, 1985). Dobson, Wendy, "Economic Policy Coordination: Requiem or Prologue?" Pol­ icy Analyses in lmemational Economics, Vol . 30 (Washington: Institute for International Economics, 1991). Funabashi, Yoichi, Managing the Dollar: From the Plaza to the Louvre (Washing­ ton: Institute for International Economics, 2nd ed., 1989). Gold, Joseph, Conditionality, IMF Pamphlet Series, No. 31 (Washington: Inter­ national Monetary Fund, 1979).

---, "Developments in the International Monetary System, the International Monetary Fund, and International Monetary Law Since 1971," in his Legal and Institutional Aspects of the International Monetary System: Selected Essays, Vol . 2 (Washington: IMF, 1984), pp. 17-254. Group of Ten. Ministerial Statemem of the Group of Ten, and Annex Prepared by Deputies (Paris: G-10, 1964). Guitian, Manuel, Fund Conditionality: Evolution of Principles and Practices , IMF Pamphlet Series, No. 38 (Washington: International Monetary Fund. 1981). Horsefield, J. Keith, The International Monetary Fund 1945-1965: Twemy Years of International Monetary Cooperation (Washington: IMF, 1969). International Monetary Fund, Annual Reporr of the Executive Directors for the Financial Year ended Apri/ 30 (Washington: IMF. various issues).

©International Monetary Fund. Not for Redistribution PRINCIPLE OF SURVEILLANCE 791

--, Articles of Agreement of the International Monetary Ftmd (Washington: IMF, 1988).

--, Selected Decisions and Selected Documents of the lntemationafMonetary Fund, (Washington: IMF, 1993). Kenen, Peter B., "Capital Controls, the EMS. and EMU," Economic loumaf, Vo l. 105 (January 1995), pp. 181-92. Kunz. Diane. ''Cold War Dollar Diplomacy: The Other Side of Containment," in The Diplomacy of the Crucial Decade: American Foreign Relations During the 1960s,ed. by Diane Kunz (New York: Columbia University Press, 1994). Lawson, Nigel, The View from No. 1/: Memoirs of a Tory Radical (London: Bantam, 1992). Nau, Henry R., The Myth of America's Decline: Leading the World Economy into the 1990s (New York: Oxford University Press. 1990). Pohl, Karl Otto, "You Can't Robotize Policymaking," International Economy . Vo l. 1 (October/November 1987), pp. 20-6. Putnam, Robert D., and Nicholas Bayne, Hanging Together: Cooperation and Conflict in the Seven-Power Summits (Cambridge, Massachusetts: Harvard University Press, 1984). Sachs, Jeffrey D., "The Current Account and Macroeconomic Adjustment in the 1970s.'' Brookings Papers on Economic Activity: 1 (1981), The Brookings Institution, pp. 201-68. Van Houtven, Leo, "Half a Century After Bretton Woods: The Role of the IMF in the [nternational Monetary System," in Monetary Stability Through !mer­ national Cooperation: Essays in Honour of Andre Szasz. ed. by Age Bakker and others (Dordrecht: Kluwer, 1994), pp. 283-96. Working Party Three, The Balance of Payments Adjustment Process (Paris: OECD, 1966). Yeo, Edwin, Ill, Attachment to letter to George Schulz, 1976. reproduced in InternationalCooperation Since Bretton Woods, by Harold James (Washing­ ton: IMF and Oxford University Press, 1996), pp. 268-9.

©International Monetary Fund. Not for Redistribution StaffPOfHrs lMF No. 4 (December Vol. 42, International Monetary1995) Fund Cl 1995

Conditionality: Past, Present, Future

MANUEL GUITL.\N*

The paper first discusses the rationale for conditionality and its analytical framework. It then traces the practice of conditionality as developed by the /M F and its membership over the first 45 years of the institution's existence. Next, the paper focuses on current is sues and practices in the imple­ mentation of conditionality. The challenges that conditionality is likely to confront as the IMF moves into the next century are examined. Finally, the paper broadly assesses the role and effectiveness of conditiorzality and discusses some of the broader is sues that the IMF will face in an in­ creasingly interdependent and global economic environment. (JEL E5, E61, F33]

HE INTERNATIONAL MONETARY FUND was created half a century ago Twith a clear and specific mandate: the promotion of economic and financial cooperation among its member countries. Such cooperation was believed to be the best means for the attainment of a number of interna­ tional economic objectives that were considered essential for the eco­ nomic welfare of the world community. These objectives, which were explicitly laid out in Article I of the IMF's Articles of Agreement, included: the expansion and balanced growth of international trade; the promotion of exchange stability; and the elimination of foreign exchange restrictions through the establishment of a liberal multilateral system of current international payments and transfers. These broad aims would

• Manuel Guitian is the Director of the Monetary and Exchange Affairs De­ partment. He is a graduate of the Universities of Santiago de Compostela (Law) and Madrid (Economics) and received a Ph.D. in economics from the University of Chicago. Conditionality is one of the subjects under permanent examination in the institution in Board reports, staff papers, and other published and internal documents. For the preparation of this paper, the author relied extensively on these materials as well as on his own paper on conditionality of a decade and a half ago; see Guitian (1981). 792

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 793 provide the basis for the promotion and maintenance of high levels of employment and real income through the development of the productive resources of all members. To these ends, the institution was "to give confidence to members by making the general resources of the Fund temporarily available to them under adequate safeguards , thus providing them with an opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity" (Article I (v); emphasis added). In this manner, the duration and degree of disequi­ Libria in country external payment balances would be contained. There­ fore, from its very inception, it was foreseen that the IMF would act as a source of financial support to member countries facing actual or poten­ tial balance of payments problems. In the process of discharging this responsibility, the institution has come to play a central role in the international monetary sphere.1 The principles to which the institution's financial assistance must conform are already explicit in the general formulation of Article I(v), which calls for the IMF to adopt policies on the use of its resources: (1) that would assist members to overcome their balance of payments difficulties; (2) in a manner consistent with the purposes of the institution; and (3) under adequate safeguards to ensure that such use would be temporary. During its long and close relationship with a varied and growing mem­ bership, the IMF has developed, and has adapted continuously, a prag­ matic and flexible body of policies and procedures to govern the access to, and the use of, its resources by member countries. This body of policies and procedures has come to be known by the term conditionality . As a principle, conditionality applies in most of the IMF's financial support operations with its members. This is clearly required by the Articles of Agreement. Its actual implementation, though, takes account of developments in the international economic environment. Therefore, practices of conditionality have been formulated throughout the IMF's existence in response to the needs of its members and those of the world economy.2 Of particular importance within this evolution has been the

1 This role encompasses responsibilities that �o beyond the purely financial domain, such as those the IMF has on the jurisdJctional and surveillance fronts. See, for elaboration, Guitian (1992a); for a detailed examination of the history an� activities of the £MF through the late 1970_s, see Horsefield (1969) and de Vnes (1976 and 1985). And for a recent revtew of the role of the IMF in international monetary cooperation, see James (1996). • 2 A !ecent a�d exc�llent examiryation of the evo)ving character of IMF's condi­ uonabty practtces wJII be found m Polak (1991); see also Finch (1989). Earlier discussiOn is contained in Guitian (1981) and Williamson (1983).

©International Monetary Fund. Not for Redistribution 794 MANUEL GUITL.\N existence of a general consensus that the IMF's financial assistance should be conditional on the adoption and implementation of adjustment policies. The purpose of this paper is to examine the principle and the practice of conditionality as both of them have evolved since the inception of the IMF. The paper first discusses the rationale for the concept of condition­ ality. Then, it outlines its analytical framework and the corresponding conceptual underpinnings and economic policy implications. Against this background, the paper traces the practices of conditionality as developed by the IMF and its membership over the first four and a half decades of the institution's existence (rhe pasr). The paper then focuses on current issues and practices in the implementation of conditionality (the present). And from there, it turns to an examination of the challenges the exercise of the IMF policies on conditionality is likely to confront as the institution moves into the next century (thefurure ). Finally, the paper concludes with a broad assessment of the role and effectiveness of conditionality as well as a discussion of some of the broader issues that the IMF will have to face in an increasingly interdependent and global economic environment. As supporting background, the Executive Board decision that lays out certain key aspects of conditionality and a summary description of IMF financial facilities are included as annexes to the paper.

I. Rationale

The prescriptions that the use of IMF resources should be temporary and that it should be made subject to adequate safeguards contained in Article (v) receive further elaboration in the Articles of Agreement, in particularr in Article V, Section 3. Essentially, the elaboration is to make clear that the institution's financial support is intended to assist members in the correction of their external imbalances "in a manner consistent with the provisions of this Agreement" (Article V, Section 3(a)). Or to put it differently, a central aim of the provision of IMF resources is to ensure the observance of the institution's code of conduct. And members seeking to use them are expected to have an actual or potential balance of payments need, which they undertake to eliminate by the adoption of measures compatible with their obligations under the Articles of Agree­ ment.3 These general principles were then translated into operational

3The references to the Articles of Agreement in the text are to those amended effective November 11, 1992; see IMF (April 1993). Broadly speaking, the analysis applies also to earlier versions of this basic charter. See 1-Iorsefield

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 795

procedures by means of Executive Board decisions, which adapted the implementation of conditionality as the world and members' economies evolved.4 Behind these prescriptions in the Articles there were a number of important aims, which provide a rationale for conditionality and warrant explicit discussion.

Moral Hazard

In common with other instances of government or official intervention in the economic sphere, the availability of IMF financial assistance could have undesirable effects on member country behavior. It is conceivable that a country's concern about the avoidance of imbalances could be unduly lessened by the prospect of the financial support that the institu­ tion is mandated to provide. IMF resource availability can carry with it, therefore, a risk of moral hazard, which needs to be contained. A first aspect of conditionality in the use of IMF resources that must be stressed, as a consequence, is its character as an instrument to contain moral hazard. Apart from the direct economic terms of IMF resources (that is, their interest and amortization payments or, as they are institutionally labeled, charges and repurchases), the access to and disbursement of those resources is contingent on the adoption and pursuit of economic adjustment measures that give an assurance that the imbalance will be redressed. This aspect of IMF financial support was behind early discussions of the right to draw on the institution's resources, which sought to balance the need to protect the member's interests with those of the IMF. On the one hand, it was felt that it would be desirable to provide members with certainty about their right to draw on the resources of the institution. In the extreme, this line of reasoning led to the advocacy of automaticity in drawing rights.5 On the other hand, an equally important consideration was the need to protect the institutional interest by ensuring that mem­ bers using IMF resources would undertake the required adjustment effort. This line of argument, of course, led in the direction of condi-

(1969), Vol. 3 for the original text of the Articles of Agreement. See also Gold (1 979) for an extensive discussion of conditionality under the first and the second amendments of the Articles of Agreement. 4 See IMF (1994), in particular Executive Board Decision No. 6056-(79/38) on Guidelines on Conditionality (reproduced in Annex 1). See also Gold (1919). sUp to a maximum limit in relation to a member's quota. Quantitative limita­ tions were, from the beginning, generally agreed by all parties; see Horsefield (1969), 1, 67 Vol. p. ff.

©International Monetary Fund. Not for Redistribution 796 MANUEL GUITIAN

tionality in drawing rights and, as already noted, is the one on which consensus was finally reached. In sum, the availability of IMF financial assistance provides a measure of insurance against adverse shocks and the emergence of external imbal­ ances. In common with aU insurance activities, it raises the specter of moral hazard in the form of less determination on the part of member countries toward the adoption or maintenance of appropriate policies. By making its resources conditional on the implementation of adjustment measures, the risk of moral hazard is thus contained.

IMF Asset Portfolio Protection

A related, but separate, aspect of the extension of financial assistance is concerned with the protection of the quality of the IMF's asset port­ folio. Through its lending to member countries, the IMF invests its resources in them and therefore, its asset portfolio cannot but be contin­ gent on their economic performance. Given its focus on the appropriate­ ness of country economic policies, conditionality is therefore also an instrument to protect the quality of the institution's assets. As such, conditionality procedures represent the essence of the "adequate safe­ guards" prescribed by the Articles of Agreement. From this standpoint, it is only natural that the IMF has consistently urged members to adopt corrective measures at an early stage of their imbalances, measures that couJd. of course, be supported by use of the institution·s resources. Here the aim has been to contain not only the adjustment cost to the member, but also the risk to the IMF's investment in the latter's adjustment effort. When members unduly postpone those efforts and allow the imbalance to grow, the costs of correction are likely to increase, and with them, the risks embedded in IMF financial support.

Cooperative Nature of Arrangement

A third aspect of IMF financial assistance on which conditionaJity has a strong bearing is the cooperative nature of the institution. Such nature, on the financial front, arises from the fundamental sources of IMF loanable funds, which are the quota subscriptions of members.6 All

6 Over time, the IMF has been able to supplement its quota-based resources by contributions of specific members under a variety of schemes. such as the General Arrangements to Borrow, the (now lapsed) supplementary financing facility, and the (also lapsed) enlarged access policy; see IMF (1994).

©International Monetary Fund. Not for Redistribution COND!TIONALITY: PAST. PRESENT, FUTURE 797

members have the same rights and obligations under· the Articles of Agreement. One implication of this is that countries facing an actual or potential external imbalance and willing to undertake a commensurate adjustment effort are all in principle entitled to request the institution's resources in support of that effort. Such potential entitlements require that the institution continuously ensure that sufficient funds are prospec­ tively available. In turn, given the overall constraint on IMF resources, this entails that the use of those resources be revolving. Not only that, but in an uncertain environment such as the one that typically character­ izes the international economy, they will also require that the IMF's resource portfolio exhibit a significant measure of liquidity. Optimally, IMF lending should be indeed temporary. But further, it should not normally exhaust the supply of the institution's loanable resources. In effect, the IMF should not generally be loaned up, so to speak, so that its portfolio, as a norm, may remain appropriately liquid. Only thus can the revolving nature of IMF resources and the monetary character of the institution both be preserved. In sum, conditionality in the use of IMF resources is indeed a mul­ tifaceted instrument. Its basic aim is to enable members to overcome their balance of payments problems by aligning their adjustment efforts with their adjustment needs. But in the pursuit of this aim, the exercise of conditionality helps balance a multiplicity of interests. Nationally, it contributes to the containment of adjustment costs relative to what they would otherwise be, thereby enhancing economic welfare. Inter­ nationally, it helps to limit moral hazard risks, protects IMF assets, and safeguards the institution's monetary identity.'

11. Framework

The attainment of all these diverse goals called for the development of an operational framework for an objective implementation of condition­ ality. And such a framework evolved over time based on the relationship that could be established between an external imbalance (what has been called above adjustment need) and the economic policy measures re­ quired to correct it (what has been referred to above as adjustment effort).

7lt should be stressed, though, that conditionality is not the only instrument on which the IMF counts to further its institutional interests and purposes. Surveillance over members' economic policies is the other, as important, if not more so, instrument at its disposal. See Guitian (1992a) for elaboration. On the monetary identity of the IMF, see Duisenberg and Szasz (1991 ). See also Guitian (1994a).

©International Monetary Fund. Not for Redistribution 798 MANU EL GUITJ..\N

This is a dynamic relationship that reflects country-specific characteristics as well as those of the international economy. But it does exhibit a basic set of features that are founded in economic theory and reality, on which conditionality practices are anchored.

Adjustment Need

A central element in the implementation of conditionality is an assess­ ment of the nature and characteristics of the disequilibrium that requires correction. On the most general leveL an actual or potential imbalance may arise in a variety of manners and for a variety of reasons, all of which will have a bearing on the strategy for its elimination. Indeed, the perceived dimensions of an economic imbalance also evolve with time, this being an important factor behind the continuing adaptations of conditionality.8 Broadly speaking, an imbalance arises whenever the sum total of demands for resources in an economy exceeds the amount of those resources that can be generated internally plus those that can be attracted from abroad on an appropriate scale and on sustainable terms. But it can also surface because of inefficiency in the use of the resources that are available; that is, on account of distortions that constrain the level or rate of expansion of aggregate supply and thus keep the economy operating under its capacity or below its potential rate of growth . Aggregate im­ balances that conform to this general description recur in all economies. And important for their resolution will be an assessment of their origin (e.g., are they the result of internal or external factors?); and their nature (e.g., do they reflect exogenous or endogenous causes? are they transitory or permanent?}" All these characteristics are critical for the design of a corrective strategy, although it must be admitted that they are not generally easy to ascertain in practice. But there is one issue that needs to be kept under close examination in this context. That is the correspondence in time between the development of an imbalance and the implementation of a plan for its correction, both of which are flow concepts. The extent to which both are commensurate in scale and timeliness will determine whether or not the imbalance is being passed on from one period to the

RThe use of varying economic terminology in connection with the process of adjustment is one of the indications of these diverse perceptions. Thus, we have seen terms like stabilization, adjustment, structural adjustment, development, and growth replace one another as key features of the process in the last five decades. 9See, forfurther discussion, Guitian (1981). Nowzad (1981). and Tanzi (1987).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FUTURE 799

next. Such adjustment delays are important, if only beaause they affect the characteristics of the imbalance by tending to let it cumulate into a (larger) stock disequilibrium. ln this manner, for example, a transitory imbalance that is not corrected opportunely may turn out to be less temporary than originally envisaged. Or one that is created exogenously, if not corrected promptly, may give rise to endogenous distortions that compound it. 10

Adjustment Goals

Strategies for the adjustment of economic imbalances invariably seek to attain a broad range of objectives. And the choice among those objectives reflects economic as well as other considerations. Goals typi­ cally pursued on economic grounds include a sound growth rate, an appropriate level of employment, domestic price and exchange rate stability, and a viable balance of payments position. There are also other policy objectives pursued on the basis of economic, social, and political considerations, such as those that relate to the equity and distributional consequences of economic policies.11 Clearly, the scope for the attain­ ment of the full range of such diverse aims is bound by the relation­ ship that prevails between the amounts of required and available re­ sources as well as by the efficiency in their allocation and use. The process calls for delicate and complex choices on the mix of objectives and the relative speed of their achievement, choices that are an integral part of economic policy decision making and government responsibility. Indeed, the appropriateness of such choices can be the critical element for the sustainabi)jty of the adjustment effort . Within the broad array of policy goals all countries seek to attain, the IMP's mandate directs its attention toward those of an external nature, e.g., exchange arrangements, exchange rates, and the balance of pay­ ments. The institution shares an interest with members on the attainment

1° Considerations of this nature are behind the emphasis that conditionality practices put on the timely introduction of corrective measures; see Executive Board Decision No. 6056-(79/38) on Guidelines on Conditionality, reproduced in Annex I to this paper. 11 For a recent example of recent discussions of these issues in the IMF, see the papers presented at a conference on Income Distribution and Sustainable Growth h eld at IMF headquarters on June 1-2, 1995, and in particular those conducted in the session on Equity Issues in IMF Policy Advice. In practice. the IMF has responded to members' concerns with their multiple aims by making explicit the relationship of economic policies to a variety of objectives, including. for exam­ ple. poverty alleviation and the environment: see Polak (1991 ).

©International Monetary Fund. Not for Redistribution 800 MANUEL GUITI.A..N of other economic policy goals, particularly since such attainment will often be important for the maintenance of a viable external payments position. The perspective, though, is their contribution to the purposes and aims of the institution, that is, to the maintenance of a balanced national and international environment in the context of a liberal multi­ lateral payments regime. For these reasons, the IMF exercises an impor­ tant measure of circumspection with regard to the domestic policy objec­ tives of member countries. Subject to this proviso, TMF conditionality practices have made increasingly explicit the connection between eco­ nomic policies and domestic policy aims, and in the process, they have underscored the relevance of these aims for the external performance of country economies and for the international system as a whole.

Adjustment Effort

A country's adjustment effort, that is, the formulation and implemen­ tation of policy measures to correct an aggregate imbalance, will be inevitably linked to the particular characteristics of that country. And conversely, these characteristics are behind the range of economic poli­ cies on which IMF conditionality practices focus. Amount of Resources and Macroeconomic Balance It follows from the above discussion of adjustment need that a key function of economic management is to keep the level and rate of growth of aggregate demand in line with those of the economy's productive capacity and resource base. For a given scale of the latter and for a given structure of relative prices and costs, this will entail domestic financial policies that are consistent with macroeconomic balance in the economy. Three major policy areas can be identified in this respect. Expansion in aggregate demand frequently reflects imbalances in the fiscal accounts and, more broadly, in the public sector finances.12 The correction of disequilibria originating in the public finances encompasses the fiscal dimension of macroeconomic management. This includes policy measures that seek to curtail fiscal spending or raise fiscal revenues to restore viability to the public sector finances. The specific mix of mea-

12This is a policy area on which views have been significantly adapted since the IMF was established, most particularly in recent years. The adaptation reflects a new consensus on the role of government in the economy, which has moved in the direction of favoring support of market forces rather than interference with them. See, for further discussion, Guitian (1996).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 801

sures chosen will influence the general performance of the economy, which will vary depending on whether the strategy is based on measures to control outlays or on revenue-raising actions. The former will tend to restore balance to the economy by lowering the weight of the public sector in aggregate demand. In contrast, raising revenue will likely entail a reduction in the share of private demand. The mix of fiscal measures will also influence the speed and certainty of the adjustment process. This is because expenditure reductions and revenue mobilization are not equally effective: a government's scope to control its outlays usually exceeds its ability to increase its receipts. There is a close relationship between fiscal outcomes and the broader sphere of financial policies, that is, those related to developments in credit and money flows in an economy. Indeed, it is not always easy to distinguish them unambiguously, as fiscal policy basically determines the public sector borrowing requirement, in general, and, in particular, its need for domestic bank financing, which is a key aspect of monetary policy. It is well known and accepted that maintaining aggregate demand on a sustainable path calls for some control over the flows of domestic financing and specifically over the rates of monetary and domestic credit expansion. This is the monetary dimension of macroeconomic manage­ ment. It stresses the importance of the link between domestic credit expansion and money supply increases, on the one hand, and their relationship with aggregate expenditure and income, on the other. Or to put it differently, it highlights the fact that a discrepancy between the supply and demand for money (a money market imbalance) has a coun­ terpart in an imbalance between expenditure and income (a goods market disequilibrium). Restoration of a sound relationship of expenditure to income, therefore, will entail keeping domestic credit expansion in line with the prospective path of desired money holdings in the economy. These considerations are behind the emphasis placed on domestic credit expansion as a policy variable in the application of conditionality.13 Consistency requires that both fiscal and monetary policies be comple­ mented by appropriate foreign borrowing strategies, the external debt dimension of macroeconomic management. The macroeconomic impli­ cations of external debt management policies derive from the direct influence they can have on the expenditure-income flow and from the substitutability that exists between foreign and domestic credit. In essence, a central aspect of macroeconomic management is to keep the

13 For further elaboration on this point, see Guitian (1973 and l994b) and Polak (1991).

©International Monetary Fund. Not for Redistribution total (i.e.. dome-;tic and foretgn) !low

©International Monetary Fund. Not for Redistribution tONDI rJO\.;t\1 11'\ PAS I PRI'SI;\1'1 ll ll RI XOJ

of i� C\Sl'ntial to en•;ure that uli::ationeconomic incentive�£' \change and and price trade signal-. n•ginH'\ fulfil!al-;(l then· function�.

111. Past

The framework just described will provide the setting for an examina­ tion of the evolution of IMF conditionality practices. In the process. it will be made clear the extent to which imbalances. the policies designed 10 deal with them, and, correspondingly. the exercise of conditionali1y have varied over time.

Rules Under Bretton Woods

The conditionality practices developed during the Bretton Woods pe­ riod (approximately 1945-1970) reflected the prevalence of the rule­ based regime agreed at the 19-t-4 Bretton Woods conference that estab­ lished the IMF. The world economy was then characterized by relative stability in inflation and growth rates and by substantial progress tOward the liberalization of international trade and payments and current account convertibility under a par value system. The principle of conditionality was soon incorporated into the in­ stitution's lending activities. 14 An early consensus emerged that the lMF's attitude toward a member's request for financial assistance would be guided primarily by the Fund's judgment on whether the mem­ ber's policies were adequate to cope with its balance of payments problem so as to enable it to repay the institution within three to five years from the date the resources were received. as prescribed by the Articles of Agreement (Article V, Section 7(c)). The IMF developed the srand-by arrangement as the main instrument to provide members with conditional access ro its financial resources. The stand-by arrangement represented a line of credit outlining the circum­ stances under which a member could make drawings on the IMF.1� In the early days, the arrangement was conceived mainly as a precautionary device to assure members that had no immediate need for IMF resources

14 See in particular Executive Board Dccbion No. 102-(52lJ I ). pp. 56-58 in IMF (1994). where general criteria arc set out for use of IMF resources. For a discussion of early conditionality practices. see Mookcrjce (1966). 1' See Annex 11 to this paper for a summary description of IMF facilities. And for a complete analy!.is of the evolution o(stand-bv· arrangements during the 1950!. and 1960s. see Gold (1970).

©International Monetary Fund. Not for Redistribution 804 MANUEL GUITI..i.N but felt they might require such financing in the near future (hence, the label "stand-by") that they would have access to it. It rapidly became evident, though, that the stand-by arrangement was also a particularly well-suited vehicle for channeling IMF conditional resources to countries with urgent balance of payments financing needs. The stand-by arrangement, in its link of a member's access to IMF funds to its adoption of a program of action, represented the first formal manifestation of the institution's belief that its assistance would be most effective if it was provided to support a member"s policies designed to correct its external imbalance. The policy programs, at that time, did not extend beyond one year, a period short enough to permit an economic forecast to be made, but long enough to allow an assessment of the im­ plementation of policy measures as well as a judgment on whether these measures called for adaptation or modification. Of course, this practice did not mean that adjustment was expected to be completed within such a limited horizon. In fact, members often entered into consecutive stand-by arrangements, a strategy that provided them with assurances of continued IMF financial support and the institution with evidence of policy continuity and determination. With experience, it became clear that to be effective, an economic policy program had to be as specific and precise as possible. In particular, the main policy targets and instruments capable of relatively accurate measurement came to be stated in the financial programs in explicit quantitative terms.16 These quantified instruments and targets in turn became the guideposts for assessing whether programs were being imple­ mented satisfactorily and their objectives were being attained corre­ spondingly. For the reasons elaborated in the discussion of the frame­ work, the variables most frequently used related to the expansion of domestic credit by the central bank or the banking system; the reliance of the government or the public sector on domestic bank financing or on short- and medium-term foreign borrowing; the management of interna­ tional reserves; and the establishment of realistic prices and appropriate incentives. A qualitative condition common to all stand-by arrangements was the avoidance of reliance on the introduction or intensification of exchange restrictions as a means of coping with balance of payments problems. The formulation of quantitative policy programs was accompanied

1°For an excellent discussion of financial programming, see Robichek (1971 and 1985).

©International Monetary Fund. Not for Redistribution CONDITIONAL!TY: PAST. PRESENT. FUTURE 805 by two developments that became common features of stand-by ar­ rangements. The first related to the way IMF resources were made available. Because policy actions were undertaken over a given period, access to the resources came to be phased over that period, to provide incentives to policy implementation and to prevent unduly rapid rates of drawing. In the process, the total amount of assistance came to be disbursed at specified intervals in installments that were linked to the periodic quantified policy implementation targets. 17 The second development was the inclusion in stand-by arrangements of provisions, referred to as performancecriteria, that had to be observed to ensure continued access to IMF resources. Here again, the rationale for these operational features of the arrangements was that failure to comply with the performance criteria served as a signal that the policy program should be reviewed. The restoration of drawing rights in case of deviations required either an amendment of performance criteria or new understandings on the policies to be pursued and measures to be taken over the remainder of the program period. In practice, the resump­ tion of the right to draw under an arrangement was based on the follow­ ing: the extension of a waiver for the noncompliance with performance criteria (typically in situations of small or reversible deviations); a mod­ ification of the criteria (when their fulfillment became unfeasible); or the replacement of an old arrangement with a new one (when modifications appeared insufficient). The conditionality practices developed during the 1950s and 1960s were reviewed by the IMF's Executive Board in 1968. This general review encompassed aJl aspects of the subject and concluded with a decision that summarized the major elements of prevailing policies on access to IMF resources.18 The decision stressed the importance of ensuring adequate safeguards to preserve the revolving nature of those resources and the need to alJow for flexible, yet uniform, treatment of alJ members. Subse­ quently, the Executive Board undertook periodic reviews of experience with specific performance criteria, such as those involving domestic credit ceilings, foreign borrowing limitations, and balance of payments tests. These reviews and the resulting Board decisions helped to further define and adapt the IMF's conditionality policies.

See Executive Board Decision No. 7925-(85/38), as amended by Decision No. 8887-(88/89),17 in IMF ( 1 994). pp. 87-89, for further elaboration. 8 See Executive Board becision No. 2603-(68/132), reproduced in Guitian (1981),1 pp. 44-45.

©International Monetary Fund. Not for Redistribution 806 MANUEL GUITIAN

Discretion in Policy Design and Implementation

By tbe time the policies on use of lMF resources were thus being consolidated, pressures had begun to mount on the international econ­ omy and, with them, on the Bretton Woods regime. 19 A succession of crises in foreign exchange markets during the late 1960s had cast growing doubts on the continued viability of the par value system. By 1971. the fixed (but adjustable) exchange rate regime established at Bretton Woods had ceased to be operative, and by early 1973. it l1ad been abandoned by the major industrial countries. For many of these coun­ tries, an international monetary system based on flexible exchange rate arrangements came into effect. Thus, the rule-based international monetary regime that had been introduced in Bretton Woods came to an end. And to replace it, a discretion-based system was adopted that has prevajJed since then. The Tu rbulent 1970s Early in the decade, it became evident that the changes in the interna­ tional economic environment and regime would require significant adap­ tations in the IMF's conditionality practices. Relatively large increases and marked shifts in external payments imbalances called for a blend of adjustment and financing that differed from that formerly incorporated in stand-by arrangements. And a reconsideration of the size and length of those arrangements was also in order.10 The payments imbalances facing many member countries in the early 1970s, judging by their scale and nature, required longer periods of adjustment than envisaged in the conditionality practices prevailing at the time. Correspondingly, they also involved larger amounts of assistance than could be made available normally under stand-by arrangements. Consequently, the IMF established in 1974 an extended facility to provide

19 For a detailed exposition of this period. see de Yries (1976). 20The adaptations mcluded the introduction of transitory facilities that pro­ vided resources with relatively low conditionality standards. Thus, in 1974 an oil facility was established to assist members in the financin& of oil import-related balance of payments deficits. This facility was extended m 1975. Recognizing, however. that adjustment to new oil price levels required specific measures, the conditionality standards were tightened. The oil facility was terminated in March 1976. In that year the IMF also introduced, on a temporary basis, the Fund, a facility for the benefit of low-income countries. As a source of directTr usr balance of payments loans the Trust Fund was terminated in 1981. See Guitian (1981) for further discussion of this subject. See also Annex Il to this paper for an updated description of IMF faci lities.

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FUTURE 807

member countries experiencing particularly severe balance of payments problems with commensurate medium-term assistance. The extended facility was designed to alleviate two main categories of payments problems: (1) severe payments imbalances that were due to structural maladjustments in production and trade, where cost and price distortions were widespread and long-standing; and (2) imbalances that were due to a combination of slow growth and an inherently weak balance of payments position that constrained the pursuit of active development policies. Arrangements under the extended facility covered periods of up to three years (which now may be extended to four years in exceptional circumstances), and the assistance they provided was to be repaid within four and a half to ten years from its receipt. The framework for conditionality discussed earlier remained applica­ ble for operations under the extended facility. The adaptation they en­ tajled was not so much in the degree of conditionality as in the strong assurance given by the member that the adjustment effort would be sustained over the medium term. The assurance was embedded in com­ prehensive economic programs that included policies of the character and scope required to correct structural imbalances in production, trade, and prices. Particular attention was given to measures intended to mobilize domestic and foreign resources, to improve their utilization, and to reduce impediments to international transactions. As the decade evolved, further adaptations had to be made, par­ ticularly in the scale of IMF financial assistance. External payments imbalances besetting the international economy were so large and persis­ tent that the need for conditional resources remained historically high. The fMF therefore took decisions that allowed for appropriate expan­ sions in the amounts of assistance it could make available to foster orderly adjustment and help promote sound expansion in the world economy.21 The important adaptations to the IMF lending practices in the 1970s led the Executive Board to undertake a second general review of condi­ tionality in 1978-1979. This review concluded with the adoption of a

21 The expansions in the scale of IMF conditional financing came first in the form of a temporary supplementary financing facility that became effective in 1979. Its resources were fully committed by early 1981, and the facility was then replaced by the enlarged access policy , which permitted the IMF to assist mem­ bers in coping with "payments imbalances that are large in relation to their quotas." Th e assistance was to be provided under stand-by or extended arrange­ ments. See Executive Board DecisiOn No. 6783-(81/40) in IMF (1994), pp. 181-4. An important adaptation in this area was the shift from the concept of limiting total cumulative use of IMF resources in relation to quota to the principle of annual use of those resources, albeit within a global maximum use. See, for elaboration. Guitian (1981).

©International Monetary Fund. Not for Redistribution 808 MANU EL GUITIAN decision setting out broad guidelines on the use of IMF resources.22 These guidelines stressed the importance of early adjustment, encouraged members to seek IMF assistance promptly, and acknowledged that the adjustment process stretched over the medium term. They also recog­ nized that measures likely to be required affected key and sensitive policy areas-fiscal, credit, incomes, foreign borrowing, and exchange rates, as well as trade and payments restrictions. They emphasized the need for the IMF to pay due regard to members' objectives and circumstances and to ensure that institutional lending conditions would be limited to those variables that were essential, because of their macroeconomic impact, for the effectiveness of the member's policies. The guidelines also confirmed the need for adequate safeguards for the institution and therefore called for reasonable assurance that programs would be carried out, a judg­ ment that might call for the prior adoption of measures critical for policy implementation. Lastly, the guidelines provided the basis for periodic reviews of conditionality practices, which have since taken place regularly. The 1980s: Decade of External Debt The 1980s will likely go down in economic history as the decade of the international debt crisis. This was a period in which debt-servicing diffi­ culties in the developing world became a virtually constant feature of the international economic scene. 23 From the outset, the IMF played a critical role in the resolution of the systemic threat posed by the debt crisis, a role that called for innovative approaches and continued adaptations to conditionality practices. At the time, the essential challenge confronting the IMF and its mem­ bers was to underpin the crucial, but elusive, relationship between adjust­ ment, financing (including external debt flows), and growth. Strengthen­ ing the relationship in a universal setting required the participation of all major parties. These considerations underlay the approach taken by the IMF to deal with problems of external indebtedness. The approach, as usual, entailed direct IMF financing for debtor countries; but its essential

22 See Executive Board Decision No. 6056-(79/38) on Guidelines on Condition­ ali reproduced in Annex I to this paper. ;rMuch has been written about the debt crisis since it irrupted onto the international stage in August 1982. A representative sample of volumes on the subject would include Mehran (1985), Frenkel, Dooley, and Wickham (198 . Dornbusch and others (1989), Calvo and others (1989), Husain and Diwan (1989w , and Stoll (1990). See also "The International Debt Crisis: What Have e Learned?"Cha pter Ill in Guitian 1992b) and the further references listed there; and Guitian (1992a). (

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FUTURE 809 focus was more than ever on the adjustment efforts of those countries. Thus, the IMF became the vehicle to which were attached all other elements of a cooperative strategy for the resolution of indebtedness problems. The central notion behind the strategy was that there was no substitute for sound domestic management in the debtor country. Thus, appropri­ ate, sustained economic policies were seen as necessary conditions for the solution of debt-servicing difficulties. The conditionality practices of the IMF sought to ensure that those necessary conditions were met and they were suitably adapted toward this end. In the early phases of the imple­ mentation of the debt strategy, the main (though not exclusive) focus was on macroeconomic management so as to balance resource demands with their availability. As it became progressively clear that the completion of adjustment would extend over the medium run, conditionality practices focused also on structural reforms and microeconomic measures to en­ sure efficiency in resource allocation and use as well as a resumption of growth. Despite their obvious importance, debtor adjustment efforts were not seen as sufficient conditions for the restoration of balance. It was also critical that those efforts be appropriately supported by the international community. This consideration led to an important innovation in IMF conditionality practices: the introduction of "concerted" lending pack­ ages.24 The IMF began to require creditors to provide firm assurances of the availability of external financing before it would move with its own support of the debtor adjustment program: a "critical mass" of commit­ ments of external assistance thus became a prerequisite for the comple­ tion of an arrangement with the IMF. In this way, the IMF helped catalyze capital flows toward countries willing and able to undertake an adjust­ ment in their economies. The aim was to attain a balanced distribution between the efforts required from debtors-that is, adjustment-and those required from creditors-that is, fi nancing. In the process, a measure of conctitionality came to be exercised on creditors as well. ' The policy programs also reflected the realities faced by debtor coun­ tries, which called for emphasis on the growth and structural reform aspects of the adjustment process. The scope of conditionality was cor­ respondingly broadened. Adaptations were made on the financing side as well. The concerted lending approach, originally focused on new loans and rescheduling or refinancing of old loans, evolved into progressively

2AFor excellent discussions of the role of debtors and creditors in the debt strategy, see de Larosicre (1986 and 1987).

©International Monetary Fund. Not for Redistribution 810 MANUEL GUJTJ.A.N sophisticated money packages, which introduced innovative financing modalities and techniques (such as debt-equity swaps, debt buybacks, exit bbnds, and the like). The approach also involved voluntary debt relief in the form of debt and debt -service reduction options. :!5 The debt crisis led to substantial innovations to conditionality prac­ tices, many of which still remain. It also demonstrated the difficulties of bringing together a group of separate and diverse interests. This simply reflected the tendency of each party to protect what it perceived as its immediate interest. Frequently, but not surprisingly, the resulting pres­ sures were directed toward the entity that pursued the common interest, in this case, the IMF. Possibly undue weight was given to the apparent substitutability between adjustment and financing. This led debtors to stress the importance of financing (the contribution of creditors), and creditors to underscore the necessity of adjustment (the contribution of debtors). There was a measure of logic in these tendencies, but from a fundamental standpoint, they concealed that adjustment and financing can be comp/ememary , in that the sounder and more credible the adjust­ ment, the more accessible the financing. This complementarity was the dimension of the relationship on which the IMF, in which both debtors and creditors are represented, based its approach to indebtedness issues. In the process, the institution contributed to the containment of moral hazard risks. Concern over the strength of adjustment efforts served to limit those risks on the part of debtors, as they would ensure the restora­ tion of debt-servicing capacity. But the risk of moral hazard can also arise with respect to creditors, and the requirement that they support the process with new financing or debt relief or both served to contain it. Thus, balance was sought in the management of these two opposite moral hazard risks. As already noted, many of the adaptations to conditionaJity practices undertaken in the 1980s have remained in effect since. The emphasis on efficiency, aggregate supply management, linkages between macroeco­ nomic and microeconomic performance, as well as the relevance of adequate financing packages were incorporated into IMF lending poli­ cies. Growth-oriented adjustment programs reflected an explicit concern with ensuring that the correction of external imbalances be effected in ways that would not impair growth prospects, a concern that had already been stressed in the extended fa cility , which underscored the importance

2�These stages of the debt strategies were associated with initiatives prOtJOSed by United States Treasury Secretaries at the time, James Baker (growth) and Nicholas Brady (debt reduction). See Erb (1990), Ortiz (1989). and Cline (1984 and 1995).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FU TURE 811 ofstructural reforms. Continued efforts were also made to protect adjust­ ment efforts from unexpected adverse contingencies. resulting the addition of a contingency component to the already existing compensa­in tory financingfacility. And the equity and distributional consequences of the adjustment process. always a matter of concern, were made more explicit.26 Operationally, this led to the establishment of concessional facilities, that is, the structural adjustmentfacility (SA F) and the enhanced structural adjustment facility (ESAF) , intended for, and available to. low-income developing member countries.

IV. Present

Recent years have witnessed marked changes in economic ideas as well as in realities in the international economy. These changes reflected a progressively widespread philosophy based on principles of democracy and freedom, which stressed the primacy of individual over collective rights. As the world economy entered the current decade, views in many countries on the role and scope of economic policy had begun to conform to this trend, exhibiting many characteristics in common with those that the IMF had been advocating in its conditionality and surveillance activ­ itiesY At the same time, the institution, together with its policies, was yet to confront one of the most difficult and important challenges of its existence: the integration into the international monetary system of the economies in transition from central planning to market-based economic regimes.

Consensus

Until relatively recently, general views on economic policy manage­ ment were based on an active, indeed dominant, presence of government in the economic process. This reflected the notion that besides providing public goods and correcting market failures, governments were also alone responsible for stabilizing cyclical economic fluctuations as well as for growth and development in the economy.

26See, for further discussion of growth-oriented adjustment programs, Corbo, Goldstein, and Khan (1987). See also Mohammed (1991) for a recent examina­ tion of IMF contingency financing and concessional support. 27 The existence and strength of this coincidence and the corresponding support of conditionality were the themes of Michel Camdessus·s (1994) closing remarks at a plenary session of the Annual Meeting on October 6 in Madrid.

©International Monetary Fund. Not for Redistribution 812 MANUEL GUITIAN

Experience with the pursuit of this principle gradually changed the perceived role of government. In the process, a certain common view developed on economic policy that fosters a new pattern of interaction between government and market forces. Rather than have government directly involved in the economic process through participation in the production and distribution of goods and services, government action should be directed toward fostering market forces, safeguarding com­ petition, and encouraging private initiatives. Thus, instead of compet­ ing and interfering with market forces, governments are to focus on establishing a framework in which those forces can operate efficiently. This economic policy consensus has spread widely across the IMF's membership, both in the industrial and developing world.28 It lays out basic responsibilities of government, which can be grouped into three broad categories. First, the establishment of a stable macroeconomic setting and the timely undertaking of policy adjustment efforts, whenever these become necessary to preserve or restore stability to the economy. Second, the protection and maintenance of the country's economic in­ frastructure, broadly interpreted to encompass investment in both human and physical capital. And third, the establishment, development, and safeguard of the economy's institutional infrastructure. This responsibil­ ity includes a host of traditional (and in some cases, not so traditional) governmental activities, such as the provision of an appropriate legal, regulatory, and social framework to support the fu nctioning of market forces . Key ingredients of such a framework, therefore, will be an ade­ quate incentive system and a favorable climate for competition, both of which will require the prevalence of an open and liberal economic regime. Clearly there are significant areas of common ground between this economic policy consensus and the policy framework underpinning IMF conditionality practices, both in concept and in practice. The attainment and maintenance of a stable macroeconomic setting invariably calls for sound fiscal and monetary policies as well as for prudent external debt management, three key dimensions of the conditionality framework. And the provision of an environment fa vorable to market forces , with its emphasis on competition and openness, coincides broadly with the IMF's emphasis on appropriate incentives, preservation of competi­ tiveness, and its institutional mandate of fostering liberal exchange and

28This point is elaborated fully in Guitian (1996). The convergence of views described in the text has come to be known as the •·Washington consensus," the label iven to it by Williamson (1990). It is also behind the "new develo ment � p3 paradtgm" in development economics; see Summe rs and Thomas (199 ) and Bruno (1995).

©International Monetary Fund. Not for Redistribution COND!T!ONALITY: PAST. PRESENT. FUTURE 813

trade regimes. As has been remarked in a different context, this coinci­ dence of ideas and policy approaches augurs well for economic man­ agement in the world in general, and for the acceptance of conditionality in particular.29Thus, the present can be described as a period in which IMF conditionality practices are enjoying a good measure of general concurrence.

Challenge

The international community faces a most difficult, but also stimulat­ ing, challenge as it seeks to integrate into its monetary system the economies of the countries in transition from central planning to market­ based regimes of economic organization. As far as the IMF as an institu­ tion is concerned, the challenge encompasses all its major activities. From the outset, the IMF's contribution started with the provision of technical assistance on monetary, exchange, and fiscal as well as statistical and legal issues. It also entails the exercise of surveillance in its extension of economic policy advice and periodic assessment of policy implementa­ tion. And it also, of course, involves the exercise of conditionality and the corresponding extension of financial support to help these members overcome the balance of payments needs arising during the process of reform. The economic policy consensus described above has provided a good setting for this purpose. Emphasis on the traditional policy areas on which IMF conditionality practices focus-that is, appropriate macro­ economic management, rational pricing, and open economic systems-is of course necessary for the process of reform to succeed. But the question of whether this is sufficient has to be addressed since transition economies initially lack markets as well as a market-supporting institutional frame­ work. Economic policy analysis generally presupposes the existence of these elements, and IMF conditionality practices have been based on them. Therefore, it has been necessary to supplement the traditional instruments of conditionality by focusing on the microeconomic and institutional aspects of reform.30 Accordingly, the IMF sought to adapt

29 On the promising prospects owing to the general acceptance of these ideas, see Schultz (199 5). And on the growing acceptance of conditionality, see Camdessus's (1994) closing remarks at the Annual Meetings in Madrid. 30The focus in the text is on lMF practices. But in addition to these, the coordination linkages with other multilateral institutions, such as the World Bank, the OECD, and the EBRD, have been strengthened. See Guitian (1992b and 1992c), where these issues as well as some other aspects of conditionality in the context of reform are discussed.

©International Monetary Fund. Not for Redistribution 814 MANU EL GUITlAN

(and innovate on) its conditionality practices to suit the needs of the reforming economies. In addition to activating fully its technical assistance and financial support capabilities from the outset of the reform process. the IMF also established a new instrument designed to address the specific needs of the transition to market-based regimes-the systemic transformation fa cility. This facility is intended to provide financial assistance to member coun­ tries with balance of payments difficulties stemming from disruptions to traditional trade and payments arrangements in the transition to multilat­ eral, market-based trade. In particular. it covers balance of payments problems resulting from a sharp fall of total export receipts as trade moves from nonmarket to market prices; a substantial and permanent increase in net import costs, owing to the same shift from nonmarket to market prices; or a combination of both. Access to the resources of this facility is subject to the general condition that the member requesting them will cooperate with the IMF in solving its balance of payments problems and that it will adopt as soon as possible policy programs that can be supported by the regular IMF facilities. Specifically, access to the systemic transformation facility requires sig­ nificant policy action in a number of areas: economic stabilization; con­ tainment of capital flight; and the implementation of structural and institutional measures needed to foster a market-based economic en­ vironment.31 The country is required to issue a policy statement describ­ ing the policy objectives, the macroeconomic outlook, the structural, fiscal, monetary, and exchange rate measures to be implemented, and, as applicable, a technical assistance program. At the latest before the second purchase under the facility, a quantified quarterly financial pro­ gram is required. Finally, the country must make a commitment to avoid introducing or tightening exchange and trade restrictions. In essence, the innovations introduc�d by the systemic transformation facility were twofold. One was to ensure that support could be provided during the stage of the reform process when the necessary conditions were not yet present for a fully quantified phased macroeconomic pro­ gram, as would be required by the regular IMF facilities, such as stand-by or extended arrangements. Instead, the stress was on specific policy actions that would help ensure that reform was proceeding and that policy implementation capabilities were being developed. The other inno­ vation was the explicit introduction of a technical assistance program into

31 See Executive Board Decision No. 10348-(93/61) STF. in IMF (1994). See also Annex Il to this paper and IMF Survey (May 3, 1993).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT. FUTURE 815 the conditionality requirements ofthe facility.32 The IMF had been mov­ ing in this direction from the outset of its relations with countries in transition. Examples of this approach can also be found in financial operations with other members in the developing world. In bringing out explicitly the link between conditionality practices and technical assis­ tance activities, the institution sought to enhance the importance of institution building as an element of the reform strategy. In this fashion, all aspects of the policy process have entered into the framework of conditionality: policy institutions, assessments and measurement of policy stances, and specific policy implementation tools.

V. Future

As the international economy moves into a new century, the IMF and its conditionality practices will continue to face challenges that will test the institution's ability to adapt and innovate. One of those challenges is, of course, to continue to support the reform programs of the economies in transition. Progress toward their goal of establishing market-based regimes and their complete integration in the international system has so far been uneven and conditional financial support from the IMF will therefore continue to be needed.

Global Markets

Another challenge, and one that has been evident already for some time, is that posed by the increasing globalization of financial and capi­ tal markets and the consequent growing interdependence of national economies that characterizes the world economic system. Progressively integrated capital markets are the logical result of the past fifty years' worth of effort by the IMF's members to fulfill their objective, and the institution's mandate, of opening and liberalizing trade and current account balances. Open national markets for goods and services, lead­ ing to growing trade flows, could not but enhance closer financial and credit ties among trading nations, which is an essential factor behind the internationalization of capital flows. The presence of global capital markets carries important implications for the IMF's activities. On the surveillance front, it affects the code of

32 For a description of the technical assistance services of the lMF. see IMF {1991).

©International Monetary Fund. Not for Redistribution 816 MANUEL GUITIAN conduct in the Articles of Agreement, which still envisages controls on capital movements as an instrument that may be used by countries to regulate them, as necessary.33 This provision, a relic from the original Bretton Woods par value regime, reflected several considerations of relevance at the time. One was the need to set priorities, which, given the disruptions to international economic relations owing to the World War II hostilities, focused on the restoration of trade and current account flows. Another was the aim of preserving a measure of independence for national economic policy and, in particular, for monetary manage­ ment. Yet another was that the scale of capital flows had until then been relatively limited. However, those flows of relevance for interna­ tional trade relationships were duly acknowledged and properly treated in the Articles of Agreement, though indirectly by including them in the definition of current transactions. 34 Despite the provision allowing for capital controls, as the years passed the international economy exhibited progressively liberal capital flows among countries. This trend, as noted above, was to a degree an in­ evitable consequence of the effectiveness of the integration of national markets for goods and services. And it was also probably fostered by the change of regime from par values to flexible exchange rates in the international economy. The challenge for surveillance, therefore, is to restore logic to the code of conduct. The gains from trade argument does not distinguish by the nature of the transactions involved, so what suits current transactions also holds for capital flows. Such restoration of logic will also serve to bring the code of conduct closer to the current reality of the international economy, instead of remaining. as at present, welt behind.35 As for conditionality, the current international economic setting poses difficult challenges indeed, both country-specific and systemic. Global capital markets increase the mutual interdependence of individual mem­ ber countries and, with it, their exposure or vulnerability to external developments. Conditionality practices will have to adjust in order to contain the resulting risks to national economies. The way to do this is to adapt conditionality practices as well as the

33See Article VI on Capital Transfers in IMF (1993). 34 See Article XXX on Explanation of Terms in IMF (1993). This Article includes as current transactions payments due in connection with normal short­ term banking and credit facilities and payments of moderate amount for amor­ tization of loans or for depreciation of direct investments, which are generally seen as capital transactions. 35For a more extensive discussion of these issues, see Guitian ( 1992c and 1993). For a recent examination of ca pital account convertibility issues and their impli­ cations for IMF policies. see IMF ( 1995).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 817 policy on access to IMF resources so that adequate finan'cial support can be made available to underpin members' policy efforts to cope with balance of payments problems associated with capital account transac­ tions. 36 This goes to the core of conditionality practices, particularly with regard to the phasing of drawings, their relationship with policy perfor­ mance, and their repayment. Capital account problems typically require a rapidly agreed and relatively large financial support package , in which a substantial share of the funds is made available up front. It may also be the case that, as the strategy works, not all the financing available to the country will be necessary. Or if it is, it may be repaid promptly. These features eaU for appropriate adaptation of IMF conditionality practices­ an adaptation that is likely to be complex, particularly in the area of as­ surances that disbursements go hand in hand with policy implementation. In a financially integrated world economic setting, an additional chal­ lenge for the IMF and its conditionality practices will be potential threats to the system's stability and soundness. Such threats could also arise, of course, in more restrictive contexts. But the more open and liberal the environment, the greater their probability and scale. These potential systemic threats will require the IMF to be ready to assist its members in a fashion similar to that of a central bank that stands ready to help its national banking system. In addition to calling for conditionality prac­ tices capable of helping to resolve capital flow problems, systemic chal­ lenges will also require adaptations in surveillance procedures to provide for monitoring of developments in the system as a whole.37

Moral Hazard and Market Failures

The evolution of the world economy has thus returned to the forefront the essential specific rationale for IMF conditionality, that is, the contain­ ment of moral hazard and the protection of IMF resources in a cooper­ ative setting. Of critical importance in this context is the interaction between governments' economic policies and the operation of market

36T his challenge brings up the close link between the code of conduct and IMF conditionality, which underscores the importance of updating the former to bring it in line with reality in the world economy. In its present version, members may not use IMF general resources to meet "large or sustained'' capital outflows. See Article VI, Section l(a) in IMF (1993). Some of the implications for IMF conditionality and surveillance in a setting of 37progressively integrated goods, services, and capital markets have been dis­ cussed in Guitian (1994c). It should also be pointed out that the IMF has already been considering initiatives in the directions indicated in the text. Thus, a short­ term financing facility, currency stabilization funds, and emergency financing mechanisms have been or are being discussed in the Executive Board.

©International Monetary Fund. Not for Redistribution 818 MANUEL GUITL.\N

forces. This means the IMF must focus on the role of government as well as on the scope of market discipline. Generally, market forces contribute to the recognition of imbalances and, as such, they provide incentives for the correction of those imbalances. But occasions also arise where market financing flows act instead to delay necessary policy adjustments. The challenge is to design and implement economic policies that will support market forces when they act to favor the correction of imbalances, but also to have policies in place that can contain those forces when they let imbalances prevail. IMF conditionality practices are the institution's contribution to this endeavor. However, market failures are easier to describe in theory than to identify in practice, making efforts to recognize and counter them risky. Market failures are rarely perceived as such by all, so action to cope with them calls for willingness to risk official financial resources in a bet that may be seen by others as unrealistic. Such odds can arise in a country-specific context, but they become critical in a systemic setting. So once again, the challenge for conditionality is to foster policies in member countries that guide, give certainty, and favor market forces; but also to promote and garner support for policies that offset market failures, when these arise. There is an important implication here for lMF surveillance, too. As has often been pointed out, in addition to market failur.es, there are also (and perhaps more frequently) government policy failures . And it may be argued that a market failure defined as an inability or an unwillingness to identify an imbalance as it arises has as a counterpart the policy failure that causes it. Such market failures are often behind the "market over­ shoots" that occur when the imbalance has been allowed to cumulate over time. The other side of this argument is the coexistence of policy failures that in fact represent "policy undershoots," these being phenomena that surveillance is intended to limit, if not eliminate. In sum, IMF conditionality practices, appropriately supplemented by the exercise of surveillance, are an instrument to protect the financial integrity of the institution by the provision of public goods to the mem­ bership. These public goods include the support and guidance of market forces, the compensation of market failures and the containment of moral hazard risks.

VI. Assessment

Possibly the most interesting question about IMF conditionality is whether it has made a difference-how effective and how efficient it has been. In practical terms, questions of this nature have led to concrete

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FUTURE 819 examinations of the experience of countries that have undertaken policy programs supported by IMF conditional resources. The nature of the subject is such, though, that it would be futile to expect universal or categorical answers to those questions from the analyses of those expe­ riences, even if there were agreement about the appropriate methodol­ ogy for dealing with them.38 As the purpose of this paper has been to discuss IMF conditionality practices. rather than to examine how they have been applied in the specific context of member countries, such an assessment will not be undertaken here. 39 Instead, an attempt will be made to evaluate the role that conditionality has played as an instrument to protect the code of conduct and to foster its observance, that is, its contribution to the functioning of the interna­ tional monetary system. The assessment will basically be normative and reflects a particular interpretation of events in the international economy over the past five decades. As conditionality is not independent of the nature of the international economic regime in effect, the assessment will be divided according to the experience during the three broad periods discussed in the paper.

Rules and Conditionality

During the Bretton Woods period, the international monetary regime was the fixed (but adjustable) par value system. The commitment to currency parities, together with the obligation of countries to establish current account convertibility for their currencies, represented both a standard and a constraint for their economic policies. On the one hand, appropriate policies were those compatible with both the par value and an open current account in the balance of payments. On the other hand, policies that conflicted with these commitments would soon run against a balance of payments constraint.'�0

3s An early treatment of methodological questions can be found in Guitian what (1981), where three possible standards of comparison were discussed: is relative to what was; what compared to what should be; and what is versus what would or might have beenis. 39There is too ample a body of literature dealing with conditionality that focuses on the subject of country-specific assessments to pay justice to it in this paper. A selected sample would include Schadler 1995), Schadler and others (1995), Santaella (1995), Polak (1991) and the references( it includes, Spraos (1986), Goldstein (1986), Finch (1989), Khan (1990), and Williamson (1983) ; for a selected list of earlier references, see Guitian ( 1981 ). "0The argument in the text is cast in terms of prospective external deficits, as it is in the context of such deficits that the issue of conditionality and IMFfinancial support arises. But the reasoning i� general. In the absence of restrictions,

©International Monetary Fund. Not for Redistribution 820 MANUEL GUITIAN

The role of conditionality in this setting was relatively straightforward. Its aim was to support either policies that would require neither par value changes nor current account restrictions (thus fostering the observance of the rules), or policies that, requiring a par value adjustment. would undertake it and ensure its sustainability at the new parity level without resort to current restrictions (thus exercising discretion, as permitted by the regime). The par value system was to a large measure self-enforcing and there­ fore the use of conditionality was relatively unobtrusive. As has been noted elsewhere, during the Bretton Woods period, the exercise of conditionality was "episodic in character" and the balance of payments problems that it addressed were quickly resolved.41 From this stand­ point, IMF conditionality practices clearly made a difference by con­ tributing to the rapid adjustment of external imbalances. To this should be added the contribution that conditionality made to the understand­ ing of the transmission mechanism between policy instruments and policy objectives,42 and the strong impulse that conditionality practices gave to the quantification of analytical relationships as the basis for policy implementation. This was a period in which the IMF membership, by its agreement to accept and observe common rules, demonstrated readiness to contain the scope of national policy autonomy for the benefit of the system at large. To the extent that, as a result, the environment proved to be more stable than otherwise, the limitations derived from the cession of autonomy were more apparent than real. After all, the basic argument in favor of political autonomy is to protect the national economy from external disturbances. If the latter occur only rarely, the advantages of autonomy diminish.

primary responsibility for policy correction falls on deficit countries; that correc­ tion, through its impact on price and quantity variables in the system, will restore balance to it. The surplus countries' responsibility is to allow those changes to operate within their economies; see GUttian (1992b). 41 See Polak (1991), p. 2, for an elaboration of this argument. Polak attributes the quick resolution of external imbalances, in part, to the rapid expansion of the world economy. But this, in turn,can be attributed to the existence of transparent rules and the instruments available to ensure their observance, key among which was IMF conditionality . 42 Reference should be made here to the institution ·s contribution to balance of payments analysis and its linkages with monetary and other policies. See. for example, the collection of articles in lMF (1977).

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT. FUTURE 821

Discretion and Conditionality

The international monetary regime, by adopting the principle of flexibility in exchange rate arrangements following the abandonment of the Bretton Woods framework, established discretion as the key concept in monetary relationships between countries. As such, the regime ceased being virtually self-enforceable and thus called for an agent of en­ forcement of international order. This was to be the IMF's mandate, a mandate that, rather than being to exercise discretion under agreed rules, became one of enforcing certain rules in a discretionary environment. As the rules became less transparent in terms of members' commitments and obligations, their enforcement called for a substantial measure of judg­ ment on the part of the institutionY The exercise of conditionality in a setting where members exercise discretion and where their obligations are therefore a matter for collec­ tive judgment and agreement became increasingly, and predictably, com­ plex. And the evolution of the international economy in the decades following the Bretton Woods period added to that complexity.44 Never­ theless, a good case can be made that IMF conditionality practices made a difference during this period. By their continuing adaptation to chang­ ing world and country circumstances, they assisted in the correction of external imbalances owing to oil price rises (and declines), in the ad­ justment of imbalances that reflected structural factors and required longer periods of corrective policy implementation, and in the effective resolution of the systemic threats posed by the debt crisis. In the process, conditionality practices helped make operational the link between macroeconomic and microeconomic management by stress­ ing its importance for balance of payments adjustment. They also brought to the forefront the importance of adequate financial support for sound adjustment efforts. And they brought home the message that exchange rate flexibility was no panacea for economic problems. Over the Bretton Woods period, the emphasis was on the preservation of national policy autonomy, an objective that was, in principle, made possible by exchange rate flexibility. On the other hand, this was the time

43The extent to which the new regime called for judgment to be exercised will be evident when reading, for example, Article IV on the obligations of members retarding exchange arrangements; see IMF (1993). In this context. the relevant question has been posed by Paul Vo lcker when he asked whether the turbulence and subpar performance that have char­ acterized the world economy since the 1970s are "related to, and aggravated by, the breakdown of the d isciplines implied by the Bretton Woods monetary system ...." See Vo lcker and Gyo hten ( 1992), p. 291.

©International Monetary Fund. Not for Redistribution 822 MANUEL GUITIAN when capital flows began to become a predominant force in the inter­ national economic environment. And IMF conditionality practices con­ tinued to seek openness in external current accounts and markets. These two factors together set limits to the scope of national policy autonomy. Thus, while during Bretton Woods there was little, if any, loss of autonomy, in the period that followed it, little, if any, gain was made on that front.

Markets and Conditionality

The presence of global markets and their impact on country economic poUcies and performance arecentral features of the world economy now and in the period ahead. The difference that IMF conditionality will make in these circumstances is, of course, yet to be seen. But a few consider­ ations merit attention even at this early stage. The wide scope given to market forces by most governments in both the national and international domains have broadened and strengthened the impact of market discipline on their own policies and economies. IMF conditionality practices and IMF surveillance are also instruments of policy discipline. Their effectiveness will depend on their ability to underpin market forces wben they point to needed adjustments. But it will also be contingent on their skill in guiding market forces away from any tendency they exhibit to delay policy correction. From these perspectives, effective surveillance can go a long way toward averting the "policy undershoots" that often are behind market misper­ ceptioos. And effective conditionality , in turn, can serve to defuse the "market overshoots" that can occur when inadequate policies are allowed to persist for long. Thus, although in a certain sense all instruments of policy discipline-market forces, conditionality, and surveillance-can be seen as competing, they can also complement one another. A good test of whether the IMF and its conditionality practices will make a difference in today's and tomorrow's world economic setting will be their ability to underpin market discipline and to prevent market forces from underwriting inappropriate policies. And an essential message to bear in mind is that the limits on national policy autonomy that have prevailed in the past-during periods of rules or of discretion-will likely be tightened severely in a setting of globalized market forces.

©International Monetary Fund. Not for Redistribution CONDITIONAUTY: PAST, PRESENT. FUTURE 823

ANNEX 1

Guidelines on Conditionalitl5

The Executive Board agrees to the text of the guidelines on conditionality for the use of the Fund's resources and for stand-by arrangements as set forth [below] . Decision No. 6056-(79/38) March 2, 1979

Use of Fund's General Resources and Stand-By Arrangements

1. Members should be encouraged to adopt corrective measures, which could be supported by use of the Fund's general resources in accordance with the Fund's policies, at an early stage of their balance of payments difficulties or as a precau­ tion against the emergence of such difficulties. The Article IV consultations are among the occasions on which the Fund would be able to discuss with members adjustment programs, including corrective measures. that would enable the Fund to approve a stand-by arrangement.

2. The normal period for a stand-by arrangement will be one year. If, however, a longer period is requested by a member and considered necessary by the Fund to enable the member to implement its adjustment program successfully, the stand-by arrangement may extend beyond the period of one year. This period in appropriate cases may extend up to but not beyond three years.

3. Stand-by arrangements are not international agreements and therefore lan­ guage having a contractual connotation will be avoided in stand-by arrangements and letters of intent.

4. ln helping members to devise adjustment programs, the Fund will pay due regard to the domestic social and political objectives, the economic priorities, and the circumstances of members, including the causes of their balance of payments problems.

5. Appropriate consultation clauses will be incorporated in all stand-by arrange­ ments. Such clauses will include provision for consultation from time to time during the whole period in which the member has outstanding purchases in the upper credit tranches. This provision will apply whether the outstanding pur­ chases were made under a stand-by arrangement or in other transactions in the upper credit tranches.

•ssource: IMF (1994).

©International Monetary Fund. Not for Redistribution 824 MANUEL GU!TIAN

6. Phasing and performance clauses will be omitted in stand-by arrangements that do not go beyond the first credit tranche. They will be included in all other stand-by arrangements but these clauses will be applicable only to purchases beyond the first credit tranche.

7. The Managing Director will recommend that the Executive Board approve a member's request for the use of the Fund's general resources in the credit tranches when it is his judgment that the program is consistent with the Fund's provisions and policies and that it will be carried out. A member may be expected to adopt some corrective measures before a stand-by arrangement is approved by the Fund, but onJy if necessary to enable the member to adopt and carry out a program consistent with the Fund's provisions and policies. In these cases the Managing Director will keep Executive Directors informed in an appropriate manner of the progress of discussions with the member.

8. The Managing Director will ensure adequate coordination in the application of policies relating to the use of the Fund's general resources with a view to maintaining the nondiscriminatory treatment of members.

9. The number and content of performance criteria may vary because of the diversity of problems and institutional arrangements of members. Performance criteria will be limited to those that are necessary to evaluate implementation of the program with a view to ensuring the achievement of its objectives. Perfor­ mance criteria will normally be confined to (i) macroeconomic variables, and (ii) those necessary to implement specific provisions of the Articles or policies adopted under them. Performance criteria may relate to other variables only in exceptional cases when they are essential for the effectiveness of the member's program because of their macroeconomic impact.

10. In programs extending beyond one year, or in circumstances where a member is unable to establish in advance one or more performance criteria for all or part of the program period, provision will be made for a review in order to reach the necessary understandings with the member for the remaining period. In addition, in those exceptional cases in which an essential feature of a program cannot be formulated as a performance criterion at the beginning of a program year be­ cause of substantial uncertainties concerning major economic trends, provision will be made for a review by the Fund to evaluate the current macroeconomic policies of the member. and to reach new understandings if necessary. In these exceptional cases the Managing Director will inform Executive Directors in an appropriate manner of the subject matter of a review.

11. The staff will prepare an analysis and assessment of the performance under programs supported by use of the Fund's general resources in the credit tranches in connection with Article TV consultations and as appropriate in connection with further requests for use of the Fund's resources.

©International Monetary Fund. Not for Redistribution CONDrTIONALITY: PAST, PRESENT, FUTURE 825

12. The staff will from time to time prepare, for review by the Executive Board, studies of programs supported by stand-by arrangements in order to evaluate and compare the appropriateness of the programs, the effectiveness of the policy instruments, the observance of the programs, and the results achieved. Such reviews will enable the Executive Board to determine when it may be appropriate to have the next comprehensive review of conditionality.

©International Monetary Fund. Not for Redistribution 826 MANU EL GUITI..\N

ANNEX II

IMF Financial Facilities46

The IMP's financial resources are made available to its members under a variety of facilities and policies, depending on the circumstances, especially the nature of the macroeconomic and structural problems to be addressed. The IMF makes resources available under both its general resources and its concessional financing facilities. Member countries use the general resources of the IMF by making a purchase (drawing)47 of other members' currencies or SDRs with an equivalent amount of their own currencies. The IMF levies charges on these drawings and requires that members repurchase (repay) their own curren­ cies from the IMF with other members' currencies or SDRs over a specified time. Concessional financing under the structural adjustment and enhanced structural adjustment facilities is made available in the form of loans.

Regular Facilities

Reserve Tranche A member has a reserve tranche position to the extent that its quota exceeds the IMF's holdings of its currency in the General Resources Account, excluding holdings arising out of drawings made by the member under all policies govern­ ing the use of the IMF's general resources. A member may draw up to the full amount of its reserve tranche position at any time, subject only to the requirement of a balance of payments need. A reserve tranche drawing does not constitute a use of IMF credit and is not subject to charges or to an expectation or obligation to repay.

Credit Tranches IMF credit is made available in tranches (or segments) of 25 percent of a member country's quota. For drawings in the first credit tranche, members must demonstrate reasonable efforts to overcome their balance of payments diffi­ culties. First credit tranche drawings are not phased nor are they subject to performance criteria. Upper credit tranche drawings are made in installments and are released, provided policy implementation is in line with an agreed program. Such draw­ ings are normally associated with stand-by or extended Fund facility arrange­ ments. These arrangements typically aim at overcoming balance of payments difficulties as well as supporting structural policy reforms where appropriate. Performance criteria and periodic general program reviews are used to assess policy implementation.

""Excerpted from 1M F (September 1995). 47Table Al in the AppendixSurvey hsts the various arrangements and resource com­ mitments of the IMF undertaken since its inception.

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST. PRESENT. FUTURE 827

Stand-By Arrangements Stand-by arrangements give members the right to draw up to a specified amount during a prescribed period. Drawings are phased on a quarterly basis, and their release is conditional on meeting performance criteria and the comple­ tion of periodic program reviews. Performance criteria generally cover credit policy, government or public sector borrowing requirements. trade and payments restrictions, foreign borrowing. and reserve levels. These criteria allow both the member and the IMF to assess progress and may signal the need for further corrective policies. If performance criteria are not met, members can make further drawings only if the IMF and the member country reach understandings on the resumption of drawings. Stand-by arrangements typically cover a 12-18 month period (although they can extend up to 3 years), and repayments are to be made within 3V. to 5 years of each drawing.

Extended Fund Facility (EFF) The extended Fund facility provides assistance to member countries for longer periods and in larger amounts than under credit tranche policies. Extended arrangements generally run for three years (and can be extended for a fourth year). They are aimed at overcoming balance of payments difficulties stemming largely from structural problems and requiring a longer period of adjustment. A member requesting an extended arrangement is expected to present a program outlining the objectives and policies for the whole period of the arrangement, and to present a detailed statement each year of the policies and measures it will follow in the next 12-month period. The phasing and performance criteria are comparable to those of stand-by arrangements, although phasing on a semiannual basis is possible. Countries using extended Fund faci lity resources must repay the currencies they have drawn within 4V2 to 10 years of the drawing.

Special Facilities

Systemic Transformation Facility (STF) This temporary facility was created in April 1993 to respond to the needs of economies in transition . It provided financial assistance to members experiencing balance of payments difficulties as a result of a shift from significant reliance on trading at nonmarket prices to multilateral market-based trade. Access was limited to not more than 50 percent of quota, which could be made in two drawings and was in addition to financing obtained under other IMF facilities. Since May 1, 1995, first drawings under the STF can no longer be made, but members that have had a first drawing approved by then are eligible for a second drawing by ef'\d-December 1995.

Compensatory and Contingency Financing Fa cility (CCFF) The compensatory component of this special facility provides members that demonstrate a balance of payments need with resources to help compensate for

©International Monetary Fund. Not for Redistribution 828 MANUEL GUITIAN temporary shortfalls in export earnings and services receipts and/or temporary excesses in cereal import costs that arise from events largely beyond their control. The contingency element helps members with IMF arrangements maintain the momentum of their reform efforts when faced wi th unforeseen adverse external shocks. Financing is provided to cover part of the net effect of unfavorable deviations in highly volatile and easily identifiable key variables affecting the member's current account. The variables could include main export or import prices and international interest rates. Workers' remittances and tourism receipts may also be covered if they represent a significant element in the member's current account. Deviations are measured in relation to baseline projections, which are established at the start of an arrangement. The contingency mechanism is triggered once net deviations exceed a threshold level and it is clear they will not be offset by positive movements of other key variables. Favorable devia­ tions-such as an unexpected increase in export prices-may trigger a symmetry provision, under which a higher reserve target, a reduced drawing. or an earlier repayment would be required.

Buffer Stock Financing Facility Under this facility, the IMF provides resources to help finance members' contributions to approved international buffer stocks, if the member demon­ strates a balance of payments need. Drawings must be repaid within 3V.to 5 years. No drawings have been made under this facility for the past ten years, and no credits under the facility are currently outstanding.

Emergency Assistance

In the context of balance of payments assistance under its credit tranche policies, the IMF provides emergency assistance, allowing members to make drawings to meet balance of payments needs arising from sudden and unforesee­ able natural disasters. These drawings do not entail performance criteria or a phasing of disbursements, although they often precede an arrangement with the IMF under its regular facilities.

Concessional Facilities

Structural Adjustment Facility (SAF) Established in 1986, arrangements under this facility provided low-income member countries with loans on concessional terms in support of medium-term macroeconomic adjustment policies and structural reforms. In November 1993, the IMF's Executive Board agreed that no new commitments would be made under the structural adjustment facility.

Enhanced Structural Adjustment Fa cility (ESAF) Established by the Executive Board in 1987, the ESAF was extended and enlarged in Fe bruary 1994. With the decision to halt new commitments under the SAF and to expand significantly the resources available under the ESAF Trust

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 829

Fund (the funding source for ESAF arrangements), ESAF arrangements became the principal means by which the IMF provides financial support, in the form of loans and on concessional terms, to low-income member countries facing pro­ tracted balance of payments problems. ESAF resources are intended to support strong medium-term structural adjust­ ment programs. Eligible members seeking the use of ESAF resources must develop, with the assistance of the staffs of the fMF and the World Bank, a policy framework for a three-year adjustment program. A policy framework paper describes the authorities' economic objectives, macroeconomic and structural policies, priorities, and the measures they intend to adopt during the three-year period. It assesses external financing needs and major sources of financing, and serves as a means of strengthening Bank-IMF collaboration and of attracting externalfinancial and technical assistance in support of the adjustment program. The paper is updated annually. Adjustment measures under the ESAF are expected to help achieve a substan­ tial strengthening of the balance of payments position and foster growth during the three-year period. ESAF arrangements are funded by ESAF Trust resources, which are derived from loans and grants as well as from transfers from the Special Disbursement Account. Monitoring under ESAF arrangements is conducted through quarterly financial and structural benchmarks. In addition, semiannual performance criteria are set for key quantitative and structural targets. ESAF loans are disbursed semiannually, initially upon approval of an annual arrange­ ment and subsequently based on the observance of performance criteria and after completion of the midterm review. ESAF loans are repaid in ten semiannual installments, beginning 5\12 years and ending 10 years after the date of each disbursement. The interest rate on ESAF loans is 0.5 percent a year. The enlargement and extension of the ESAF Trust became effective on Feb­ ruary 1994, and is being financed by a broad cross-section of the IMF's 23, membership. To date, commitments have been received from more than 40 countries, about half of which are developing countries. The target amount of lending for the enlargement is SDR 5 billion, which, together with the SDR 5.1 billion available before the enlargement, would expand the total lending capacity to SDR 10.1 billion. The commitment period extends to December 1996, with disbursements to be made through the end of 1999.

©International Monetary Fund. Not for Redistribution 00 Vl 0 APPENDIX

Arrangements Approved During Financial Years Ended April Table Al. 30, 1953-95 1 Amounts Committed Under Arrangements< Number of Arrangementsb (In millions of SDRs) Financial Year Stand-by EFF SAF ESAF STF Total Stand-by EFF SAF ESAF STF Total 19S3 2 2 ss ss 19S4 2 2 63 63 :: 19SS 2 2 40 )> 40 z 19S6 2 2 48 c 48 m l9S7 9 9 1,162 1,162 r 11 1,044 0 19S8 11 1,044 c 1959 15 1S 1,0S7 1,057 3 1960 14 14 364 364 >· 1961 1S 1S 460 z 1962 24 24 1,633 1,633460 1963 19 19 1,S31 l,S31 1964 19 19 2,160 2,160 196S 24 24 2,159 2,159 1966 24 24 S7S S7S 1967 25 25 591 591 1968 32 32 2,352 2.3S2 1969 26 26 541 541 1970 23 23 2,381 2,381 1971 18 18 502 502 1972 13 13 314 314

©International Monetary Fund. Not for Redistribution 1973 13 13 322 322 1974 15 15 1,394 1,394 1975 14 14 390 390 1976 18 2 20 1,188 284 1,472 1977 19 1 20 4,680 518 5,198 1978 18 18 1,285 1,285 1979 14 4 18 508 1,093 1,600 1980 24 4 28 2,479 797 3,277 (') 1981 21 11 32 5,198 5,221 10,419 0 1982 19 5 24 3,106 7,908 11,014 0z 1983 27 4 31 5,450 8,671 14,121 3 27 4,287 95 4,382 0 1984 25 2 z 1985 24 24 3,218 3,218 > 1986 18 19 2.123 825 2,948 r 1987 10 32 4,118 358 4,476 22 � 1988 14 1 15 30 1,702 245 670 2,617 "0 >(/) 1989 12 1 4 7 24 2,956 207 427 955 4,545 :1 1990 16 3 4 26 3,249 7,627 37 415 11,328 "0 3 ;o 3 20 2,786 2,338 15 5,593 ['1'1 1991 13 2 2 454 (/) 1992 21 2 1 5 29 5,587 2,493 2 743 8,826 ['1'1 1993 11 3 1 8 23 1,971 1,242 49 527 3,789 z 1994 18 2 1 7 18 28 1,381 779 27 1,170 2,725 6,082 ;-1 ,., 1995 17 3 11 14 31 13,055 2,335 1,197 1 , 123 17,710 c 2 732 51 37 32 865 91,466 42,678 1,585 5,461 3,848 145,038 ;o Total 45 ['1'1 Sources: IMF, Annual Report (1995); IMF Memorandum; IMF staff calculations. • For abbreviations of facilities, see Annex II. hFor STF, number of drawings; most countries have drawn twice ·

00 \;.) ......

©International Monetary Fund. Not for Redistribution 832 MANU EL GU!TIAN

REFERENCES

Bruno, Michael, "Development Issues in a Changing World: New Lessons, Old Debates, Open Questions,'' Proceedings of the World Bank Annual Con­ fe rence on Development Economics (Washington: World Bank, 1995), 1994 pp. 9-19.

Calvo, Guillermo A., and others, eds . . Debt, Stabilization and Development: Essays in Memory of Car/os Diaz-Alejandro (New York: Basil Blackwell, 1989). Camdessus, Michel, Closing Remarks at Plenary Session of the IMF/World Bank Annual Meetings, October 6. 1994, Madrid. Reproduced in IMF Survey (October 17, 1994), p. 328. Cline, William R., InternationalDebt: Systemic Risk and Policy Response (Wash­ ington: Institute for International Economics, 1984).

---, International Debt Reexamined (Washington: Institute for International Economics, 1995). Corbo, Vittorio, Morris Goldstein, and Mohsin Khan, eds., Growth-Oriented Adjustment Programs (Washington: International Monetary Fund and World Bank, 1987). de Larosiere, Jacques, "The Debt Situation,'' Remarks before the Bankers' Association for Foreign Trade, Phoenix, Arizona, May 16, 1986; reprinted in IMF Survey , Vol. 15 (June 2, 1986), pp. 164-67.

---, "Progress on the International Debt Strategy," Finance and Develop­ ment, Vol. 24 (March 1987), pp. 10-11. de Vries, Margaret Garritsen, The internationalMonetary Fund, The 1966-1971.· System Under Stress (Washington: International Monetary Fund, 1976). ---, The International Monetary Fund, Cooperation on Trial 1972-1978: (Washington: International Monetary Fund, 1985). Dornbusch, Rudiger, John H. Makin, and David Zlower, eds., Alternative Solu­ tions to Developing-Country Debt Problems (Washington: American Enter­ prise Institute for Public Policy Research, 1989). Duisenberg, W.F., and A. Szasz, "The Monetary Character of the IMF," in International Financial Policy.· Essays in Honor of Jacques J. Polak, ed. by Jacob A. Frenkel and Morris Goldstein (Washington: International Mone­ tary Fund and De Nederlandsche Bank, 1991), pp. 254-66. Erb, Richard D., "Current Developments in the Debt Crisis,'' in Interna­ tional Finance and Financial Policy, ed. by Hans R. Stoll (New York: Quorum Books, 1990). Finch, C. David, The IMF: The Record and the Prospect, Princeton Essays in International Finance No. 175 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, September 1989). Frenkel, Jacob A., Michael P. Dooley, and Peter Wickham, eds., Analytical Issues in Debt (Washington: International Monetary Fund, 1989). Frenkel, Jacob A., and Morris Goldstein, eds., International Financial Policy:

©International Monetary Fund. Not for Redistribution CONDITIONALITY: PAST, PRESENT, FUTURE 833

Essays in Honor of Jacques J. Polak (Washington: International Monetary Fund and De Nederlandsche Bank, 1991). Gold, Joseph, The Stand-By Arrangements of the InternationalMonetary Fund: A Commentary on Their Formal, Legal, and Financial Aspects (Washington: International Monetary Fund, 1970).

--, Conditionaliry, IMF Pamphlet Series, No. 31 (Washington: International Monetary Fund, 1979). Goldstein, Morris, The Global Effects of Fund-Supported Adjustment Programs, IMF Occasional Paper No. 42 (Washington: International Monetary Fund, 1986). Guitian, Manuel, "Credit Versus Money as an Instrument of Control," Staff Papers , International Monetary Fund, Vol. 20 (November 1973), pp. 785-800; reprinted in International Monetary Fund (1977). --, Fund Conditionality: Evolution of Principles and Practices , IMF Pam­ phlet Series No. 38 (Washington : International Monetary Fund, 1981).

-- (1992a), "Remarks on the Debt Crisis," in Changing Fo rtunes: The Wo rld's Money and the Threat eo American Leadership , by Paul A. Vo lcker and Toyoo Gyohten (New York: Times Books, 1992), pp. 318-22. -- (1992b), Rules and Discretion in International Economic Policy , IMF Occasional Paper No. 97 (Washington: International Monetary Fund, 1992).

-- (1992c), The Unique Nature of the Responsibilities of che International Monetary Fund, IMF Pamphlet Series, No. 46 (Washington: International Monetary Fund, 1992). --, "The Issue of Capital Account Convertibility: A Gap Between Norms and Reality." Paper presented at a Seminar on Currency Convertibility sponsored by the Arab Monetary Fund and the IMF Institute (Marrakech, Morocco: IMF; Arab Monetary Fund, December 1993). -- (1994a), "The IMF as a Monetary Institution: The Challenge Ahead," Finance and Development, Vol. 31 (September 1994), pp. 38-41. -- (1994b) , "The Role of Monetary Policy in IMF Programs," in A Frame­ work for Monetary Stability, ed. by J. On no de Beaufort Wijnholds and others (Dordrecht, The Netherlands: Kluwer Academic Publishers, 1994), pp. 185-202.

---, "Scope of Government and Limits of Economic Policy," in Macro­ economic Dimensions of Public Finance, ed. by Mario Blejer and Teresa Ter-Minassian (London: Rutledge, forthcoming 1996). Gwin, Catherine, and others, The International Monetary Fund in a Multipolar World: Pulling Together (New Brunswick, New Jersey: Transaction Books, 1989). Horsefield, Keith, ed., The International Monetary Fund, /945-1965: Twenty J. Years of International Monetary Cooperation (Washington: International Monetary Fund, 1969). Husain, Ish rat, and Ishac Diwan, eds., Dealing with the Debt Crisis (Washington: World Bank, 1989).

International Monetary Fund, The Monetary Approach to the Balance of Pay­ ments: A Collection of Research Papers (Washington: IMF, 1977).

©International Monetary Fund. Not for Redistribution 834 MANUEL GUITL.\N

---, Technical Assistanceand Training Services of the International Monetary Fu nd, IMF Pamphlet Series, No. 43 (Washington: International Monetary Fund, reprinted in May 1991).

--, Articles of Agreement (Washington: IMF, April 1993).

---, Selected Decisions and Selected Documents of the Jmemational Monetary Fund, Nineteenth Edition (Washington: IMF, June 30, 1994). ---, Capital Account Convertibility: Review of Experience and implications for /M F Policies, IMF Occasional Paper No. 131 (Washington: International Monetary Fund, 1995). James, Harold, internationalMonetary Cooperation Since Bretton Woods (Wash­ ington: International Monetary Fund; and New York; Oxford: Oxford Uni­ versity Press, 1996). Khan, Mohsin S., '·The Macroeconomic Effects of Fund-Supported Adjustment Programs," Staff Papers , International Monetary Fund, Vol. 37 (June 1990), pp. 195-231. Mehran, Hassanali, ed., ExternalDebt Management (Washington: International Monetary Fund, 1985). Mohammed, Azizali, "Recent Evolution of Fund Conditionality," in Interna­ tional Financial Policy: Essays in Honor of Jacques J. Polak, ed. by Jacob A. Frenkel and Morris Goldstein (Washington: International Monetary Fund and De Nederlandsche Bank, 1991), pp. 244-53. Mookerjee, Subimal, "Policies on the Use of Fund Resources," Staff Papers, International Monetary Fund, Vol. 13 (November 1966), pp. 421-40. Nowzad, Bahram, The IMF and Its Critics, Princeton Essays in International Finance, No. 146 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, December 1981). Ortiz, Guillermo, "The IMF and the Debt Strategy,'' in The InternationalMon­ etary Fund in a Multipolar World: PullingTo gether, by Catherine Gwin and others (New Brunswick, New Jersey: Transaction Books, 1989). Polak, Jacques J., The Changing Nature of 1M F Conditionality, Prince ton Essays in International Finance, No. 184 (Princeton, New Jersey: Princeton Univer­ sity, Department of Economics, International Finance Section, September 1991). Robichek, E. Waiter, "Financial Programming: Stand-By Arrangements and Stabilization Programs" (unpublished; Washington: International Monetary Fund, January 6, 1971).

---, "Financial Programming as Practiced by the IMF" (unpublished; Wash­ ington: World Bank, 1985). Santaella, JuJio A., "Four Decades of Fund Arrangements: Macroeconomic Stylized Facts Before the Adjustment Programs," IMF Wo rking Paper 95/74 (Washington: International Monetary Fund, July 1995). Schadler, Susan, and others, IMF Conditionality: Experience Under Stand-By and Extended Arrangements-Part 1: Key Issues and Findings, IMF Occa­ sional Paper No. 128 (Washington: International Monetary Fund, 1995). Schadler, Susan. ed., /MF Conditionaliry: Experience under Stand-By and Ex­ tended Arrangements-Part 1/: Background Papers, IMF Occasional Paper No. 129 (Washington: International Monetary Fund, 1995).

©International Monetary Fund. Not for Redistribution CONDITIONAUTY: PAST, PRESENT. FUTURE 835

Shultz, George P., "Economics in Action: Ideas, Institutions,'Policies," Ameri­ can Economic Review , Papers and Proceedings , Vol. 85 (May 1995), pp. 1-8. Spraos, John, /M F Conditionality: Ineffectual, Inefficient, Mistargeted, Princeton Essays in International Finance, No. 166 (Princeton, New Jersey: Prince­ ton University, Department of Economics, International Finance Section, December 1986). Stall, Hans R., ed., international Finance and Financial Policy (New York: Quorum Books, 1990).

Summers, Lawrence H . . and Vinod Thomas, "Recent Lessons of Development," World Bank Research Observer, Vol. 8 (Washington: World Bank, July 1993), pp. 241-54. Tanzi, Vito, "Fiscal Policy, Growth, and Design of Stabilization Programs," in External Debt, Savings, and Growth in Latin America, ed. by Ana Maria Martirena-Mantel (Washington: International Monetary Fund and Instituto Torcuato di Te lla, 1987), pp. 121-41. Volcker, Paul A., and Toyoo Gyohten, Changing Fo rtunes: The World's Money and the Threat to American Leadership (New York: Times Books, 1992). Williamson, John, ed., IMF Conditionality (Washington: Institute for Interna­ tional Economics, 1983).

---, "What Washington Means by Policy Reform," in Latin American Adjust­ ment: How Much Has Happened? ed. by John Williamson (Washington: Institute for International Economics, 1990), pp. 7-20.

©International Monetary Fund. Not for Redistribution IMF Staff Papus Vol. 42. No. 4 (December 1995) International Monetary Fund � 1995

Wage Structures in the Tr ansition of the Czech Economy

ROBERT J. FLANAGAN*

From the perspective of market economies, central planning produced distinct distortions in the wage structures ofsocialist countries. This paper examines the extent to which wage structures in the state and private sectors have adjusted to remove such distortions during the econornic transition, using micro data from the Czech Republic. There is strong evidence that Czech wage structures are moving toward pauems observed in market economies, and the change is being led by developments in the private sector. At the same time, the establishment of collective bargaining does not appear to be introducing countervailing distortions imo Czech wage structures. [JEL JO, J3, J4, JS. P2, PSl

N THE EARLY POSTWAR PERIOD, the central planning systems of eastern Iand central Europe pursued higher growth by raising labor force participation and increasing capital investment. Over time, however, this contribution to growth was countered by systemic failures to allocate workers to the sectors in which their potential productivity was highest and to elicit high levels of worker performance within organizations. The resulting decline in total factor productivity growth, which figured promi­ nently in the collapse of central planning systems. was largely caused by the failure of the system to devise adequate incentives to solve the labor allocation and performance problems.

• Robert J. Flanagan is the Konosuke Matsushita Professor of International Labor Economics at the Graduate School of Business, Stanford University. Parts of this paper were written while he was a visiting scholar at the Research Depart­ ment of the IMF in October 1994. The author is grateful to Jiri Vecernik for generously providing the fi les of the Survey of Economic Expectations and Attitudes, to David Coe and Vladislav Flek for comments on an earlier draft, and to the Stanford Graduate School of Business for research support. 836

©International Monetary Fund. Not for Redistribution WAGE STRUCfURES TN THE CZECH ECONOMY 837

There is a strong presumption among economists that the transition from central planning to more market-oriented economics should elim­ inate the inefficiencies associated with central planning. Given the initial pattern of distortions, this process should produce radical changes in relative wages and an increase in overall wage inequality. Nevertheless, there is no clear understanding of how or how rapidly the required wage structures and enterprise wages policies will emerge. Nor has the possibility that the transition process might produce coun­ tervailing distortions-common in market economies but absent under central planning-received much attention. Foremost among the poten­ tial sources of wage structure distortions accompanying the economic transitions is the reconstitution of tabor unions into genuine collective bargaining organizations. However. the adjustment of wage structures in transition economies may also be thwarted by the wage rules applied through incomes policies-a clear case of short-run stabilization policies conflicting with the fundamental objectives of the transitions. The conse­ quence of reduced relative wage flexibility from any source is an in­ creased reliance on quantity adjustments, notably unemployment and tabor force withdrawal, in transition economies. This paper examines the extent to which wage distortions under central planning have been reversed during the transition and the extent to which countervailing distortions have been introduced, using micro data from the Czech Republic. The data arc from employment and tabor force surveys conducted between June 1988 and November 1994. In addition to providing information on wage. schooling, and demographic charac­ teristics of respondents, the most recent surveys include information on union status, presence of foreign investment. and other institutional features of tabor markets in the Czech Republic. Section I proceeds from a review of the tabor market setting under central planning. which pro­ vided the point of departure for the transition. to an analysis of the recent evolution of the wage structure in the Czech Republic. A defining feature of a transition is the growth of a private sector. and Section 11 explores how this development contributes to the observed changes in wage struc­ tures. The data permit an assessment of developments in the two major components of the private sector-former state enterprises that have been privatized and new private firms. Section III addresses differences in company wage policies between the state and private sectors and between two components of the private sector-privatized state firms and newly created private firms. The influ­ ence of foreign investment on wage policies is also examined. Section IV of the paper examines the pattern of union representation and the extent to which the development of collective bargaining has influenced wage

©International Monetary Fund. Not for Redistribution 838 ROBERT J. FLANAGAN structures. The final section discusses the policy implications of the findings.

I. Evolution of the Wage Structure

This section examines the evolution of the wage structure in the Czech Republic between 1988, a pretransition year in which virtuaJJy all official economic activity was in the state sector, and late 1994, when the rapidly emerging private sector and privatized state enterprises accounted for about half of employment. The point of departure is the distinctive wage structure achieved under central planning. In the early postwar period, central planners wished to encourage the development of goods-produc­ ing sectors of the economy (particularly heavy industry) and discourage production in services and most intellectual activities outside the hard sciences. In establishing a national wage structure-the so-called "tariff" wage scale for each job category-in the early postwar period, central planners gave some weight to education, the unpleasantness of the job, and other factors familiar to Adam Smith, but also altered relative wages to reaJlocate tabor in accord with their objectives for industrial develop­ ment. By the mid-1950s east European economies experienced large relative wage increases in industry and construction and relative wage decreases in trade and services (Adam (1984)). The devaluation by central planners of "mental" work relative to "physical" work also influenced the schooling and training of the labor force. Vocational training was emphasized over advanced educational attainment and the limited university slots were often allocated on the basis of political affiliation and activity. From the perspective of market economies, east European economies entered their economic transitions with overinvestments in vocational-apprenticeship training and under­ investments in university education. As a result, the average returns to schooling in eastern Europe under central planning were lower than returns to schooling in most market economies (Fianagan (1994b)). The official tariff wage structures, with the preferences that they incor­ porated, have been abandoned during the transition process, leaving relative wages free to adjust subject to new institutional constraints. The rest of the paper examines the nature of these adjustments using data for June 1988 (18 months before the revolution), November 1993, and November 1994 (about four and five years into the transition). The analysis for 1988 uses over 16,700 observations from the microcensus conducted by the Czech Federal Statistical Office. The data for 1993 and 1994 are from the Survey of Economic Expectations and Attitudes con-

©International Monetary Fund. Not for Redistribution WAGE STRUCTURES IN THE CZECH ECONOMY 839

ducted by the Institute of Sociology at the Czech Academy of Sciences. These surveys cover adults 18 to 60 years of age and use two-step quota sampling, with the first step defining the region and size of locality, and the second defining gender, age, and education. The survey responses provide basic economic and personal data as well as the attitudes of respondents to economic and political aspects of the transition. The November 1993 survey includes questions on several aspects of human resource management, and the November 1994 survey includes questions on union membership. The November 1993 (1994) survey interviewed 1,113 (1,307) respondents from the Czech Republic, and data for the 754 (864) employed respondents are analyzed in this paper. To examine the post-transition evolution of the wage structure, these data are analyzed with a human capital earnings function, in which the natural logarithm of the monthly full-time wage (net of social benefits) is regressed on years of schooling and potential experience (age minus school years minus six) variables. The regressions also include a dummy variable for gender. The average rate of return to another year of school (the coefficient on the school years variable) moved from 4.3 percent in mid-1988 to 6.2 percent in late 1994.1 Much of this increase occurred during 1994, as employment began to increase in the Czech Republic for the first time since the beginning of the economic transition. Aggregate estimates may obscure key aspects of the transition process, since new private firms are unlikely to adopt the wage structures and practices developed in state enterprises under central planning. Moreover, the private sector itself is not monolithic. Pay structures and policies in privatized state enterprises are likely to be subject to organizational inertia not encountered by new private firms, and the wage and human resource management policies of foreign firms may differ from those of domestic private firms. Estimated returns to schooling in November 1994 (1993) were 7.3 (6.7) percent in new private employment, 6.7 (5.4) percent in privatized state enterprises, and 6.5 (5.4) percent in ongoing state enterprises. While these figures seem to imply that wage structures in the state sector are responding sluggishly to wage structure develop­ ments in the private sector, the differences in the estimated returns to schooling between the three sectors are not statistically significant. Given the distortions in educational choices discussed above, it is also of interest to examine changing returns to specific levels of educational

attainment. The regression results in Table 1 report the wage premiums

1 The regressions producing these estimates are not reported here, but are available from the author. The standard errors on the estimated returns to schooling for 1988 and 1994 are .095 and .62. re�pectively.

©International Monetary Fund. Not for Redistribution Table l. Evolution of Returns to Schooling (/-statistics in parentheses)

Standard Adjusted error of estimate 1988 CONSTANT VOCAT HS UN/V EXP EXPSQ PRWORK PROWN FEMA LE R2 (1) Total 5.439 0.063 0.!53 0.370 0.036 -0.00070 -0.377 0.44 0.283 = 16,730) (569) (9.87) (22.98) (42.48) (56.83) (55.65) (83.66) (N 1994 (2) Total 8.488 0.180 0.316 0.550 0.017 -0.00032 -0.382 0.25 0.437 = 863) (125) (3.88) (6.54) (9.61) (3.80) (3.39) (12.67) (N (3) State 8.220 0.222 0.422 0.593 0.018 -0.00021 -0.316 0.32 0.354 = 349) (102) (3.63) (6.89) (8.35) (2.79) ( 1.58) (8.14) (N (4) StPriv 8.474 0.034 0.166 0.551 0.023 -0.00039 -0.329 0.30 0.376 (N=238) (61) (0.45) (2.03) (5.80) (2.74) (2.18) (6.68) ;!1 (5) Private 8.679 0.183 0.336 0.634 0.020 -0.00045 -0.437 0.23 0.515 > z (N=241) (58) (1.81) (3.05) (2.25) (2.42) (6.46) > (4.80) C) (6) Total 8.394 0.158 0.319 0.549 0.019 -0.00033 0.140 0.439 -0.359 0.32 0.418 > z = 863) (126) (3.56) (6.90) (9.99) (4.26) (3.67) (3.69) (8.92) (12.41) Sources:(N Czechoslovak Income Survey (Microcensus) for 1988; Survey of Economic Expectations and Attitudes, November 1994, for 1994. Notes: Dependent variable is the natural logarithm of the full-time monthly wage. VOCAT =vocational degree is highest complete schooling. HS = high school degree is highest completed schooling. UN/V= university degree is highest completed schooling. EXP =potential experience (age minus school years minus 6). EXPSQ =square of potential experience. PRWORK =employee in private sector. PROWN =employer in pnvate secor. FEMA LE = female worker.

©International Monetary Fund. Not for Redistribution WAGE STRUCfURES IN THE CZECH ECONOMY 841

relative to primary school completion associated with completion of vocational (VOCAT), high school (HS), or university (UN/V) degrees. The regressions include the standard quadratic specification of potential experience (EXPand EXPSQ) and a dummy variable (FEMA LE) tak­ ing the value 1 if the respondent is a woman and 0 otherwise. Regressions 1 and 2 report the economywide evolution of returns to different levels of school completion between 1988 and 1994. Between mid- 1988 and late 1994, the wage returns to vocational, high school, and university levels of schooling increased relative to primary school education by statistically significant amounts. Regressions 3 through 5 report 1994 returns by three major sectors­ traditional state enterprises (State), recently privatized state enterprises (StPriv), and new private firms created during the transition (Private). (Limited sample size prevents separate regression analyses of private sector wage structures in foreign and domestic firms.) The returns to vocational education are measured most reliably in the state sector and are not economically or statistically significant in privatized enterprises. (One year earlier none of the three major sectors reported a statistically significant effect of vocational education on wages.) While returns to a university degree appeared to increase most rapidly in the private sector in the early stages of the transition, by November 1994 there were no statistically significant differences in the returns between the three major sectors. Under central planning, returns to (potential) experience were also lower than in market economies. During the transition. these returns became even lower-economywide earnings-experience profiles flat­ tened further since 1988. By 1994, ten years of labor force experience produced half the wage effect it would have produced five years earlier. This is consistent with the hypothesis that recent labor market experience is more valuable than experience acquired under central planning. How do wages paid in the private sector compare with wages paid in current and former state enterprises? The November 1994 Survey of Economic Expectations and Attitudes identifies both owners and work­ ers in new private enterprises, as well as workers in privatized state enterprises. After controlling for schooling and potential experience. workers (owners) in new private firms earn 14 (44) percent more than respondents in current or former state enterprises (Table 1, regression 6).2 (No significant wage differential had yet emerged between workers in current and former state enterprises. however.)

1The slope coefficients do not vary with worker status in the private sector. Interacting PROWN with the schooling and experience variables produced no significant interactions.

©International Monetary Fund. Not for Redistribution 842 ROBERT J. FLANAGAN

To summarize, since 1988 average returns to schooling increased, and new private firms appear to have played a leading role in reversing the wage structure distortions that developed under central planning. Re­ turns to university education increased, initially in the private sector but later more broadly. Most recently, there have been some changes in wage structures in the state sector that may reflect the effects of tabor market competition with the private sector (e.g., some increase in the returns to high school and university degrees). There is mounting evidence that both the relative wage adjustments and the leading role of the private sector reported above are not unique to the Czech Republic. Wage inequality and returns to schooling have risen in Poland, led by sharply increasing relative wages for white­ collar workers with a university education (Rutkowski (1994)). Returns to schooling in Hungary have also increased (Commander and others (1993)). Unlike the Czech Republic, both of these countries had signifi­ cant sectors of private activity operating well before the economic tran­ sitions. In contrast, in Romania, where there has been little development of a private sector, there was little change in returns to schooling in the early transition years, and a university degree only provided a 30 percent wage premium relative to high school completion in 1992 (Earle and Oprescu (1994)). Moreover, vocational trained workers have experi­ enced both a falling relative wage and relatively high unemployment rates in several transition economies. Given the leading role of the private sector, a closer examination of private-sector wage policies is in order.

II. The Role of the Private Sector

The Czech Republic entered the transition process with less private sector activity than the other major east European countries-Hungary and Poland. But by early 1993, almost 30 percent of the employed respondents to the Survey of Economic Expectations and Attitudes worked in private employment. The privatization of large state enter­ prises followed over the next few months, and by the November 1994 survey only about 41 percent of employment remained in traditional state enterprises (Table 2). Another 28 percent of employment was in privatized state enterprises, most of which appear to have been state enterprises ten months earlier. Broadly construed, the private sector (including these enterprises) provided almost 60 percent of the nation's employment in late 1994. Table 2 also provides a snapshot of human resources in the three major sectors in November 1994. Average educational attainment (EDYEARS) is actually lower in new private firms than in the state or privatized state

©International Monetary Fund. Not for Redistribution WAGE STRUCTURES IN THE CZECH ECONOMY 843

Table 2. Means and Standard Deviations, Employfd Wo rkers, November 1994 Total Private State StPriv (N = 864) (N = 242) (N = 350) (N = 239) EDYEARS 11.72 11.51 11.95 11.64 (2.49) (2.33) (2.65) (2.43) PRIM 0.15 0.15 0.15 0.14 (0.35) (0.35) (0.35) (0.35) VOCAT 0.41 0.47 0.33 0.46 (0.49) (0.50) (0.47) (0.50) HS 0.31 0.27 0.35 0.27 (0.46) (0.45) (0.48) (0.44) UNJV 0.14 0. 11 0.17 0.13 (0.35) (0.32) (0.38) (0.34) EXP 22.29 19.94 23.14 22.26 (1 1.81) (11.95) (11.14) (11.48) AGE 40.01 37.45 41.09 39.90 (11.53) (11.94) (10.63) (11.26) FEMA LE 0.49 0.43 0.55 0.44 (0.50) (0.50) (0.50) (0.50) EARN 7185 8837 6321 7014 (5217) (5224) (3555) (5999) LEARN 8.74 8.90 8.65 8.73 (0.50) (0.59) (0.43) (0.46) PRIVATE 0.28 1 0 0 (0.45) STATE 0.41 0 1 0 (0.49) PR/V, FORMER STATE 0.28 0 0 (0.45) Note: EDYEARS =years of schooling. PRiM= primary school degree is hi hest completed schooling. AGE= age in years. See text and notes to Table 1 �or other definitions. Standard deviatiOns are in parentheses.

sectors. New private firms, which provide the highest returns to a univer­ sity degree, have the lowest share of workers with university degrees. Employees in the private sector also have less experience than in the state sector, although the returns appear to be greater. Despite their lower level of human capital investment. full-time em­ ployees in the private sector earned considerably more than their coun-

©International Monetary Fund. Not for Redistribution 844 ROBERT J. FLANAGAN terparts in the state sector in the Czech Republic. (In Table 2, EARN is the monthly wage of full-time workers in Czech koruny and LEARN is the natural logarithm of the monthly wage.) Moreover, the regression analysis reported in Table 1 also confirmed the presence of higher wages in the private sector after holding human capital investments constant, raising the question of how and why such large wage differentials emerge. One hypothesis is that the differential is necessary to attract tabor resources into the emerging private sector. Research in western tabor markets has found that tabor mobility responds to both relative wage and job vacancy mechanisms. In tabor markets with considerable unemploy­ ment, the existence of job vacancies may be enough to attract workers from unemployment. With unemployment low, as it has been during the transition in the Czech Republic, wage differentials may also be necessary to attract workers. Moreover, wage differentials are likely to be necessary when private firms hire directly from state firms rather than from the pool of unemployed, as appears to be the case in the Czech Republic and other transition economies (Flanagan (1994a)). 3 A second possibility is that the private sector must also offer wages that compensate for differences in nonwage benefits and/or job security be­ tween the state and private sectors. Unfortunately, the data in the Survey of Economic Expectations and Attitudes do not include sufficient in­ formation to permit direct assessment of this hypothesis. By 1994, it was by no means clear in which direction a compensating wage differen­ tial should go, however. The job security advantage of the state sector had clearly declined. Moreover, a 1992 survey of private manufacturing firms in the Czech and Slovak Republics reported that while workers received "few fringe benefits beyond the minimum mandated by labor laws ...[f)ringe benefits were about three times higher in new start-ups than in privatized state enterprises.'' (Webster and Swanson (1993), pp. 33-4). Further progress on this issue awaits detailed data on the composition of total compensation packages in each sector.

Company Wage Policies During the Transition Ill. The foregoing discussion focused on the evolution of a relative wage structure that would raise productivity through improved allocation of tabor. In a world of long job tenures and "career'' employment relation­ ships, productivity also depends on incentives established in the internal

'Because of housing restrictions, effective labor market competition i� likely to be local, rather than national. :.o that large differentials are not to attract workers over long distances. necessary

©International Monetary Fund. Not for Redistribution THE CZECH ECONOMY WAGE STRUCTURES IN 845

labor markets of enterprises. As yet, there is little systematic institutional information about the human resource management policies adopted during the transition, and most information must be inferred from wage structure developments in the various sectors. The Survey of Economic Expectations and Attitudes enables consideration of three potential sources of differences between the sectors: (I) the nature of career wage profiles; (2) the role of foreign investment on wage policies; and (3) the use of screening and signaling devices. As before, it will be instructive to decompose the private sector into new private firms and privatized state enterprises.

Career Wage Profiles

Several economic theories stress the desirable incentive effects associ­ ated with steep career earnings profiles. The oldest is the notion that steeper career earnings profiles are associated with longer periods of human capital investment, since with a shorter career span, earnings must increase more rapidly to provide a sufficient return to encourage the investment. In addition, the principal-agent literature has stressed the performance incentives provided by relatively steep earnings profiles when work cannot be monitored continuously. Workers who are paid less than they are worth early in their career have an incentive to work sufficiently hard to receive more than they are worth later in life. Flatter profiles produce more equity between younger and older workers, which may encourage more on-the-job cooperation, but otherwise offer weaker

incentives. The regression results reported in Table I clearly indicate that career wage profiles are most flat in state enterprises, consistent with the greater use of wage incentives to foster career attachments in the private sector.

Wage Policies of Foreign Investors

The process by which new employers in a transition economy come to adopt wage structures and policies that differ from their prior experience with state enterprises is not well understood. ln the markets for goods and services, world prices often provide an immediate benchmark for a tran­ sition economy, bur channels of information on wages are less obvious. One potential source of transmission of market-oriented policies is for­ eign ownership and/or investment. Foreign owners and investors may encourage or import from operations in other countries market-oriented wage structures and/or apply wage policies intended to encourage greater

©International Monetary Fund. Not for Redistribution 846 ROBERT J. FLANAGAN

Foreign lnvestmeflt and Wages by Sectors, November Table 3. 1993 Total State StPriv Private Percent of respondents working in estabUshments with high foreign investment 12 6 25 10 Wage impact associated with substantial foreign investment (HIFORKAP) (t-statistics in parentheses)

Controls : Schooling and 0.114 0.132 0.093 -0.010 experience (2.13) (1.70) (1.20) (0.07) Schooling, experience, 0.049 O.Q78 0.086 -0.043 and industry dummies (0. 91) ( 1.00) ( 1.11) (0.32) Source: Survey of Economic Expectations and Attitudes, November 1993.

effort. In the latter case, there have been anecdotal reports that foreign firms tend to adopt efficiency wage policies in transition economies in an effort to attract and motivate high-quality workers. If this is a general phenomenon, a high foreign presence in a firm should be statistically associated with high-wage policies. The November 1993 Survey of Economic Expectations and Attitudes includes questions about the importance of foreign capital in the respon­ dent's firm but does not explicitly ask whether the firm is foreign-owned. Twelve percent of the respondents indicated that a major part of the financial capital of their organization came from abroad (HIFORKAP); another 2 percent indicated that foreign capital played a small role (Table 3). Relatively few employees of state enterprises (6 percent) report a major foreign financial influence. There are notable differences in foreign financial presence within the private sector, however. Of em­ ployees in privatized state enterprises, 25 percent report such influence, in comparison with 10 percent of employees in other private firms. In part, this reflects the tendency of foreign capital to flow to the larger enterprises. Foreign capital also tends to be found in relatively high-wage industries. The sample of employees working for firms with HIFORKAP is too small to analyze the effect of foreign investment on the overall wage structure. Nevertheless, it is possible to test for a relationship between foreign investment and efficiency wage policies. To examine whether Czech firms with substantial foreign investment pay relatively high wages for given human capital attributes, human capital earnings functions with

©International Monetary Fund. Not for Redistribution WAGE STRUCfURES IN THE CZECH ECONOMY 847

a dummy variable for HIFORKAP were estimated for the total sample and each of the three sectors. Only the results for the HIFORKAP variable are reported in Table 3. We see that, economywide, workers in firms with substantial foreign investment appear to earn a wage premium of about 11 percent relative to similarly qualified workers in other firms. This result reflects the tendency of foreign investment to flow toward high-wage industries in the Czech Republic, however, and the finding disappears with the addition of industry dummy variables (Table 3). No significant relationship between wages and HIFORKAP is found for any of the individual sectors with or without controls for industry. It appears that there is no general tendency for firms with substantial foreign in­ vestment to follow efficiency wage policies. Consistent with these wage patterns, firms with HIFORKAP do not appear to attract em­ ployees with particularly high investments in human capital. HIFORKAP is most likely to be found in relatively large firms employing compara­ tively few workers with university education and offering comparatively little training.

Use of Screening and Signaling Devices

In virtually all hiring decisions, employers are unable to observe all aspects of an applicant's productive potential, and the applicant is incom­ pletely informed about the prospects of the firm as a place of work. Both parties have an interest in relying on observable characteristics to reduce uncertainty. For example, profit-maximizing employers who view prior unemployment experience as signaling unobservable quality deficiencies will offer workers with such experience lower wages or lower-wage posi­ tions. There is already evidence from transition economies that private­ sector employers prefer to hire new workers directly from state enter­ prises rather than from the pool of unemployed (Flanagan (1994a)). Additional evidence may be obtained by testing for wage differences associated with prior unemployment experience. The Survey of Eco­ nomic Expectations and Attitudes elicits information on unemployment experience in the past two years. When a dummy variable for such experience is added to the basic human capital wage specification, prior unemployment experience does tend to depress wages by about 10 per­ cent in the private sector. In other sectors, prior joblessness has no significant influenceon wages.

©International Monetary Fund. Not for Redistribution 848 ROBERT J. FLANAGAN

IV. Union Wage Effects During the Transition

In Czechoslovakia, as in other former Soviet bloc countries, union membership was compulsory but labor unions had a distinctly circum­ scribed role under central planning. As workplace organs of the Commu­ nist Party, the unions' major role was to assist in the production process, and representational functions were Jjmjted to a kind of grievance han­ dling in rare dismissal cases. (Under central planning, labor unions also distributed "workplace public goods," including recreation and cultural benefits.) Collective bargaining over the terms and conditions of work, as it is understood in the west, did not exist. As in the rest of the region, Czech unions rapidly transformed to conduct normal wage bargaining following the velvet revolution. There was a further change in the insti­ tutional arrangements after the division of the country into the Czech and Slovak RepubJjcs in early 1993. This institutional development holds the potential for introducing dis­ tortions into transition wage structures that were not present under central planning. In market economies, unions have had two types of relative wage effects. First, they tend to raise the wages of union workers relative to nonunion workers. In economies in which both a union and nonunion sector can be observed statistically, union wage levels exceed nonunion wage levels for a given quality of worker. The union wage premium averaged 15 percent in the United States between the mid-1950s and late 1970s, but declined somewhat in the 1980s (Lewis (1986)). The union-nonunion wage differential is somewhat lower in the United King­ dom (Metcalf (1993)). Second, unions tend to narrow pay distributions in ways that reduce returns to schooling. Looking across countries, this effect appears strongest in economies with centralized bargaining, where unions can coordinate nationwide "solidaristic" wage policies. But even in countries with decentralized collective bargaining (e.g., the United States), returns to schooling tend to be relatively low among unionized workers. In principle, then, collective bargaining might inhibit the transformation of wage structures inherited from central planning. Thjs section examines the extent to which unions have had such an effect. Until recently, assessments of union strength in eastern Europe came mainly from the membership claims of trade unions. Evidence from surveys of privatized firms in the region indicated that union repre­ sentation was largely Limjted to the state sector. The November 1994 Survey of Economic Expectations and Attitudes asked respondents whether they were union members, thus permitting a more precise assess­ ment of the pattern of unionization and of the relationship between union

©International Monetary Fund. Not for Redistribution WAGE STRUCfURES IN TilE CZECH ECONOMY 849

4. Union Membership Rates, November Table 1994 (Standard errors in parenthe es) s Men Women Total .55 .55 (.50) (.50) State .75 .74 (.43) .44 ( ) StPriv .82 .70 (.38) (.46) Private .08 .08 (.27) (.28) Source: Survey of Economic Expectations and Attitudes, November 1994.

representation and wages in the Czech Republic about five years into the transition.

Patterns of Union Representation

By these measures. about 55 percent of workers in the Czech Republic were union members in November 1994 (Table 4). in contrast with virtu­ ally 100 percent unionization under central planning. The membership estimate is in the upper end of union density rates in west European countries but exceeds membership rates in North America (OECD ( 1994), Chapter 5).4 At first glance, the extent of union representation indicates a potential for substantial union influence on wage structures during the transition. We now examine the data for clues as to whether the potential is likely to be realized. The estimate of average union membership masks a huge difference in union presence in the state and private sectors. While unions have re­ tained most of their members in ongoing and recently privatized state enterprises, they barely have a toehold in the new private enterprises that have emerged during the transition (Table 4). At 8 percent, the member­ ship rate in the private sector falls below the lowest density figures for

'The coverage of collective bargaining agreements often exceeds union mem­ bership. although the difference varies from a few percentage points in North America to over percentage points in Austria, France. Germany. and Spain 50 5). h (OECD (1994). Chapter T e Survey of Economic Expectations and Attitudes does not inqUire into collective bargaming coverage in the Czech Republic.

©International Monetary Fund. Not for Redistribution 850 ROBERT J. FLANAGAN

Table 5. Union Wage Effects, November 1994

(t-statistics in parentheses) Controls Total State StPriv Private

-.177 -.060 -.032 -.208 (5.45) (1.24) (.49) (1.59) PRIVATE -.055 (1.40) PREMP, PROWN -.046 -.15 (1.18) (1.21) FEMALE -.182 -.259 -.10 -.206 (6.13) (1.33) (1.64) (1.70)

FEMALE, PRIVATE - .082 (2.27) FEMALE, PREMP, -.073 PROWN (2.05) Source: Survey of Economic Ex ectations and Attitudes, November 1994. Note: All r essions also inclu J: e following control variables: HS, VOCAT, UNIV, EXP, �XPSQ.

the private sector in market economies. 5 Efforts to determine a more extensive set of determinants of union membership (including education, experience, and other individual characteristics) using probit analysis were unsuccessful. Broad sector of employment (state versus private) remained the dominant influence. Thus, the future of union representa­ tion in the Czech Republic may depend on the ultimate balance between newly created private firms and privatized state enterprises.

Union Wage Effects

If the assumption of normal collective bargaining activities by unions during the economic transitions has retarded the emergence of com­ petitive wage structures, there should be measurable union wage effects. Table 5 reports on efforts to measure such effects among em­ ployed workers in the Czech Republic in November 1994. The union

1993 'Similar patterns occur in a late survey of 200 manufacturing rirms in Poland. The survey report s no union representation in the 40 ew private firms in the sample, although significant union membership rates existn in current and former state enterprises (Belka and others Table 15). (1994),

©International Monetary Fund. Not for Redistribution WAGE STRUcrURES IN THE CZECH ECONOMY 851

membership dummy variable (MEM) has been added to'standard human capital earnings functions to test for union wage impacts. Table 5 reports only the coefficients and !-statistics for the union membership (MEM) variable. In western market economies union wage effects tend to be on the order of 10-15 percent, with considerable variation by sector, reflecting variations in bargaining power. From this perspective, the remarkable finding in the economywide estimate in Table 5 is that for a given level of education and experience, union members appear to earn 18 percent less than nonmembers in the Czech Republic. Negative union-nonunion wage differentials are essentially unheard of in research on western market economies! Averages can be deceptive, however. In particular, we have already reviewed evidence indicating that private sector workers earn more than state-sector workers, and that union membership rates are much higher in state enterprises. The regressions clearly rule out the possibility that relatively high wages in the private sector can be attributed to the pres­ ence of unions, but they do not rule out the possibility that the negative union wage differential is an artifact of the sectoral distribution of union members and thus must be explained by whatever factors explain the difference in state-private wages. In fact, this appears to be the case. With the addition of dummy variables for private sector employment (PRIVATE in row 2 and PREMP and PROWN in row 3 of Table 5), the sign on the union representation variables becomes statistically insignif­ icant. Sector-specific regressions also find no significant union wage effect in either the state or private sectors (see the last three estimates in first row).6 In the early years of the transition, there is no indication of a positive union wage effect. One explanation is that unions in !his transition economy have little power since neither subjective measures of worker support nor objective measures of impact provide evidence of union power. A second interpretation is that incomes policies have prevented the exercise of union economic power. Since incomes policies were, prior to the summer of 1993, applied only to the state sector of the Czech Republic, this might also contribute to an explanation of the wage differ­ ences between the state and private sectors. Unfortunately, there are insufficient degrees of freedom to disentangle these interpretations, al­ though the, absence of a union wage impact in the private sector, where

6 Again, there is a parallel to the case of Poland, where managers in all sectors report that unions are not a significant factor in wage determination (Belka and others (1994), Table 14).

©International Monetary Fund. Not for Redistribution H52 FLANACii\N I�OBE- RT J

income:, were not applied until prO\ idc:, \01111: �urport for the low-union-powerp\1licic� h) pothc..,i�. llJlJ3.

V. Conclusions

Four years into the economic transition, there is strong evidence that wage structures in the Czech Republic are moving away from the patterns produced by central planners' preferences and toward patterns observed in market economies. The change was led by developments in new private firms but somewhat retarded by the initially sluggish response in state enterprises. There are also signs that new private firms pioneer with more advanced incentive policies. although firm-level data would permit a much richer analysis of the extent of such changes. A key question for future labor market adjustment during the transi­ tion is why wage structures in state enterprises. which account for about half of employment in the Czech Republic. were slow to adjust. ln the face of competition from new private firms for highly educated personnel, both current and former (recently privatized) state enterprises have only recently adjusted wage structures sufficiently to match outside private offers. To what extent may market forces be muted by institutional developments? A possibility suggested by experience in market economies is that collective bargaining, which occurs mainly in state enterprises. retards the adjustment of wage structures in the state sector. As of late 1993, however, there was no indication of countervailing wage distortions from the establishment of genuine collective representation at the workplace. Indeed, unions appear to have no impact on wages in either the state or private sectors. (Unions may, of course, negotiate direct quantitative restrictions on labor reallocation via dismissal regulations, etc .. but the surveys provide no information on this issue.) A second possibility is that incomes policies, which also usually apply only to the state sector. have retarded the adjustment of wage structures. The application of incomes policies in transition economies is somewhat paradoxical. The case for the policies rests on the argument that state managers facing soft budget constraints and an uncertain future have little incentive to resist union wage demands. Yet, the tax-based incomes policies that are typically applied in the region arc designed to encourage n managers facing hard budget constraints to bargain harder!7 r any event. we have seen that the case for incomes policies seems particularly weak

in Flanagan 'Tht:�� point'> ar� Uc\dopcu further ( Jl)l)-la ).

©International Monetary Fund. Not for Redistribution WAGE STRU(TURI::S IN I HI· ('/ICH f:CONOt-.n 853 in the Czech epublic . where there is no evidence of union wage effects R in either the state sector. where such policies appl). or the private sector. where they did not until mid-1993. This does not quite prove that incomes policies did not restrain union wages in the state sector. but layoffs and diminishing job security were an important counterweight to weak management. To the extent that incomes policies are successful. their potential for retarding labor market adjustment is great. Contrary to transition objec­ tives, the particular wage rules adopted tend to encourage wage freezes or further compression of internal wage structures in enterprises. Ulti­ mately this has perverse macroeconomic consequences. as it forces more of the burden of market adjustment into quantity responses. such as unemployment and laborforce withdrawal. rather than wage adjustment.

REFERENCES Adam. Jan. Employment and Wage Po/icie.1 rn Poland. C:.ec!toslm ·aAw mu/ Hungary Since /950 (New York: St. Martin\ Pre��- IIJH-l). Bclka. Marek. and others. "Evtdencc.: from a Survey' of Statc.:-Owned.Privatized and Emerging Private Firm!>.. . Papc.:r preparett for Work�hop on Enterprise Adju�tment in Eastern Europe. September 22-23, I t Europe and Rus�ia. October 1993 (unpubli�hed; Wa�hing.tnn:World Ban!-.. 1993). Earle, John S., and Gheorghe Oprescu. "Aggregate Labor Market Bchavior in the Restructuring of the Romanian Economy." Paper prepared for World Bank Research Project on the Labor Market in Transitional Sociali�t Economies (unpublished: Washington: World Bank, 1994). Flanagan. Robert.. J. ( 199..\a ). "Lahnr Market Response� to a Change in Economic System . in Proceedings of the World BanA Annual Conference on Del'£•1op­ ment EconomiC\ (Washington: World B

--- ( 199-lb). "Were Communisb Good Human Capttaliqs'! The Ca�e of the Czech Republic ... Stanford Uni,Cr\it} Graduatc So.:hool of Bu,incs� Work­ ing Paper (Stanford. California: Stanfnrd Univc.:rsity. June 19l)..t). Union Relatil·e Wage Effects: Surrey Lcwi�. I!. Grcgg, t\ (Chtcago: Univcrsity of Chicago Press. 19H6). Metcalf, David. ·Transformation of British Industrial Relations? Institutions, Conduct and Outcomes 1980-90 ... CEPR Discussion Paper No. 151 (Lon­ don: London School of Economics. Centre for Economic Performance. June 1993). Employml'lll Owlook l9tJ3 OECD. (Pari�: OECD. and ltJIJ-l).

©International Monetary Fund. Not for Redistribution 854 ROBERT FLANAGAN J.

Rutkowski, Jan. "Changes in Wage Structure and in Returnh to Education During Economic Transition: The Case of Poland'" (unpublished; Prince10n. New Jersey: Center for International Studies. Woodrow Wilson School. Princeton University. March 1994). Webster, Leila M., and Dan Swanson, "The Emergence Private Sector Man­ of ufacturing in the Former Czech and Slovak Federal Republic: A Survey of Firms," Wo rld Bank Technical Paper No. 230 (Washington: World Bank, 1993).

©International Monetary Fund. Not for Redistribution IMF Staff Papers Vol 42. No. 4 llJ

Pricing to Market and the Real Exchange Rate

* HAMID FARUQEE

This paper investigates the consequences of pricing to market for exchange rate pass-through and real exchange rate dynamics across diffe rent pauerns of trade. Under two-way, intraindustry trade-where home prices display greater linkage with those of fo reign competitors-domestic and export prices exh ibit lower pass-through and greater desrinarion-specific adjust­ ment compared to intersectoral trade. With both trade patterns, pricing-to­ market behavior intensifies the degree of persistence in the real exchange rate under nominal rigidities, and allows monetary shocks have perma­ to nent effe cts on relative prices so long as goods markets remain segmented. [JEL £12, F12, F41]

NYPLAUSIBLE THEORYof real exchange rate determination must be Aable to resolve two prominent stylized facts regarding international relative prices. First, at ao aggregate level, real exchange rate movements tend to be very persistent. Second, at the micro level, individual prices tend to be sticky in terms of local currency.1 Of course, open-economy models in the Keyoesian tradition have long since emphasized sticky

• Hamid Faruqee is an Economist in the Research Department. He received his Ph.D. from Princeton UniversitY.- Work on this paper began while the author was a lecturer in the Woodrow W1Ison School of Public and International Af­ fairs. He would like to thank William Branson, Larry Ball, Rex Ghosh, Michael Knetter, and IMF Research Department Seminar participants for helpful comments and discussions. 1 Comparing the relative price of different goods within the same country versus the relat1ve pnce of the same good across different countries, Engel (1993) finds the former measure to be less variable in all but a few cases such as energy prices and primary commodities. Moreover, the second relative price tends to be several times more variable on average, confirmin� that the local prices prevailing in a given market destination remain comparatively quite stable. 855

©International Monetary Fund. Not for Redistribution 856 HAMID FARUQEE prices in goods and labor markets. Perhaps the best-known example is the Dornbusch (1976) overshooting model, wherein sticky goods prices in conjunction with flexible, forward-looking exchange rates lead to real exchange rate overshooting and transitory deviations from purchasing power parity (PPP). However, by invoking the law of one price, this literature incorporates price stickiness at the point of origin rather than at the market destina­ tion. In other words, the standard Keynesian approach generates price­ level inertia in terms of national currency (seller's currency) rather than local currency (buyer's currency). Furthermore, by relying on the law of one price, these models typically cannot generate the degree of persis­ tence that is observed in actual real exchange rate time series. Not only do real and nominal exchange rates move together over the short term, they often move together over much longer horizons as well-sometimes over many years. The law of one price has of course had a long-standing tradition in international economics. Interpreted as a condition for spatial arbitrage, the law of one price requires that the price of a good be the same in all locations when measured in a common currency. Extending this condi­ tion to all traded goods, PPP theory posits an aggregate version of the Jaw of one price, applied at the level of national price indices. The theory maintains that the price of entire baskets of goods should be equal (or proportional) across countries. To gether, PPP and the law of one price have served as the basis for some of the most widely-held theories in international economics, greatly influencing the discourse on exchange rate determination. Meanwhile, empirical evidence supporting the law of one price has remained elusive. Numerous empirical studies have consistently docu­ mented the absence of "one price" in practice, net of transport costs. tariffs, and other trade impediments.1 In combination with the very mixed evidence on PPP,3 some have begun to call into question those theories that rely upon the concept of spatial arbitrage in goods.4 Given

2 See Engel (1993) for a study on the G-7 countries. Giovannini (1988) and Marston (1990) document evidence of pricing to market practices within partic­ ular Japanese tndustries. Other studies on the failure of one price include Isard (1977), Mann (1986), and Knetter (1989 and 1993). 3 See Breuer (1994) for a recent survey of the PPP literature. "Krugman (1989, p. 43) summarizes this djssatisfaction: [W e must now admit that international Keynesianism, while more like reaj ity than international monetarism, itself turns our to have a problem: lt does not go far enough in rejecting international arbitrage. Not on l y does the Law of One Price fail to hold at the level of aggregate national price indices ...it doesn't even hold at the level of individual goods.

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 857 its dubious validity, the law of one price has been subject to various modifications,5 including the allowance of systematic violations to one price as manifested by the phenomenon of pricing to market. With pricing to market, forces that would normally assure spatial arbitrage are absent, allowing corresponding prices to diverge across markets. This view-which is adopted in this paper-essentially breaks from standard models by favoring market segmentation rather than inte­ gration, and thereby acknowledging the presence of economic barriers and structural rigidities that restrict convergence in intermarket prices. Pricing to market may thus provide an alternative to the traditional view that differential rates of adjustment in goods prices versus asset prices alone underlie persistence in the real exchange rate. Much of the empirical evidence on market segmentation has drawn from the experience of United States in the 1980s. In particular, several studies have focused on the behavior of U.S. import prices during the period of massive appreciation(> and subsequent depreciation' of the dollar. For example, Krugman (1987) examines U.S. import prices dur­ ing the period of the dollar's rise. Constructing dollar price indices for U.S. import bundles and comparing them with the dollar price of the same bundle exported to third-country markets, he concludes that more than 30 percent of the real appreciation of the dollar was reflected in a divergence between these prices. In light of the empirical support for the proposition that certain mar­ kets may be segmented, the theoretical avenues regarding the issue of pricing to market have not yet been fully developed. Typically, models

5 Aizenman (1984), for example. illustrates that when transport or informa­ tion costs impede arbitrage over the very short term, PPP holds up to constant plus white noise. However, it is the persistence of relative-price movements that dominate the time series of the real exchange rate that requires further explanation. During the dollar upswing between 1980 and 1985, Dornbusch (1987, p. 104) observes the following fact regarding U.S. import prices: [T]he order of magnitude of the decline [in import prices] remains relatively small compared to the change in relative unit tabor costs. With a change in relative unit labor costs of more than 40 percent, the decline in the relative price is in most cases less than 20 percent. That is not at all out of line with the theory once some degree of "pricing to the American market'' is taken into account ....

7 For the period 1985-87. when the dollar fe ll precipitously, Hooper and M ann (1987) find that import-price increases. in percentage terms. were well short of the nominal depreciation. Krugman (1989) reports a similar finding in the specific case of Japanese exports in manufactures. fi nding that export prices (in dollars) were relatively stable in the destination market despite sharply rising unit labor costs (in dollars) at the point of origin.

©International Monetary Fund. Not for Redistribution 858 HAMID FA RUQEE of pricing to market present static, partial equilibrium results from the perspective of an exporting industry or firm .x Hence, these models cannot address the dynamic price interactions and adjustment underlying the evolution of relative prices and the real exchange rate at the macroeco­ nomic level. Using a general equilibrium model of monopolistic competition and market segmentation, this paper investigates the macroeconomic conse­ quences of pricing to market for real exchange rate behavior across different patterns of trade. Allowing producers to price discriminate between local and foreign markets. the monetary model examines ex­ change rate pass-through and pricing-to-market behavior under seg­ mented markets, and relative-price dynamics under nominal rigidities. By also incorporating the structure of trade, the paper examines the cross-sectional implications of intersectoral versus intraindustry trade for macroeconomic adjustment.9 The central result can be summarized as follows. The pattern of indus­ try specialization and trade largely determines the degree of strategic complementarity or price linkage between producers from different countries. In particular, under intraindustry trade-wherein home and foreign products are close substitutes-there exists a higher degree of linkage between domestic and foreign prices prevailing in the same market than under intersectoral (cross-industry) trade. Consequently, domestic and export prices exhibit greater responsiveness to exchange rate fluctuation, a lower degree of pass-through. and a greater degree of pricing to market under two-way trade. In a setting of imperfect integration, 111 a greater degree of price linkage across borders translates into stronger mean reversion in the real ex­ change rate. The intuition is as fo llows. With menu costs and price staggering, prices display inertia in terms of national currency (unit of account) through a lack of coordination among decentralized price set­ ters. In an open economy, however, producers are concerned with rela­ tive prices both at home and abroad. With the law of one price, the added concern with the foreign currency price of output compels price setters to partly overcome the coordination failure underlying domestic price

x A partial exception is Delgado (1991), which develops a dynamic menu cost model of pricing to market, albeit in a partial equilibrium setting. '1 See also Faruqee ( 1994 ). 10 Under imperfec t integration in world markets for goods and services. coun­ tries differ in their national consumption patterns and in the units of account in which they set prices-favoring both the1r own goods and their own currency. However, the law of one price still equates currency-adjusted prices across markets. For a general discussion of imperfect integration. see Krugman ( 19R9).

©International Monetary Fund. Not for Redistribution PRICiNG TO MARKET 859 inertia. Prices become less sticky in terms of national currency in order to reduce variation in terms of foreign currency. Hence, the stronger international linkage in prices under two-way trade red11ces the variability and persistence of real exchange rate fluctuation compared with inter­ sectoral trade.'' However, once the assumption of spatial arbitrage is removed, this comparative result no longer holds true. Once export prices detach from domestic prices with segmented mar­ kets, prices are in effect set in terms of local currency rather than national currency, and price linkages predominantly exist within a particular market rather than across markets. If home and foreign prices are inertial in the same unit of account, the pattern of trade no longer affects the comparative degree of variability and persistence in relative prices. Instead, the level of persistence becomes solely a function of the fre­ quency and timing of price adjustment. As a result, market segmen­ tation allows international relative prices to exhibit equally persistent deviations across diverse patterns of trade, despite inherent differences in the substitutability between home and foreign goods. Comparative issues aside, pricing to market ensures grearer price sta­ bility in terms of local currency than otherwise. regardless of the trade pattern. Consequently, segmented markets allow greater inertia in the price level domestically and slower adjustment in relative prices interna­ tionally than under the law of one price. Thus, pricing to market provides an important propagation mechanism for explaining the large and pro­ tracted swings in real exchange rates characteristic of the post-Bretton Woods era. Moreover, in the presence of market segmentation, mone­ tary shocks have lasting effects on relative prices. Without goods arbi­ trage, absent are the economic forces that would otherwise guarantee the equivalent price of a basket of goods across countries. Instead of long-run PPP, monetary disturbances have permanent effects on the real exchange rate in the absence of one price. Persistence and permanence in relative prices have important time­ series implications for exchange rates. When monetary shocks have per­ sistent effects on relative prices as a result of price level inertia, nominal and real exchange rates move together in the short run. as has been emphasized in earlier sticky-price models. When monetary shocks have "'permanent" effects on relative prices. nominal and real exchange rates move together over the longer run as well-so long as markets remain segmented and the law of one price systematically fails.

The paper is organized as follows. Section I develops a two-country model of monopolistic competition and market segmentation under in-

11 See Faruqee (1994).

©International Monetary Fund. Not for Redistribution 860 HAMlD FARUQEE traindustry and intersectoral trade. Using this static framework, Section 11 addresses the issues of exchange rate pass-through and pricing to market. A dynamic version of the model is developed in Section Ill. ncorporating nominal rigidities into the analysis, this section examines theT dynamic adjustment of relative prices and the time-series behavior of the real exchange rate. Section IV offers some concluding remarks.

Pricing I. Model of to Market Consider a world economy consisting of a home and a foreign country, with each country composed of producer-consumers. These individual agents produce and sell a differentiatedN good in order to purchase and consume the variety of goods made by all agents, taking others' prices as given. Meanwhile, each agent acts as a monopolistic competitor, setting a price and choosing a level of production depending on the demand that the individual producer faces. Building on the constant elasticity of substitution (CES) approach of Dixit and Stiglitz (1977), the model focuses on two basic variants of this central framework: the cases of intersectoral and intraindustry trade. 12 Under intersectoral trade, countries specialize and trade at the industry level according to international differences in comparative advantage. This pattern of trade-often associated with the exchange between north and south-emerges as a result of underlying differences in relative factor proportions and gains from cross-industry trade. Thus, this framework essentially embodies the traditional Heckscher-Ohlin view of interna­ tional trade, although, in this context, each sector is characterized by monopolistic competition rather than by perfect competition. Under intraindustry trade, the premise is quite different. Countries are now essentially identical with respect to factor composition, yet they still gain from specialization and trade at the variety level because of scale economies and product differentiation. This modern interpretation of trade offers an explanation for the two-way, intraindustry exchange often observed between OECD countries, but left unaccounted for under the standard factor proportions theory. Thus, the second case draws from some recent developments in the noncompetitive theory of international trade. 13 Under both trade patterns, the local and export demands facing an individual home producer i are given in Ta ble 1.1� Note that monetary

11 See Appendix for details and the basic setup of the model. " See, for example, Helpman and Krugman 14 1985). See Appendix for details. (

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 861

Table I. Domestic and Export Demand Fu nctions Under Monopolistic Competition

Pattern of trade Home market Foreign market

P, EM* Intersectoral trade Y, = p (1 - a)- ( ) •( p ) lntraindustry trade Y, = � (8) '( �) variables with asterisks are expressed in units of foreign currency, and is the nominal exchange rate expressing the domestic price of foreignE currency. Also in Table 1, E > 1 represents the constant elasticity of substitution between any two varieties from the same industry, while et and j3 are taste parameters, measuring the extent to which countries symmetrically favor their own goods in consumption. These latter two measures roughly capture the nominal expenditure shares allocated to locally produced goods from each country. 15 Assuming that countries predominantly consume the goods that they produce themselves, the model assumes that 0.5 < et,j3< 1 under local goods preference. From Table 1,demand for output in each market under interindustry trade is a function of two components:Y; relative price and aggregate demand. Demand for a particular variety is a decreasing function of its price P; relative to other prices in the industry, where P is the relevant index of producer prices set by producer i and his (n 1) compatriots. Second, product demand for each variety depends - on aggregate de­ mand facing the industry in each country, where real money balances measure real aggregate demand in each market through a simple quantity equation relationship (see Appendix).

More precisely, with interindustry trade, the domestic CP! is a function of prevailing15 home and foreign producer prices given by

Q (P)"(EP*)1 «, = where a is the exact expenditure share on home goods. Under intraindustry trade, the home CPI and expenditure share on home goods are 1 * 1 ' 1 , Q = (rlP (1 - r.!.)EP j1 an d a + �-" '+ �-" ---� (l - �)(EP*I� P)1 ' Given relative producer prices in general equilibrium. is chosen so that ,; = a, keeping expenditure patterns identical across trade patterns.[3

©International Monetary Fund. Not for Redistribution 862 HAMID FA RUOEE

In the case of intraindustry trade, variety demand is once again a function of relative price and aggregate demand. However, since foreign goods now compete in the same industry, the relative-price component is a function of the individual price relative to other home and foreign industry prices. This modification in demand is seen by the ratio of P; to the domestic CPI Qat home and the foreign CPI EQ* abroad, where the consumer price indices and producer price indices naturally coincide under trade in one industry. Moreover, real money balances also reflect the appropriate deflator to capture real aggregate demand in the relevant market. Note that the demands under two-way trade assume from the outset that country size is not an issue (n = n*). Otherwise, a measure of relative country size would also enter the demand under intraindustry trade, further distinguishing it from the case of intersectoral trade. le. Introducing market segmentation into the analysis, the model also assumes that each producer can price discriminate across markets, setting a different price in each location if the agent so chooses. This assumption represents the critical point of departure from Faruqee (1994), where the law of one price applied. Now, with segmented markets, the implicit arbitrage mechanisms that assured the law of one price are absent. Consequently, each home and foreign producer can price to each market separately, depending on the market conditions that prevail in each location. In standard models of market segmentation, firms price to market by making destination-specific adjustments to price-cost margins in re­ sponse to exchange rate changes. However, in this CES framework, markups are constant, closing the usual channel through which prices may systematically differ. Hence, to ensure that producers have an incen­ tive to price discriminate given their ability to do so, it is also assumed that costs are separable (and convex) in the production of domestic and export goods. 17 Otherwise, producers would always choose the same price for each market under either pattern of trade. 18

16See Helpman and Krugman (1985). 17 There are many justifications for the premise of differential costs. lf there exist market specific costs in transportation, distribution, production. advertising and/or servicing, then costs can differ at the margin for the home and exported good. For example, foreign markets may require different product specifications and/or have different governmental regulations that differentiate costs of produc­ tion; producing the export good may even take place in the destination country itself. involving a completely separate plant and production run. These and similar explanations may also help explain why markets are actually segmented in the first place. IR Constant differential markups could be introduced into this CES framework by assuming differential elasticities of substitution across markets =I= In that case. there would exist a constant degree of pricing to market. ( E E").

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 863

Qualitatively, pricing behavior under cost separability is similar to the pricing-to-market responses under differential markups. Usually, firms reduce markups in markets where the currency has weakened in order to stabilize price in terms of the local currency. 19 In this context, exchange rate movements elicit a very similar response. For example, a deprecia­ tion of the foreign currency lowers demand for home exports, which are now more expensive abroad. Consequently, faced with reduced demand and cost, home producers respond by lowering export prices in units of home currency in order to offset the price increase in terms of foreign currency. Producers still desire local price stability, but now through a marginal cost channel rather than the markup channel. Note that this simple representation of price discrimination places no boundaries on the degree of divergence between prices for the same good. Clearly, this is an extreme assumption. More likely, there would exist some neutral band-based upon the level of transport or adjust­ ment costs-within which price differentials would persist. Beyond that range, "gray markets" would emerge as agents found it profitable to circumvent the producer's own distribution channels. However, for small enough demand discrepancies (large enough adjustment costs) this representation is certainly a reasonable assumption. As for notation, the home and foreign destinations are designated as market 1 and 2, respectively. So, for example, the prices set by home producer and foreign producer j for the home market are denoted by i Pl and P}' expressed in units of national currency. Solving the producer's problem, by maximizing real revenues from domestic and export sales minus the utility costs of production (see Appendix), yields the optimal price for each agent for each market destination. Table 2 presents the optimal price-setting rule (ignoring constants) for each market and under both trade patterns from the perspective of the representative home producer. Note that lowercase letters denote loga­ rithms of variables (q =In Q). From Table 2, note that in every case the optimal (log) price p; in units of home currency1u is a weighted average of local money, consumer prices at home, and local producer prices for the industry, with all weights being strictly positive.11

''�Typically. with differential markups, demand is less convex than the constant elasticity case. See Marston ( 1990). 20 Krugman ( 1984) finds that most countries invoice exports in terms of domes­ tic currency when relative country-size differences are not significant. The excep­ tion is developing country exports, which are predominantly invoiced in U.S. dollars. 21 Based on taste parameters, the coefficients in Table 2 are given by = -y -1 9 = I _ + 1 and - E. 1T E.(-y 1) 1T, where both coefficients and their sum arc between (O,J ).

©International Monetary Fund. Not for Redistribution 864 HAMID FARUQEE

Table 2. Optimal Price-Setting Rules Destination Pattern of trade Home market Foreign market lntersectoral trade p,1 ='1Tm+9q+(l-'1T -9)p1 p; ='IT(e +m•) +9q+(l-1T-6)p2

Intraindustry = 'ITm +(I-'IT)q p;= 'IT(e +m*)+ 6q 'IT-6)(e + q*) trade jJ,1 + (1-

Under intersectoral trade, the price rule for the home market places a weight, 'TT, on domestic money, which represents nominal aggregate demand in that market. An increase in aggregate demand raises demand for each individual variety, effecting an increase in price. The weight placed on domestic consumer prices captures a real-income effect. Since each producer is also a consumer-ultimately concerned with utility-an increase in domestic consumer prices translates into a loss of real pur­ chasing power in general and a desire to raise nominal price in response.

The weight placed on industry prices, 1 - 1r- captures a relative-price effect. An increase in industry prices lowers producere, i's relative price, raising the agent's product demand and price. Under intraindustry trade, the price rule for domestic prices has the same general form. Once again, producers are concerned about domestic consumer prices to the extent reflected by Now, however, the relative­ 1- e. price coefficient 1r- capturing the positive interaction between industry prices, is also placede, on the home CPI under two-way trade. Together with the real-income effect, the weight placed on sums to q 1-1r in Table 2. So whereas price setters place a weight 1-1r- on p1 under cross-industry trade, that weight now falls on p and e (sub­ 1, p 1", sumed in q ). Hence, there exists a greater degree of linkage between home and foreign producer prices prevailing in the domestic market. In the terminology of Cooper and John (1988), there is a greater degree of strategic complementarity between home and foreign pricing decisions under intraindustry trade.22 The intuition behind this result is relatively straightforward. Home and foreign varieties are closer substitutes under two-way trade. so that movements in international relative prices have larger consequences for individual demands. Hence, home price setters arc more concerned with foreign prices set for the home market when choosing their own price. The same conclusion holds true for export prices as well. Home export prices under two-way trade display greater influence and interaction with

zz See Faruqce !994) for u the r discussion. ( f r

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 865 foreign local prices, since the relative-price effect also includes foreign industry prices and the nominal exchange rate.

11. Pass-Through and Pricing to Market Using the static framework of the previous section, the effects of various disturbances on nominal prices can be measured given a partic­ ular calibration of the model. These comparative static exercises arc carried out in this section. In particular. the responsiveness of the optimal price to an exchange rate disturbance in all four cases, given that all other variables remain unchanged, is calculated. The first measure of interest is the degree of exchange rate pass­ through. Pass-through is defined as the percentage change in home prices-measured in terms of foreign currency-resulting from a change in the nominal exchange rate. If the foreign-currency price of a good changes one for one with the exchange rate, the degree of pass-through is one. In other words, the exchange rate change is fully reflected in the price change while the home-currency price remains unaffected. At the other extreme, if a price responds fully to offset the change in the exchange rate, leaving the foreign-currency price unaffected, the degree of pass-through is zero. Mathematically. this sensitivity measure is calcu­ lated (in absolute value) by la,o;''/ae - 11 E(O,l), for m= 1.2. Figure 1 displays the degree of pass-through in home prices, holding expenditure shares fixed and allowing the elasticity of substitution to vary; dashed and solid lines represent intraindustry and intersectoral trade, respectively. Note that the pass-through responses shown also apply to foreign producers as well through symmetry 23 When the elastic­ ity of substitution is exactly one, the two types of trade are identical. This result follows from the characterization of international trade subsumed in consumer preferences. Explicitly, preferences are Cobb-Douglas across goods from different industries, while CES preferences apply to goods within an industry. Hence, when countries engage in one-way trade across sectors there exists a low degree of substitutability between home and foreign goods. Meanwhile. with two-way intraindustry trade. the CES specification posits that home and foreign goods are closer substitutes-unless E equals unity, in which case the distinction in utility vanishes.24 Conversely. as E

Pass-through abroad is defined as lap;" 11 for m = 1.2. �3 file+ 1·'Applying L ' H pital 's rule to in verific!> this equivalence resulto in the lim1tpreferences as E-+ described the Appendix 1.

©International Monetary Fund. Not for Redistribution 866 HAMlD FARUQEE

Figure I. Exchange-Rate Pass-Through and rhe Pattern of Trade

Pass- throu gh 1.0 ,---___;;;...______,

0.9

0.8

0.7

0.6 _ _ Export prices _ - - - -- 0.5 ------

2 3 4 5 6 7 E increases, the disparity between the patterns of trade widens in terms of how agents perceive domestic goods compared to foreign goods. This characterization of preferences embodies the basic idea behind the Armington model.25 The separability assumption under the Cobb­ Douglas specification in utility makes the distinction that goods from different industries-appropriately defined-belong to separate com­ modity classes, representing distinct products. At the industry level, each product is then further differentiated into many varieties with a constant elasticity of substitution between them. Several results from this framework can be inferred from Figure L. Naturally, the degree of pass-through is lower in export prices regardless of the trade structure since export demand is always more sensitive to exchange rate fluctuation than is domestic demand. ln contrast, note that in case of the law of one price, pass-through in domestic and export prices must be identical in each case, by definition. Also in Figure 1, as E increases, the degree of pass-through generally declines under two-way trade, but increases under intersectoral trade. The intuition behind the latter result follows from the previous discussion of preferences and trade. As the elasticity of substitution increases. producers display greater strategic complementarity with their industry

25ln Armington goods are imperfect substitutes according to country 1969). · rather than industry.(

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 867 rivals. This stronger linkage in prices translates into greater sensitivity to any changes in relative prices. Under intersectoral trade, as industry varieties become closer substitutes, price setters become increasingly concerned with keeping prices in line with other producers. At that margin, exchange rate fluctuations are of no consequenceresident for relative prices as industry prices are set in the same currency (same unit of account). Hence, producers are more content to allow a greater degree of exchange rate pass-through. As for intraindustry trade, as the elasticity of substitution increases, price setters prefer to keep prices more closely in line with those of industry competitors, as before, but now the industry includes foreign producers in each destination. Consequently, home producers display a greater responsiveness to exchange rate fluctuations by making offsetting changes in their home-currency prices to minimize movements in relative prices. Figure 2 displays the degree of price discrimination under each trade pattern resulting from an exchange rate disturbance. This degree of pricing to market is measured as the percentage deviation between the domestic price and export price for the same good as a percentage of the change in exchange rate.11' Mathematically, this measure is given (in absolute value) by lap}lae - apl!ael. Once again, when E = 1, the two patterns of trade are really the same and the degree of pricing to market is also identical. As the elasticity of substitution increases, producers under intraindustry trade decide to price discriminate across markets to a greater extent. The reasoning follows as before. Stronger price linkages within the industry motivate home price setters to keep price more closely in line with competitors. including foreign producers, in each destination (see again dashed lines in Figure 1 ). However, since demand abroad for the domestic product is more sensitive to the rate of exchange, movements in the value of cur­ rency induce a greater response in prices set for the foreign market than for the home market, increasing the extent of price discrimination. This result is seen in Figure 2 by an increasing degree of pricing to market under two-way trade. As for intersectoral trade, the total absence of an import-competing industry in each country greatly limits producer concerns about interna­ tional relative prices. Furthermore, as E increases. closer ties between industry price setters who happen to be from the same country further reduce the level of price discrimination. This is seen by a decreasing

21' An alternate but related definition of pricing to market used elsewhere when considering multiple markets is the discrepancy between various export prices for a given producer. See, for example, Krugman ( 1987) and Knetter ( 1993).

©International Monetary Fund. Not for Redistribution 868 HAMID FARUQEE

Figure 2. Pricing to Market and the Pa11ernof Trade

Pricing to market 0.5 �------�

0.4 1- - lntraindustry ------0.3 1------0.2� --- [ ------lnrersecroral 0.1 1------:...�-- -

o �------��----���----���----��------��------� I 2 3 4 5 6 7 e degree of pricing to market in Figure 2. As the relative-price effect receives greater weight compared to the aggregate demand and income effects, exchange rate changes are increasingly ignored in both market destinations under intersectoral trade. This finding is clearly at one extreme. From the setup of the model, complete specialization at the industry level characterizes intersectoral trade between two nations that could not be more different. In sharp contrast, complete overlap in industry pro­ duction represents intraindustry trade between identical trading part­ ners. In an intermediate structure between these two boundary cases, each country would produce predominantly in one sector or the other, but would at least have some presence in both. In that instance, the typical degree of pricing to market would likely be increasing in E, lying somewhere in between the limiting cases shown in Figure 2. Interpreted as a boundary range, Figure 2 illustrates the significant variati-on in pricing-to-market behavior associated with various patterns of international trade. ln a multisector setting, the same general result would obtain. In those sectors where the degree of intraindustry trade is high, price setters would display a greater degree of pricing to market and a lower degree of pass-through. Recent empirical evidence documenting the degree of cross-industry variation in pricing-to-market behavior is reported by Knetter (1993). Examining various export prices for Ger­ many, Japan. the United Kingdom. and the United States. Knetter

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 869

(1993) finds that is the critical dimension explaining the variation in the pattern of industrypricing-to-market practices within and across countries.

Ill. Relative Prices and Dynamics

This section extends the model to incorporate dynamics. In the pres­ ence of nominal frictions such as menu costs, producers do not adjust prices continuously. Instead, price setters respond to changing market conditions only intermittently. Moreover. when decentralized producers adjust, they do so at different times because of a lack of coordination. This lack of synchronization in pricing decisions and revisions induces inertia in the overall price level. When price setters move at different times, agents who respond and change their prices do not adjust com­ pletely in an effort to maintain relative prices with those who wait. Consequently, the repercussion of demand shocks may extend long after the initial impulse through this overlapping nature of price adjustment. To capture this overlapping adjustment process, a simple staggered timing structure for price setting is imposed. Specifically, it is assumed that half of all firms, home and fo reign, revise their domestic and export prices in even periods when necessary, while the remaining half adjust in odd periodsY Ta king the frequency and timing of intermittent price changes as given, the comparative dynamics that emerge from this com­ mon staggering structure under contrasting patterns of trade and market segmentation can then be examined. Dropping the subscript and representing price setters by their time location instead, ithe dynamic optimal price rule (without discounting) for a particular market--domestic or export-under both trade patterns is (I) = + il x, O.Sp, O.S,p" where is the actual price28 set for the periods and 1 by home firms x, t t + moving at time 29 Also in equation (1), p, and ,f5,+1 represent the actual and expected optimalt. price at time t and t + 1, where expectations are

-z7See Ball and Cecchetti (1988), and Ball and Romer (1989) for further discus­ sion on the optimal timing and frequency of price adjustment and explanations for the existence of price staggering. 211 Agents are not nsk neutral here. and equation (1) omits a risk premium that is a function of the conditional distribution of all nominal variables. For example. if money. prices, and the exchange rate are log-normal. the risk premium is a constant (comprised of variance and covariance terms) and can be ignored. Alternatively, the dynamics can be interpreted as deviations from a (stochastic) trend reflecting time-varying risk-which has very little impact on relative prices through symmetry across home and foreign price setters. �v In a general sense, one can view equation (I) as the outcome of minimizing

©International Monetary Fund. Not for Redistribution 870 HAMID FA RUQEE formed rationally, based on all the information available at time t. Hence, time-dependent pricing leaves prices predetermined for two periods. Furthermore, the even-odd timing structure introduces asynchronized adjustment similar to that in Taylor (1980). Illustrating the overlapping nature of prices, the index of home producer prices prevailing in the 1 domestic market takes the (approximate) form: p} = 0.5x}_ 1 + 0.5x, • Substituting the static price rules from Table 2 and their corresponding foreign counterparts into the dynamic price-setting equation (1) defines a system of equations with four state variables, consisting of home and foreign local prices and export prices, where each price equation takes the form of a second-order stochastic difference equation as a result of the simple two-period overlapping structure of price adjustment. How­ ever, with the focus on real exchange rate dynamics. the four-variable system can be reduced to one difference equation in relative prices, given the behavior of the three forcing variables consisting of home and foreign 1 money and the nominal exchange rate:1 1 See the Appendix for details. In the case where home and foreign money follow a random walk," the dynamic equation for the CPI-based (log) real exchange rate r under either trade pattern can be written compactly as

(2) Note that the simple even-odd staggering structure in nominal price adjustment gives rise to a simple first-order autoregressive process in relative prices, where the coefficient >.. in equation (2) can be interpreted as a measure of persistence or inertia in the real exchange rate. �1

a quadratic loss function defined by squared deviations in actual price from optimal price over the period for which prices are predetermined. 30The behavior of the nominal exchange rate is identical across both trade patterns. Given money market clearing and balanced trade (see Appendix), the (log) nominal exchange rate is given by * e, = rn, - m, . Comparing this expression to equation (2). which follows. highlights the fact that asset market prices and goods markets prices adjust at differential rates. 31 That is, it is assumed that * * rn, = 1 + v, and m, 1 v, . * = m,* m, + (where v, v are random disturbances), in which case the nominal exchange rate will also follow a random walk. 32 One can show that

1 - 7T + V 28(1 - o:) E(Q>,!). >.. = 1 + V1r + 28(1 - o:) See Appendix.

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 871

As alluded to, the dynamics governing the behavior of the real ex­ change rate in equation (2) apply to both patterns of trade. At first glance, this equivalency result may seem a bit surprising given the com­ parative differences shown earlier in Section 11. However, the findings are mutually consistent. In fact, the differences regarding the response of nominal prices described earlier are what guarantee the equivalence result regarding the behavior of relative prices. For illustration, compare the optimal price rules for the home exporter shown in Table 2. The single difference across trade patterns is found in the relative price component. Under intraindustry trade, the home pro­ ducer is concerned with competing foreign prices, the exchange rate, and other home export prices, whereas that same agent under cross-industry trade is concerned only with other home export prices. Under the law of one price, the added concern with the foreign­ currency price of output would compel price setters in the first instance to reduce the degree of domestic price inertia resulting from menu costs and price staggering. Prices would be less sticky in terms of national currency in order to reduce variation in terms of foreign currency, leading to less variable and persistent deviations in the real exchange rate under two-way tradeY However, as indicated by equation (2), once the assump­ tion of spatial arbitrage is removed, this comparative difference regarding the behavior relative prices no longer holds true. Instead, once export prices detach from domestic prices, goods prices are in effect set in terms of local currency rather than national currency .since price linkages predominantly exist within market destinations rather than across them. As a result, under two-way trade. the destination prices set by industry rivals from different countries arc inertial in the same unit of account, as if they were from the same country, which just happens to be the case of intersectoral trade. Consequently. real ex­ change rate dynamics are the same across differing patterns of trade. ceteris paribus, and depend only upon the timing structure of price adjustment. Overall, pricing-to-market behavior increases the degree of persis­ tence in the real exchange rate compared with the setting of one price. Regardless of the trade pattern, if price setters effectively stabilize price in terms of local currency, destination prices become less responsive to exchange rate disturbances. Pricing to market in effect allows price setters to keep export prices more stable in terms of foreign currency without requiring that domestic prices be more ncxiblc in terms of home currency, unlike the law of one price. Consequently, the real exchange

Faruqcc "Sec ( 1994 ).

©International Monetary Fund. Not for Redistribution 872 HAMID FARUQEE

Table 3. Real Exchange Rate Dynamics Persistence A Openness a E=2 E=6 E= Permanence 10 0.99 0.40 0.51 0 57 0.02 0.95 0.35 0.46 0.52 0.10 0.90 0.29 0.40 0.47 0.20 0.80 0.21 0.33 .40 0.40 0.70 0.14 0.26 00.34 0.60 0.60 0.09 0.21 0.29 0.80

rate displays greater variability and persistence.3� For example. with an import share equal to one quarter, the degree of persistence under intersectoral trade would be twice that under the law of one price, and the corresponding increase under two-way trade would be even larger.J5 Based on equation (2), Table 3 presents a sample of computed persis­ tence values for the real exchange rate across different degrees of open­

a a. ness and substitutability E.. Note that a decrease in representing an increase in expenditure on imports as a share of national income. reflects an increase in openness. The last column measures the long-run impact of a monetary shock on the real exchange rate. This permanence factor measures the change in the steady-state value of the real exchange as a percentage of the change in the equilibrium nominal exchange rate. 311 So. for example, if a = 0.9, a permanent percent increase in the domestic J0

34 In a closed-economy context, Ball and Romer (1990) show that real rigidi­ ties-such as efficiency wages-reinforce the effects of nominal rigidities. induc­ in� a greater degree of persistence in domestic prices. In an open economy, pncing to market provides the source of real rigidity-allowing firms to stabilize relative prices in each market-which increases the degree of stickiness in nom­ inal prices (in terms of local currency) and magnifies the degree of persistence in the real exchange rate. 35 The equivalent solution for intersectoral trade under one price for the degree of inertia is (2a - l)A, compared with A under market se�mentation. In a closed economy (a= 1), the dynamics in the two instances are 1dentical, as one would expect. Meanwhile , when a=0. 75, inertia under one price is half that under pncing to market, and the increase in persistence is even larger for intraindustry trade. See Faruqee (1994). 36The steady-state value (denoted with a bar) for the real exchange rate as seen from equation (2) is given by _ - r=w, +w� - • ) � (m-m , where the coefficient in this expression measures the long-run impact of a penna­ nent monetary shock. See Appendix for details.

©International Monetary Fund. Not for Redistribution PRICJNG TO MARKET 873 money stock (and proportional nominal depreciation) would induce a steady-state real depreciation of 2 percent.-'7 In Table 3, note that persistence declines with increasing openness. The basic reasoning is as follows. As the import share rises, a nominal depreciation leads to greater CPI inflation, and real income is thus affected more by exchange rate changes. Since producers are also con­ sumers concerned with their purchasing power, greater concern with the foreign-currency value of revenue reduces price stickiness in terms of local currency, thereby reducing inertia in the real exchange rate (a(l - a) j �A � ). Table 3 also shows that persistence increases with the elasticity of sub­ stitution between goods. The argument follows similarly to that above, but via the relative-price channel rather than the income channel. A higher degree of substitutability between varieties implies that demand is more sensitive to a change in relative price (see Table 1). As pro­ ducers become more concerned with their price relative to the industry, efforts to stabilize price in terms of local currency in each market inten­ sify. The resulting increase in local price stickiness induces greater overall persistence in tbe real exchange rate. Lastly, Table 3 illustrates that permanence is an increasing function of openness. As import expenditure shares increase, the asymmetric effects of a lasting domestic monetary expansion on prices across segmented markets become larger. More simply, this result can be interpreted in the context of tradability. If the product is essentially not traded (a-?1), the effects of pricing to market on the long-run real exchange rate are small. Conversely, the degree of permanence rises with increasing openness since the effects of pricing to market are more pronounced with a higher degree of international trade. Finally, note that the permanent effects of monetary shocks are a result of market segmentation and not nominal rigidity and would exist even if prices were perfectly flexible (A = 0; see Appendix). Persistence and permanence in relative price responses have important time-series implications for exchange rates. As with earlier sticky-price models, differential rates of adjustment in goods prices versus asset prices allow nominal and real exchange rates to move closely together in the short term. Fol lowing an initial impulse. the real exchange rate then

HThe model thus includes a role for factors in determining equi­ librium real exchange rates. so long as marketsmonetary remain segmented. See Krugman ( 1990) for a recent discussion on real determinants of equilibrium exchange rates.

©International Monetary Fund. Not for Redistribution 874 HAMfD FA RUQEE eventually self-corrects toward equilibrium, reflecting the persistent ef­ fects of monetary disturbances on relative prices. The current framework also predicts that the nominal and real exchange rate move together over the longer run as well,311since monetary shocks have permanent effects on the nominal rate of exchange and on equilibrium relative prices to the extent that goods markets remain segmented. Alternatively, if the law of one price ultimately holds in the long run, the "permanence" factor can be viewed as long-term persistence in the real exchange rate, extending beyond the period required for prices to adjust.

Concluding Remarks IV. With the collapse of Bretton Woods and the abandonment of fixed exchange rates in 1973, the two decades since have seen a great deal of turbulence in foreign exchange markets. Yet the large swings in exchange rates under the era of floating rates have not brought about large move­ ments in inflation or disinflation as one would expect with pass-through. Instead, volatile nominal exchange rates have translated into volatile real exchange rates in the post-Bretton Woods world, while local prices have remained remarkably stable. In an attempt to understand these developments, this paper develops a model of market segmentation and pricing to market, with important macroeconomic consequences for the behavior of international prices broadly in line with the stylized facts. Motivated by a preponderance of the empirical evidence disavowing the law of one price, the model details some of the economic implications of pricing-to-market behavior. The results include both cross-sectional implications regarding nominal prices and time-series implications regarding relative prices. Across different patterns of trade, significant variation exists in the degree of pass-through and pricing to market depending upon the degree of intraindustry trade and the substitutability between home and foreign goods. Across time, dynamic adjustment in prices suggests that nominal and real exchange rates move togeth�r over the short run. as well as over the longer run to the extent that markets remain segmented. Overall, pricing to market provides a potentially important source of local price stickiness and real exchange rate persistence.

·'x Sec Adams and Chadha (1991) for empirical evidence.

©International Monetary Fund. Not for Redistribution PRIClNG TO MARKET 875

APPENDIX

Two-Country Model

The Consumer's Problem For the home country, agent i's utility function is given as follows: � � V; = (�Y( 7��y-� - (Y,1)�- (Y7t - F; - zD,; O1, (Al) where C; is a consumption basket of home and foreign goods, M, represents money holdjngs of home currenc� (no currency substitution), Q is the domestic consumer price index, and Y} , Y; are agent i's level of output for the domestic and export markets. F, denotes fixed costs in the production of a single variety, while z captures the "menu cost" for changing price. Also in equation (Al), D, is a decision dummy variable, equating one if agent changes his or her own price and zero otherwise. Lastly, and represent thei constant expenditure shares JJ. 1 1 - f.L of goods and money, while 'Y- measures the elasticity of marginal disutility with respect to output. In the case of intersectoral trade, agents have CES subutilities over home and foreign varieties of goods, respectively. Explicitly, C; is given as C{ 1 C.= - -- ·-l. (A4) In these last two subutility expressions, c;; and C{, represent agent i's consump­ tion levels of home good j and foreign good j, while € measures the constant elasticity of substitution between any two home or any two foreign varieties, respectively. Consider now the case of intrrundustry trade. Home and foreign goods no longer belong to separate commodity groups as countries exchange goods witrun the same industry. To modify preferences accordingly, equations (A2) through (A4) are replaced with: • � 1 C; (n )111-· 131' I (1 - 13r'· (C�)�! :I ; < 13< ' (AS) • ± cc:;)·� + ± = [ ,-J ,.1 J where 13measures home goods preference. Although agents consider home and fo reign varieties to be of the same product type. residents in each country still prefer local goods. The budget constraint, identical in both instances. completes the formulation of the consumer's problem facing the home agent:

©International Monetary Fund. Not for Redistribution 876 HAMID FARUOEE

, tl 2:P;C;j + 2:EP,*C£ +M,=1, (A6) j-1 ;-• where P, is the price of home good j (in home currency) and Pt is the price of foreign good j (in foreign currency) prevailing in the home market (designated later by a superscript 1), E is the nominal exchange rate (home currency price of foreign currency), and /, is agent i's level of nominal wealth.

Intersectoral Trade Solving the consumer's problem by maximizing (Al), given equations (A2) through (A4), with respect to c;;, C{, , and M, subject to (A6) yields the following individual demand functions for domestic goods, imports and money: , " 111 . c:� = forj=1. . . n; where P= [1 P11 • , (A7) (;p) -•(ocnPfJ-J.) ;,� ] �: • (l'���/ 1- · C�; = ( t ( ,) for j = 1 ...n; where P* = [�,*.Pt -·J' , (A8) and M; = (1 - �J-)1;. (A9)

Intraindustry Trade Again, solving the consumer's problem by maximizing equation (Al) subject to equation (A6), but given equation (AS) instead, one can write the individual demands for each home and foreign variety under two-way trade as

C1; = for = 1 ...n , (AlO) n " (Qpj) -·(�fJ-/Q1 ) 1. and

EPt -• ( 1 - . r _ _ - for - 1 .. . n. (All) nQ Cii ( Q ) ( 13)fJ-/I) 1 Money demand remains unchanged from that under interindustry trade in equa­ tion (A9). Summing up individual demands for each home variety in equation (A 7) or (AlO) over all home consumers along with the equivalent export demands over foreign consumers yields the product demands facing the representative domestic producer as a function of relative price and real wealth at home and abroad.

Th e Producer's Problem Producer i's revenues plus his or her initial money holdings make up the individuars nominal wealth: P,1 Yi + P;Y� +M,. Using this definition of wealth, the indirect utility function/, = (ignoring menu costs) is 1 Y1 1 = PI I + · · - F, M (AI2) u; 0 pZyl �(Y,1r-�cY1 ;)Y - + 0.

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 877

For stability, "marginal cost"-in terms of the marginal disutility of output­

must be non-decreasing, requiring -y- 1 � 0. Hence, scale economies in produc­ tion in the model refer to decreasing average rather than marginal costs. The producer's problem can be stated as maximizing the modified profit function (A12) with respect to each price given demand for output in each market shown in Table 1. The explicit solutions to the producer's problem are shown (in logs) in Table 2.

General Equilibrium The following conditions characterize general equilibrium under both patterns of trade.

Money Market Equilibrium Using the money demand function in equation (A9) and the definition of wealth, domestic money market equilibrium is given by 1 : I = (1 ) I, � (A13) M = M = 2:M, � 2: 2: (P}Y,' + P;Y;), 2: - ( ) = where the total money stock held by home agents equals aggregate money demand and is proportional to nominal GNP. Using this quantity equation relationship at home and similarly abroad, one can derive the demand facing each home producer in both the local and export market shown in Table 1, where � M= .!_ M is equal to GNP per capita. 1 - n -� Goods Market Equilibrium In symmetric equilibrium at the industry level, identical producers set identical prices. Consequently, the following relative prices are unity in general equi­ librium: P}IP1 = Pt/P2 = 1 and correspondingly for foreign producer prices. Adding up product demands in Table 1-given relative producer prices in equi­ librium-yields an income-expenditure equality condition for the home country: (A14) plyt + p2yz = QC, where quantity variables without i subscripts indicate measures summed over all

home agents (e.g., C = LC;). Note that goods market equilibrium in (A14) equates GNP at market prices with aggregate consumption, requiring balanced trade (NX = 0) in the absence of capital mobility. Exchange Rate Equilibrium Given goods and money market clearing and balanced trade ( P1.Y1• = P2Y2 the nominal exchange rate in equilibrium is given by E ), M E = M* . (ALS) The exchange rate adjusts to ensure balance of payments equilibrium. In symmet­ ric equilibrium at the country level, national money supplies are assumed to be equal and, hence, local and export prices are also equal at home and abroad,

©International Monetary Fund. Not for Redistribution 878 HAMID FARUQEE respectively. Thus, the initial symmetric steady-state equilibrium has both the nominal and real exchange rate equal to unity (E = R = 1). Relative Price Dynamics To proceed, note that the consumption-based (log) real exchange rate r-de­ fined simply as the currency-adjusted ratio of CPis in each country�an be written as a weighted measure of relative local and export prices (in logs): 2 r =a (e + r 1) + (1 - a)(p - e - 1 " ). 39 Using this measure of the real ex­ change ratep and- p defining a measurep of domestic-foreign price differentials, • d =a(x 2 - x1) + (1 - a)(x2 - x '"), the difference equation governing the dynam­ ics of nominal price differentials under both patterns of trade can be written as ljlo \jJ1 * * , = ( ) + m, -m,+,m,;1-,m,+ ) +\j/2( -e, -,e,+l) , Al6) d 2 d,_1+,dr+l 2 ( 1 2 ( where

1 -1T-26(1 - a) 21T 4(1 - a)(6(2a - 1)- = ' ljl 1r) ljlo 1 + 1T + 26(1 - ex 1 + 1T + 26(1 -a)' 2 = 1 + 1T + 29(1 - ) ljl,= ex) Note that in equation (A16), the "homogeneity" condition + ljl , + ljl2 = 1 is not satisfied. The implication of this result is that monetary disturbancesljlo will have permanent effects on relative prices (long-run non-neutrality).40 Thus, PPP fails to hold under market segmentation (see below). Solving the second-order difference equation in (2), the (non-explosive) fundamental solution is given by Aljl "' d, = X.d,_1 +-' X.'(,m,�, - ,m,+; + ,rn,*+i�l - ,fnr+i+l) ljlo ;-o2: "' X. jl l ( + 2:7 >..'(-,et+i- ,et+i+ .). Al7) ;�o '+'0 where 1-V1T + 28(1-ex) X.= e(O,l). l + Y1T + 28(1. - ex ) Using the fact that r = e + O.Sd + O.Sd_" the dynamic solution for the real exchange rate is

A\jJJ ; * r, = X.r,- +- L... A ,_,m,+,- ,-Jmr+i- + t-lmt+i* + ) + 1 2 ( 1 - 1 r-lnl,,; \jio (�•=O

' + 'A (,m,�, - ,m,+; ,mit;+ J- ,m,;i+J) + r•Oi ) 311 With two-way trade, this definition serves as a linear approximation. Out of steady state, spending patterns are constant under the approximation. neglecting the (typically second-order) effects of relative price movements on budget shares under mtraindustry trade. See also footnote 15. 40 Real quantities are of course also affected. In general equilibrium. the composition of (log) output is related to the steady-state real exchange rate by r = ex(y' -yh) + (L - a)(y1*-y!).

©International Monetary Fund. Not for Redistribution PRICING TO MARKET 879

+ e, - Aet-1 . (A18) Assuming home and foreign money, and consequently the exchange rate, are martingale processes (i.e., ,m,+, = 0 for i :=:: 0) in equation (A18), one can obtain the specific solution for the real exchange rate shown in (2), where

_ _ 2 1 _ I!J w, - 1-}.._A lji, I!Jo+ and + �ji, lj/2 w2=-}.. l-}.. }.. �· The steady-state or long-run impact of a monetary shock on the real ex­ change rate from equation (2) is measured by (w1 + w2)/( l -A), which equals 1-(ljl, + I!J2)/(1 -ljlo). Hence, long-run PPP (monetary neutrality) is violated unless the homoge­

neity condition = 1 is satisfied. Under the law of one price, this homogeneity conditionI!Jo + ljl, +is I!Jzindeed met (see Faruqee (1994)), but is no longer assured under market segmentation. Instead, a domestic monetary expansion will usually lead to a lasting nominal and real depreciation with pricing to market.41

REFERENCES

Adams, Charles, and Bankim Chadha, "Real and Nominal Exchange Rates in the Long Run," IMF Working Paper 91/63 (Washington: International Monetary Fund, June 1991). Aizenman, Joshua, "Modeling Deviations from Purchasing Power Parity," inter­ national Economic Review, Vol. 25 (February 1984), 175-91. Armington, Paul S., "A Theory of Demand for Products Distinguishedpp. by Place of Production," StaffPapers , International Monetary Fund, Vol. 16 (March 1969), pp. 159-78. Ball, Laurence, and Stephen G. Cecchetti, "Imperfect Information and Stag­ gered Price Setting," American Economic Review, Vol. 78 (December 1988), pp. 999-1018. Ball, Laurence, and David Romer, "The Equilibrium and Optimal Timing of Price Changes," Review of Economic Studies , Vol. 56 (April 1989), pp. 179-98.

41 One can show that ljl<• s; l since 2(cx - I, suggesting that + + s; -y- nominal and real exchange rateslj/1 ljiexhibit2 positive long-run1) comovemcnts from monet ary shocks. In the second condition, the left-hand side is bounded above by zero (local goods preference) while the right-hand side is bounded below by zero (rising marginal costs). Homogeneity obtains in the knife-edge case when both expressions hold as equalities.

©International Monetary Fund. Not for Redistribution 880 HAMID FARUQEE

---, "Real Rigidities and the Non-Neutrality of Money," Review of Economic Studies, Vol. 57 (April 1990), pp. 183-203. Blanchard, Olivier J., "Price Asynchronization and Price-Level Inertia," in Inflation, Debt, and Indexation, ed. by Rudiger Dornbusch and Mario Henrique Simonsen (Cambridge: MIT Press, 1983), pp. 3-24.

--- , and Stanley Fischer, Lectures on Macroeconomics (Cambridge: MIT Press, 1989). Blanchard, Olivier J., and Nobuhiro Kiyotaki, "Monopolistic Competition and the Effects of Aggregate Demand," American Economic Review, Vol. 77 (September 1987), pp. 647-66. Breuer, Janice Boucber, "An Assessment of the Evidence on Purchasing Power Parity," in Estimating Equilibrium Exchange Rates , ed. by John Williamson (Washington: Institute of International Economics, 1994), pp. 245-77. Cooper, Russell, and Andrew John, "Coordinating Coordination Failures in Keynesian Models," Quarterly Journal of Economics, Vol . 103 (August 1988), pp. 441-63. Delgado, Francisco A., "Hysteresis, Menu Costs and Pricing with Random Exchange Rates," Journal of Monetary Economics, Vol. 28 (December 1991), pp. 461-84. Dixit, Avinash, and Joseph E. Stiglitz, "Monopolistic Competition and Opti­ mum Product Diversity," American Economic Review , Vol. 67 (June 1977), pp. 297-308. Dornbusch, Rudiger, "Expectations and Exchange Rate Dynamics," Journalof Political Economy, Vol. 84 (December 1976), pp. 1161-76. ---, "Exchange Rates and Prices," American Economic Review, Vol. 77 (March 1987), pp. 93-106.

--, Exchange Rates and Jnflarion (Cambridge: MIT Press, 1988). Engel, Charles, "Real Exchange Rates and Relative Prices: An Empirical In­ vestigation," Journal of Monetary Economics, Vol . 32 (August 1993), pp. 35-50. Faruqee, Hamid, "Real Exchange Rates and the Pattern of Trade: Comparative Dynamics for North and South" (Ph.D. dissertation; Princeton, New Jersey: Princeton University, 1994; and Journal of International Money and Finance (forthcoming)). Giovannini, Alberto, "Exchange Rates and Traded Goods Prices,'· Journal of Inrernational Economics, Vol. 24 (February 1988), pp. 45-68. Goldstein, Morris, and Mohsin Khan, "Income and Price Effects in Foreign Trade," Chapter 20 in Handbook of InternationalEconomics , Vol. 2, ed. by Ronald W. Jones and Peter B. Kenen (Amsterdam: North-Holland, 1985). Helpman, Elhanan, and Paul R. Krugman, Market Structure and Fo reign Trade: Increasing Returns, Imperfecr Competition, and the International Economy (Cambridge: MIT Press, 1985). Hooper, Peter, and Catherine L. Mann, The U. S. External Deficit: lrs Causes and Persistence, Board of Governors of the Federal Reserve System Interna­ tional Finance Discussion Paper No. 316 (Washington: Federal Reserve Board, November 1987).

©International Monetary Fund. Not for Redistribution PRlCJNG TO MARKET 881

Isard, Peter, "How Far Can We Push the 'Law of One Price'?'' American Economic Review, Vol. 67 (December 1977), pp. 942-48. Knetter, Michael M., "Price Discrimination by U.S. and German Exporters," American Economic Review, Vol. 79 (March 1989), pp. 198-210.

--, "International Comparisons of Pricing-To-Market Behavior," American Economic Review, Vol. 83 (June 1993), pp. 473-86. Krugman, Paul R., "The International Role of the Dollar: Theory and Prospect," in Exchange Rate Theory and Practice, ed. by John F.O. Bilson and Richard C. Marston, NBER Conference Report (Chicago: University of Chicago Press, 1984), pp. 261-78.

-- , "Pricing to Market When the Exchange Rate Changes," in Real­ Financial Linkages Among Open Economies , ed. by Sven W. Arndt and J. David Richardson (Cambridge: MIT Press, 1987), pp. 49-70.

--, Exchange Rate Instability (Cambridge: MIT Press. 1989). --, "Equilibrium Exchange Rates," in lntemational Policy Coordination and Exchange Rate Fluctuations, ed. by William H. Branson, Jacob A. Frenkel, and Morris Goldstein (Chicago: University of Chicago Press, 1990), pp. 159-91. Mann, Catherine, "Prices, Profit Margins, and Exchange Rates," Fe deral Reserve Bulletin , Vol. 72 (June 1986), pp. 366-79. Marston, Richard C., "Pricing to Market in Japanese Manufacturing." Journal of International Economics, Vol. 29 (November 1990), pp. 217-36. Taylor, John B., "Aggregate Dynamics and Staggered Contracts," Journal of Political Economy, Vol. 88 (February 1980), pp. 1-23.

©International Monetary Fund. Not for Redistribution IMF Vol. StaffNo.Papers (December 1995) 42. 4 @ 1995 International Monetary Fund

Shorter Paper and Comment

Unemployment Hysteresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium

REZA MOGHADAM and CAROLINE VAN RlJCKEGHEM*

This paper examines unemployment hysteresis in the Belgian labor market. It estimates models of wage determination using aggregate and firm -level panel data. The conclusions are: (i) the long-term unemployed do not exert a negative impact on wages; and (ii) the incumbent workers, the "insiders, " exercise market power in wage determination, taking greater account of their own interests than those of the unemployed "outsiders; " and (iii) the wage indexation system can cause a downward rigidity in real wages. Recent initiatives, including programs aimed at the long-term unemployed and the young, are app ropriate in view of the existence of insider power. (JEL E65, 130, P21)

HE IDEA OF UNEMPLOYMENT HYSTERESIS can be traced to Phelps T(1972), who argued that economic shocks could have long, lingering effects on the labor market. Accordingly, a rise in unemployment, by itself, could cause a temporary, or even permanent, increase in the non-accelerating inflation rate of unemployment (NAIRU). The history of unemployment can affect the NAIRU in a number of ways. First, if

* Reza Moghadam is an Economist in the European I Department and holds a Ph.D. from Warwick University. Caroline Van Rijckeghem is an Economist in the Research Department and holds a Ph. D. from the University of California at Berkeley. This paper was written while she was in the Fiscal Affairs Depart­ ment. The paper is a condensed version of Moghadam and Van Rijckeghem (1994). The authors would like to thank, without implication, Ke-Young Chu, Paul Masson, and the researchers at the National Bank of Belgium for helpful comments and suggestions.

882 ©International Monetary Fund. Not for Redistribution UNEMPLOYMENT HYSTERESIS IN BELGIUM 883 wage bargainers are primarily concerned with the interests of those who are employed-the "insiders"-rather than those who seek employ­ ment-the "outsiders"-then they may use union power or take advan­ tage of high turnover costs or the uncertainties associated with employing an outsider to maintain high wages in the face of adverse shocks.' A second way in which the history of unemployment can affect the NAIRU is through the composition of the pool of job seekers. According to this hypothesis, those who experience a long period of unemployment may lose their human capital through enforced inactivity and thereby become less attractive to employers. The long-term unemployed may also become less active job seekers if the unemployment benefit system does not provide adequate incentives to seek employment. Finally, past unemployment-associated with a slowdown in economic activity--canaffect the NAJRU because a higher level of unemployment is associated with a lower level of capital formation and, therefore, with a lower rate of growth of the warranted real wage. Consequently, it may take a long period of high unemployment before wage expectations are brought down to a sustainable level.

I. Empirical Evidence

The presence of high and persistent unemployment, low employ­ ment growth, and a high incidence of long-term unemployment suggests that hysteresis theories can contribute to the understanding of the Bel­ gian labor market.2 In order to assess the contribution of these theories, we carried out empirical tests of two theoretical models of wage determi­ nation using both aggregate time series and a panel of firm-level data.

The Data

The aggregate time-series data cover the period 1970:1-1990:4 and were kindly provided by the Belgian National Bank.3 The disaggregate panel data for employment, wages, and sales by enterprise were obtained from a sample of 312 manufacturing firms for the period 1978-84. They were collected by the ASLK, the Belgian savings bank. The data cover

See Lindbeck and Snower (1988), Chapter 10. 21 Mulkay and Van Audenrode (1993) show that the Belgian tabor market is characterized by a low rate of employment growth. 3 The short-term and long-term unemployment variables only exist from 1987:1. The earlier data were interpolated from annual observations.

©International Monetary Fund. Not for Redistribution 884 REZA MOGHADAM and CAROLINE VAN R!JCKEGHEM

19 industries in the 9 provinces. The sample covers 190,000 blue-collar and 79,000 white-collar workers, about one third of those in manufactur­ ing employment. The firm-level data were supplemented by industry level employment, wages, and unemployment data.

Time-Series Results Table 1 presents direct estimates of the Blanchard and Summers ( 1986) insider-outsider wage equation using 84 quarterly observations from aggregate data for the period 1970:1-1990:4. This equation gives the rate of wage inflation, W-t. as a function of price inflation, !:::.p, and employment, laggedw- once and twice. The strong insider-outsider hy­ pothesis-undern, which union members care only about the utility of currently employed members-implies that wage growth depends on the change rather than on the level of employment.4 The estimated coeffi­ cients appear to lend support to the strong insider-outsider theory: em­ ployment lagged both once and twice is significant (although the latter only marginally), and the hypothesis that the coefficients of employment lagged once and twice are equal in magnitude and opposite in sign cannot be rejected. The second equation replaces employment lagged once and twice with the Jagged change in employment; the third equation adds a time trend. The conclusions remain unchanged: in both equations the change in employment is significant. The last two wage equations in Ta ble 1 have the change in the real wage as the dependent variable rather than the change in the nominal wage. The results again appear to support the strong insider-outsider hypothesis. However, the results in Table 1 are not conclusive evidence of the insider-outsider theory because: (1) the pure insider-outsider wage equations do not allow for a rich set of economic variables, particularly outsider variables such as the long-term unemployment rate; (2) the equations in Table 1 are all dynamic with no long-run properties; and (3) the diagnostic tests imply that the equations in Table 1 have poor econometric properties. 5 Before turning to panel data analysis, we explore hysteresis further in the time-series context by estimating an alternative wage equation based on the bargaining model of Nickell (1987). In this model real wages

y - depend on productivity (output per person), n; employer-employee

4This test is strictly valid only if labor demand follows a random walk. 5 All the equations fail the normality test and the Lagrange multiplier test for serial correlation. Equations (3) and (5) also fail Ramsey's test for functional form.

©International Monetary Fund. Not for Redistribution Table 1. insider-Outsider Wage Equation, 1970:1-1990:4 OLS with �w as dependent variable � z Ramsey's Normality [TI n � T*100 XZ(l)• XZ(2)b .,;:s: n_, -z n_, DW Rz LM(4) RESET r �P- 1 07 0.98 -0.74 75 0.55 79.6 0 1. 1. 29.6 1.3 -< (1) (7.95) (2.39) (1.75) :s: [TI 1.57 0. 7 38.9 z (2) 1.21 1.14 4 36.5 2.1 --1 (8.60) (2.55) ::X: Vl-< (3) 0.83 1.08 -0.02 1.94 0.63 20.6 14.6 10.7 --1 (6.15) (2.89) (6.04) [TI �Vl OLS with �(w - p) as dependent variable Vl -0.02 2.5 0.64 -0.43 2.27 0.43 11.4 13.3 z (4) en (2.13) (1.41) (6.13) m r 1.88 5 7.7 (5) 0.84 -0.02 0.30 1 .4 12.1 0 (2.50) (6.02) � Ramsey's RESET test is based on the square of the fitted values. :s: bThea normality test is based on a test of skewness and kurtosis of residuals.

00 00 VI

©International Monetary Fund. Not for Redistribution 886 REZA MOGHADAM and CAROLINE VAN RIJCKEGHEM

Table 2. Long-Run Wage Equations, 1970:1-1990:4" OLS estimates with as dependent variable w -p y-n SUR LUR c -2.96 0.83 Pm0.14 - P 0.01 -0.03 -0.01 0.03 (10.80) (9.90) (2.50) (5.20) (9.04) (2.10) (8.40)

b 0.97 DW = 1.40 DF = -6.39 ADFC= -4.35 lP = this table (the real wage); n (smoothed productivity); and p (the• In import wedge)w- pare in logs, whiley- (the direct tax wedge); (the indirectPm - tax wedge); SUR (the short-term unemployment12 rate); and LUR f3(the long-term unemployment rate) are in percent. b D1ckey-Fuller test of stationarity of the residuals. c Augmented Dickey-Fuller test of stationarity of the residuals.

social security tax and indirect tax wedges (t., t2, t3); the import price wedge, - and unemployment. Here we separate out unemploy­ Pm p; ment into short-run, SUR, and long-run, LUR, unemployment rates to test the impact of hysteresis through the duration of unemployment. We first estimate a long-run level equation and then derive the dynam­ ics. In our preliminary estimates, we typically found that the replacement ratio and the employers' tax wedge were incorrectly signed. Excluding these variables gives the long-run, or eo-integrating, vector reported in Table 2. 6 All the coefficients are correctly signed and have plausible magnitudes. The equation also appears to have satisfactory long-run properties judging by the test statistics. The coefficient on the productivity term is close to unity and the equation indicates that if the import price wedge rises, wage earners would push up the real consumption wage. The interesting coefficients are those of the short-term and long-term unemployment rates. Short­ term unemployment has the expected negative impact on wages. How­ ever, this impact is rather small: if the short-term unemployment rate rose by 1 percentage point, wages would only fall by 1 percent. Furthermore, the impact of the unemployment rate is considerably diminished by the long-term unemployment rate, which has a positive impact on wages.7 This evidence supports the unemployment hysteresis argument based on the duration composition of unemployment.

bWe found that all the variables appearing in Table 2 were nonstationary; in fact, they were all integrated of order one. The long-run equation is estimated using ordinary least squares (OLS). 7 We also tested for cointegration of SUR and LUR. This was rejected.

©International Monetary Fund. Not for Redistribution UNEMPLOYMENT HYSTERESIS IN BELGIUM 887

We now turn to the dynamics of wage determination in Belgium. The existence of indexation suggests that it is proper to model real rather than nominal wages; moreover, indexation also implies that wages are back­ ward rather than forward looking with potentially long lags (Banque Nationale de Belgique (1991)). Also, since indexation and collective bargaining were suspended in 1982 and only fully restored in 1987, it is necessary to investigate how differently wages behaved during this period relative to the rest of the sample period. Ta ble 3 presents our estimated dynamic model for real wages. The equation is estimated using the instrumental variable technique because the change in productivity term is a contemporaneous determinant. In addition to this variable, the model contains Jagged dependent variables; changes in the tax wedge variables; the lagged residual from Table 2, res_,; and a zero-one dummy for the 1982-86 period, D. A number of features of this model are worth noting. The suspen­ sion of indexation during 1982-86 had a significant and negative long­ run impact on real wages and, as expected, the two-yearly bargaining round and the indexation system induce long lags in wage determina­ tion. In contrast to the pure insider-outsider models of Table 1, the dynamic model of Table 3 has very satisfactory diagnostics: tests for the validity of instruments, serial correlation, functional form, normality, and heteroscedasticity give very satisfactory results. When we included the Jagged change in employment on the right­ hand side to test the insider-outsider hypothesis, it was insignificant (t-statistic = 0.7). The same result was obtained when including the

lagged change in the level of unemployment (c-statistic = - 0 . 5) or the lagged change in the unemployment rate (t-statistic = -0.5).

Table 3. Dynamic Wage Equations, 1970:4-1990:4 Instrumental variables estimation

�(w -p) = 0.003 0.16�(w - p)_ + 0.39�(w - p)- 0. 12A(y - n + 3 4 + ) ( 1.9) (4.1) (2.0) - 0.01M3 - 0.003At2 - O.Olres-• - O.OlD (3.2) (2.1) (2.6) (2.6) JP = 0.55 DW 2.21 SEE = 0.007 = Sargan's (6) 3.92 xz = LM (4)) = 2.97 Ramsey's l (1) = 0.89 (x2 Normality l 0.83 Heteroscedasticity l (I)= 2. 76 =

©International Monetary Fund. Not for Redistribution 888 REZA MOGHADAM and CAROLINE VAN RIJCKEGHEM

In summary, the time-series results suggest that there is unemployment hysteresis, caused mainly by the duration composition of unemployment. Furthermore, the negative impact of unemployment on wages is very moderate in Belgium in comparison with other industrial countries. However, the time-series results were not conclusive in assessing the relevance of the insider-outsider model in Belgium. Therefore, it is useful to explore this model further using micro data. In fact, since the insider­ outsider model is based on wage determination at the firm level, cross­ section or panel analysis can lead to a more conclusive test of the hypoth­ esis. In the next section we use micro data and consider the wages paid by an individual firm in view of the insider variables such as firm level productivity and profitability, and outsider variables such as unemploy­ ment, in order to be able to test appropriately for the relevance of this theory to the Belgian labor market.

Firm-Level Results We now attempt to test for the unemployment hysteresis theories by using the firm-level data set described above. These data are from a sample of 312 manufacturing firms for the period 1978-84. The data set provides us with 2,184 observations; however, since there are some gaps in a few industry-level variables, the regressions utilize about 1,520 observations. The firm-level results are presented in Table 4. The first panel wage equation contains mainly lagged variables and is estimated using OLS. The second panel includes current employment, n, and is therefore estimated using the instrumental variables (IV) technique. All the vari­ ables are at the firm level apart from aggregate real wages ( ), and industry-level unemployment rate ( UIND). The other explanatorywagg vari­ ables are the firm's productivity ), measured as sales per worker; (y- n and the firm's real profit (1r).To test the insider-outsider hypothesis, the OLS equation also contains employment lagged once and twice and the IV equation has current and lagged employment. All the variables apart from the industry level unemployment rate, UIND, are in logs. A number of features of the regressions in Table 4 are worth noting. The industry-wide unemployment rate exerts downward pressure on firm-level wages even though we allow the firm's employment to enter the regression. This is at variance with the pure insider view of wage deter­ mination. At the same time, both employment lagged once and twice (or current and lagged employment in the instrumental variable case) as well as the insider variables such as the firm's profits and productivity, are

©International Monetary Fund. Not for Redistribution UNEMPLOYMENT HYSTERESIS LN BELGIUM 889

Table 4. Wages Equation Using Firm Level Data: Observations from 312 Firms over 7 Years (1978-84) OLS

w-p = 0.31 0.91(w -p)_1 0.05(y - n)- 1 0.03'TT_1 + + + (89.5) (2.2) (2.1) - 0.09UJND-I 0.85wagg 0.05n-l - 0.04n-2 + + (4.6) (7.6) (5.9) (5.1)

N 1,520 R2==0 .90 DW= 2.06 SEE = 0.064 Instrumental variables" w-p = 0.32 + 0.91(w -p)-1 + O.OS(y - n)-1 + 0.05'TT-1 (89.3) (2.1) (2.8)

- 0.08UIND-I 0.94wagg 0.06n - 0.05n-1 + + (4.1) (8.2) (2.4) (2.1)

N = 1,520

R2 = 0.90 D =2.08 = 0.064 W SEE "Past values of n and ( w - p) were used as instruments.

significant. In both the OLS and the IV estimates, the hypothesis that employment lagged once and twice (current and Jagged in the IV case) have equal and opposite coefficients cannot be rejected. This implies the existence of some insider power. To test for the hysteresis effect through the duration composition of unemployment, we added the proportion of those who have been out of work for more than a year, LTU, to the equations in Ta ble 4.8 The results of this exercise are presented in Table 5. The long-term unemployment variable is highly significant and positive in both the OLS and the IV estimates. A number of variations on the results presented in Tables 4 and 5 were also attempted. Instead of aggregate wages, we included industry­ level wages. This variable was highly significant and the results re­ mained unchanged but the diagnostics were not as good as those in Tables 4 and 5. When we substituted the aggregate unemployment rate

8The aggregate wage variable, wagg , was left out of these regressions because of collinearity with LTU.

©International Monetary Fund. Not for Redistribution 890 REZA MOGHADAM and CAROLJNE VAN RIJCKEGHEM

Table 5. Wage Equation Using Firm -Level Data: Observations from 312 Firms over 7 Years (1978-84) (including the long-term unemployment variable) OLS w-p = 0.01 0.91(w -p)_1 0.05(y - n) 0.03'TT_1 + + + (89.4) (2.2) (2. 1) - 0.09U/ND_, 0.05n-• - 0.04n- + 0.23LTU. + 2 (4.7) (5.9) (5.1) (7.4)

N = 1,520 R2 = 0.90 DW =2.06 SEE = 0.064 Instrumental Variables" w-p = 0.32 + 0.91(w - p)- • 0.05(y - n)_, 0.04'TT-1 + + (89.2) (2.5) (2.8) - 0.08UIND_ 1 + 0.06n-0.05n_ 1 + 0.27LTU. ( 4.2) (2.4) (2.1) (8.0)

N = 1,520 R2 = 0.90 DW= 2.08 SEE = 0.064 • Past values of n and ( w - p) were used as instruments.

for the industry-level unemployment rate, the results remained identical except for the coefficient of the aggregate unemployment rate, which was much lower ( -0.01, !-statistic = 4.3). Since our data spanned only seven years, we chose to keep the industry-level variable, which provides more variation over the data set. Finally, the long-term unemployment variable was highly significant and positive in every single regression that we estimated regardless of the method of estimation or the other variables included. In short, the firm-level investigation supports the time-series result of hysteresis occurring through the duration composition of unemployment in Belgium. Furthermore, although the results reject the strong insider­ outsider hypothesis, they do indicate the existence of some insider power in wage determination.

Conclusion and Policy Implications 11. The empirical investigation of the previous section is very preliminary and further analysis is necessary for concrete policy implications to emerge. However, the results imply: (1) that the long-term unemployed

©International Monetary Fund. Not for Redistribution UNEMPLOYMENT HYSTERESIS IN BELGIUM 891 do not have a negative impact on wages; and (2) that the incumbent workers, the "insiders," exercise market power in wage determina­ tion. The latter is not surprising given the high degree of unionization and coverage of wage bargaining in Belgium. How should tabor market policy respond to these characteristics in order to effectively generate employment, reduce unemployment, and, more generally, lower the underutilization of tabor? Given the presence of insider power and the large pool of long-term unemployed, market forces by themselves cannot be expected to lead to a quick reduction in unemployment; thus there is a need for government action in reducing unemployment, and particularly long-term unemploy­ ment. This can be achieved in two ways: tightening up the unemployment benefit system for the long-term unemployed in order to encourage job search; and providing targeted training and employment opportunities for the long-term unemployed. The Belgian Government has recently taken a number of measures to this end. The benefit eligibility rules for the part-time or temporarily unemployed have been tightened. Also, in cooperation with the regions and communities, the Government has introduced the plan d'accompagnement, which provides training or em­ ployment opportunities for all the full-time unemployed under the age of 46 after they have been unemployed for ten months. Although these are positive developments, the following further steps could be taken to reduce long-term unemployment:9 (1) In spite of tighter unemployment benefit regulations, benefits are still available for a longer duration in Belgium than in most other indus­ trial countries. In this respect, unemployment compensation also acts as a form oflong-term income support. However, unlike the income support system, MINIMEX, unemployment compensation is not means tested. Also, a large fraction of individuals receiving unemployment benefits are employed part-time, are on career breaks, or have retired early. This imposes a heavy financing burden on unemployment insurance, leading to a high wage wedge. There is a need for more transparency here. This could be achieved by separating the cyclical, short-term, function of unemployment insurance from its permanent and redistributive income support function. (2) The plan d'accompagnement could be extended. For example, the suspension of benefits for those whose duration of unemployment ex­ ceeds twice the regional average could become automatic rather than

9 Reforms along the following lines may have been put in place after the paper was written.

©International Monetary Fund. Not for Redistribution 892 REZA MOGHADAM and CAROLINE VAN RIJCKEGHEM being left to the discretion of regional unemployment offices (unemploy­ ment benefits are financed by the central government but are adminis­ tered locally). Currently, the initiative applies to those who commence their tenth month of unemployment, but it could be extended to all those who have been unemployed for over ten months. Another conclusion of this paper is the need to reduce insider power in order to help employment generation. Lindbeck and Snower (1988) discuss in some depth the policies that could be effective in this respect. They include: (1) reducing hiring and firing costs and relaxing job security regulations; (2) lowering the cost of employing outsiders through, for example, reduced employer payroll taxes for the young and the long-term unemployed;10 (3) enhancing competition in the product market through reducing the barriers to the entry of new firms by, say, providing tax incentives and business startup funds to hire the unemployed; ( 4) voca­ tional training schemes that provide marketable skills to the unemployed, making them more attractive to firms; and finally, (5) profit sharing, which reduces the marginal cost of employment. An additional complicating factor is wage indexation. In Belgium there is a consensus among the social partners in favor of indexation as a means of maintaining stable pay expectations and the competitiveness law is a way of taking action if necessary. 11 The empirical results of Section I demonstrate that the suspension of indexation in 1982, in the wake of the devaluation of the Belgian franc, was effective in reducing the real wage. The recent decision of the Belgian Government-announced in the "global plan"-to exclude petrol, tobacco, and alcohol prices from in­ dexation and to freeze real wages for the period 1995-96 will help in restraining wages and improving competitiveness. With multiyear con­ tracts, indexation prevents an adjustment in real wages during the con­ tract period. Thus, the existence of hysteresis implies that an adverse shock leads to a permanent increase in unemployment. Therefore, it may also be desirable to strengthen the competitiveness law, so that the suspension or modification of indexation becomes automatic in the event of an adverse shock or a loss of competitiveness.

10 Some payroll tax reductions for the young have been put into effect since this pa er was written. R1 Under this law the Government can suspend indexation if competitiveness deteriorates significantly.

©International Monetary Fund. Not for Redistribution UNEMPLOYMENT HYSTERESIS IN BELGIUM 893

REFERENCES

Banque Nationale de Belgique, "Estimation de la Reaction des Salaires aux Variations des Prix la Consommation" Bulletin de la Banque Nationale de Belgique, Vol. 66 (Decembera 1991) pp. 37-42. Blanchard, Olivier Jean, and Lawrence H. Summers, "Hysteresis and the European Unemployment Problem," NBER Macroeconomics Annua/ 1986 (Cambridge: MIT Press, 1986), pp. 15-78.

Engle, Robert F., and C. W.J. Granger, "Cointegration and Error Correction: Representation, Estimation. and Testing," Econometrica, Vol. 55 (March 1987), pp. 251-76. Layard, Richard, and Stephen Nickell, "Unemployment in Britain," Economica, Vol. 53 (Supplement 1986), pp. S121-69. Layard, Richard, Stephen Nickell, and Richard Jackman, Unemployment: Macroeconomic Performance and the Labour Market (Oxford: Oxford Uni­ versity Press, 1991 ). Lindbeck, Assar, and Dennis J. Snower, The Insider-Outsider The01y of Employ­ ment and Unemployment (Cambridge: MIT Press, 1988). Moghadam, Reza, and Simon Wren-Lewis, "Are Wages Forward Looking?" LSE Centre for Labour Economics Discussion Paper No. 375 (London: London School of Economics, January 1990); also published in Oxford Economic Papers, Vol. 46 (July 1994), pp. 403-24. Moghadam, Reza, and Caroline Van Rijckeghem, "Unemployment Hysteresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium," IMF Working Paper 94/150 (Washington: International Monetary Fund , December 1994). Mulkay, Benoit, and Marc A. Van Audenrode, "Creation, destruction d'emplois et chomage: le cas de la Belgique," Economie Prevision, Vol. 2, No. 108 (1993), p. 19-29. & Nickell, Stephen J., "Why is Wage Inflation in Britain So High?" Oxford Bulletin of Economics and Statistics, Vol. 49 (February 1987), pp. 103-28. Phelps, Edmund S., Inflation Policy and Unemployment Theory: The Cost­ Benefit Approach to Monetary Planning (New Yo rk: Norton, 1972).

©International Monetary Fund. Not for Redistribution StaffPapers IMFVol. 42. No. 4 (December International 1995) © 1995 Monetary Fund

German Unification and Asymmetry in the ERM

Comment on Gardner and Perraudin

JEROME HENRY and JENS WEIDMANN*

N AN INTERESTING ARTICLE in Staff Papers , Edward Gardner and IWilliarnR.M. Perraudin (1993)-henceforth GP-examined whether the shock associated with German unification had any impact on the dominant position that Germany may hold within the European ex­ change rate mechanism (ERM). The authors tackled the question by running vector autoregressions (YARs) on daily changes in the one­ month French, German, and American onshore and offshore interest rates and by looking at the derived impulse-response functions and long-run multipliers. This corresponds to an analysis in terms of the Sims (1980) type of causal linkages. In both the preunification (October 1987-December 1989) and postu­ nification periods (January 1990-August 1992), GP rejected asymmetry, defined as unidirectional causality, since "[t]he effect of French innova­ tions on Germany is significant, albeit smaller than the German effect on France." Moreover, the examination of their year-by-year estimation results suggests that Germany handed over its predominant position 1 to France during 1990, never completely regaining it thereafter. This struc­ tural break is also corroborated by Lagrange multiplier tests, which indicate that a big change did occur at the end of 1989.

*At the time the paper was written, Jerome Henry was an Economist at the Macroeconomic Research Unit of the Bank of France . Jens Weidmann is a Research Fellow at the Institute for International Economics, University of Bonn. The authors would like to thank M.J.M. Neumann and the reviewers for helpful comments. The views presented in this paper are strictly those of the authors and should not be interpreted as reflecting the views of the Bank of France. 1 In this context it is essential not to confuse "predominant position" with "asymmetry." To be semantically rigorous, we define predominant position 894

©International Monetary Fund. Not for Redistribution COMMENT ON GARDNER AND PERRAUDIN 895

In this note we attempt to show while there has indeed been a structural break, it appears that the major asymmetric shock that happened with German unification has not weakened Germany's position in the ERM. On the contrary, since German unification the ERM has moved toward an even more asymmetric functioning, namely, German dominance. In addition, we believe that our results are also more in line with the recent European monetary system (EMS) crises. The fact that Germany's lead­ ership role in the ERM has not eroded with German unification, in conjunction with the asymmetric shock of the latter, is definitely a central point in explaining the so-called "collapse" of the European monetary system in July 1993. Using both Granger (1969) and Sims (1980) causality concepts, we confirm that German unification caused a structural break in a system consisting of three interest rates. We use data similar to those of GP, that is, daily one-month offshore rates for the French franc, the U .S. dollar, and the deutsche mark, but we cover a somewhat extended sample, between April 1983 and December 1993.2 Contrary to GP, we derive this result by considering the of interest rates in a cointegrated VAR, which we believe statisticallylevels and economically to be more suitable than a model in first differences. It follows from the Granger representation theorem that, when the variables are nonstationary and cointegrated, an unrestricted VAR in first differences is misspecified because the lagged equilibrium errors are not accounted for. In other words, we allow for cointegration relations when analyzing Granger and Sims causalities. This model leaves room for mean-reversion properties in the system, and at the same time allows an analysis in terms of common trends. In addition, we not only compute the long-run impulse response function but also test for zero restrictions. We thus perform the "neutrality"

(similar to GP) as bilateral causality with the German influence on France being more important than the other way around. Asymmetry designates a unilateral causality from Germany to France, whereas German dominance means that Germany's influence on France is "monopolistic, " that is, we face asymmetry and the absence of causal linkages between France and the rest of the world (here incorporated by the United States). 2The sample period in GP is October 1987-August 1992. The inclusion of the 1993 observations will help us to assess whether Germany still exerts a leadership role in the "new" ERM, allowing for fluctuation bands of ± 15 percent, instead of ±2.25 percent, around the central parities. The widened fluctuation bands could have left more room for maneuver to the other member states' monetary policy. In any case, it should be emphasized that our results on the structural breaks in both cointegration relations and causal linkages are not qualitatively altered when the sample period is shortened so as to correspond to the one analyzed by GP.

©International Monetary Fund. Not for Redistribution 896 JEROME HENRY and JENS WEIDMANN

Figure 1. Normalized Trace

2.2

2.0

1.8

1.6

. 1.4 , = ""' 1 , ,.... '"'"' ./" ""' 1.2 /" ,,... I - --...,__ / \ I \ 1.0 \ '-- ""' - ' - 0.8 - ""'- \ \ 0.619881 4 89/1 89/2 89/3 89/4 90/1 9012 90/3 90/4 9111 9112 9113 91/4

Notes: Let r* stand for the number of cointegration relations. The normalized Trace is defined as the Trace statistic divided by its critical value. The dates along the x-axis are the starting dates for regressions ending in December 1992. analysis in a cointegrated VAR (see Stock and Watson (1989) in the bivariate case and its extension to the trivariate case in Bruneau and Nicolal (1992)). Put differently, we test whether a persistent shock, for instance on the level of the German rate, has a long-lasting impact on the level of the French rate.

I. The Changes in Long-Run Relations

Following Hansen and Johansen (1992), we first perform a recursive analysis in a multivariate setting to test for a structural break in the long-run linkages, namely, the number of cointegration relations. We run a backward recursive estimation of Johansen's (1988 and 1991) Trace and 1 4 Amax statistics.3 Figure shows that the normalized Trace test for the null

3 For the univariate unit-root tests, see Henry and Weidmann ( 1994a and 1994b), where the following results are discussed more extensively and detailed results on all of the causality tests-that is, for both long- and short-run consid­ erations-are presented. Once the existence of some long-run relations is admit­ ted, it is advisable to use full-information maximum likelihood (FIML) estimates of the Johansen (1988 and 1991) type , even though the variables were affected by non-normality (see Gonzalo (1994)). 4 The normalized Trace statistic is defined as the Trace statistic divided by its critical value and should increase with the sample size under the null hypothesis, that is, from right to left in Figure 1. Hence, we reject the null hypothesis if the normalized Trace statistic exceeds unity.

©International Monetary Fund. Not for Redistribution COMMLN.I 0'\i Gi\RDNER AND PERR/\l1DIN 897

Table I. Jolwnsen Te sts for Cointegration in the Gaussian VA R

Statistics p 2 and H0 p = 0 p=l = " (r p) 29.0"** 17.7"'* 4.4 Amu.< = * 51.1*** 22. 1** 4.4 Trace r s p Notes:( The presented) results are derived from the model and DataGenerating Process with a restricted constant and no trend with ten lags. Two asterisks stand for a null hypothesis rejected at a risk of 5 percent. and three asterisks, a risk at 1 percent.

of one cointegration relation drops sharply when observations prior to the third quarter of 1989 are included in the sample, thus bearing out GP's result. Nevertheless, the nature of our structural break differs from the one found by GP as it is equivalent to a shift in the system's degrees of freedom. There is only one cointegration relation before the structural break, whereas thereafter we find two long-run relations.5 One would expect that this also has an impact on the causal linkages at work. In light of the previous findings, we split the sample into two sub­ periods: April 4, 1983 to November 29, 1990, and December 3, 1990 to December 22, 1993-the break point corresponding to the most obvious shock to the German rate that is, a leap occurring in late November 1990. Regarding the second subperiod,, the two stationary components of the system can be identified: they are the U .S. rate and the French-German interest rate d iffe rential See Tables 1 and 2 for details. Indeed the p-value of the likelihood . ratio test statistic for this joint overidentifying restriction is only 11 percent, but much higher values are found when both restrictions are tested separately. Both the stationarity of the U .S. rate in a univariate framework and the stationarity of the French-German interest rate differential in a bivariate model are easily accepted. Given the straightforward interpretation of these two restrictions, we consider them relevant to our analysis.

5 �This r esult stands up whatever lag length between and 25-the limits t corresponding to the "optimal" Jag leng h fou nd by the Schwarz and Akaike not c�tteria, respectivel y- is chosen. The conclusions are affected by the exclu­ i 1986 SIOn of the data before Apr l (the date of the last major realignment)or after 199:1 t Janu.arv he date of t e EMS crisis and ··new" ERM) from the sample. Our ( ha iti between 19R9 1990. d�·tatlcd results suggest a t r ns on and rather than a sudden in line to nc t. m break. which would be with GP. According the latter, a less o - t �· b gi ing 1991, pronoutH.:�·d in !>ystem·:; stntt:ture ot:currcd at t 1e e nn of r t:han c IlL· 1 p uhabl�· � ! tlw break p11int \\'�' Cllth·uiTillg wtth :;dccted.

©International Monetary Fund. Not for Redistribution 898 JEROME HENRY and JENS WEIDMANN

Table 2. Estimation Results with an Identified Normalized Long-Run Model ECM coefficients (percent) Cointegration vectors Variable for constant) (rFRF• ro£M• ruso. 5.1 0.39 13t = (1 -I 0 0.55) ro£MrFRF 0.17 0.15 13z = (0 0 1 -3.77) ruso 0.33 -1.2

Notes: roeM, rFRF, and ruso denote the German, French, and U .S. rate respec­ tively. 13is of dimension 2 x 4, because, aJong with the three rates, one has to take into account the restricted constant that enters the long-run parameters. cxl3' is the ECM component of the VAR dynamics, i.e., in Johansen's notation. Only two long-run parameters are freely estimated (theTI two intercepts in the l3s).

II. The Changes in Long-Run Causal Linkages

We now check whether the changes in the system's long-run linkages coincide with some parallel modification in its causal linkages. The causal linkages can be examined either exclusively-through the error-correc­ tion terms in the cointegrated VAR or through the long-run multipliers in the moving average representation-or together with the effects of changes in the variables of interest (as in Granger (1988), or Toda and Phillips (1993)). Here, we focus on the former approach, and thus on the long-run connections between the levels of Eurorates (see Henry and Weidmann (1994a) for further analyses including the changes in the interest rates). The first significant finding relates to the weak exogeneity properties, that is, the long-run component of the Granger noncausality. According to Johansen's likelihood ratio tests, the U.S. rate is weakly exogenous prior to German unification, but the German rate becomes the sole weakly exogenous variable in the system thereafter (see Table 3 for the results of the weak exogeneity and neutrality tests). The current change in the German rate depends neither on the past level of the U .S. rate nor on the lagged French-German interest rate differential, indicat­ ing that from 1990 on German monetary policy has become more inde­ pendent. This holds for the transitory dynamics as well, since the changes in the German rate are not driven by previous changes in either of the other two rates. (The p-value for the joint hypothesis that the other two rates do not Granger-cause the German one is 35 percent using the Mosconi and Giannini (1992) procedure.) Quite similar results are obtained when a Wald test for neutrality, that is, long-run Sims noncausality, is applied (see Figure 2 for an assessment

©International Monetary Fund. Not for Redistribution COMMENT ON GARDNER AND PERRAUD!N 899

Table 3. Weak Exogeneity and Neutrality Tests Neutrality tests vis-a-vis Weak rmF exogeneity tests 1983-90 1991-93 Variables 1983-90 1991-93 caused by causing caused by causing roEM 1.5 64 7 0 86 0.9 ruso 52 0 49 0.1 irrelevant irrelevant

Notes: roEM , rFRF , and ruso denote the German, French, and U.S. rate respec- tively. All numbers are p-values in percent. The null can be accepted at any risk " rFRF X. percent when the p-value is greater than X.. The test results for the null weakly exogenous" are not presented, as the correspondin& p-values are always close to zero. The weak exogeneity tests are performed w1thjn an unrestricted model. The neutrality tests are computed given the restricted long-run relations (1 before 1990 and 2 thereafter). In the postunification period, the U.S. rate is by definition neutral to the other two: it is I(O) and therefore cannot contain any stochastic trend. of the results for both subperiods).6 For the first subperiod, we reject German dominance but asymmetry seems to prevail, since the innova­ tions in the German rate and the U .S. rate have persistent effects on the

French rate. 7 For the second subperiod, the neutrality test in the trivariate system boils down to a simpler test that can be implemented in a bivariate system involving only the two European rates. Since the U .S. rate ap­ pears to contain no stochastic trend, it cannot contribute to the nonsta­ tionary component of the other two rates, with respect to which it is therefore neutral. This brings us back to the Stock and Watson (1989) analysis of money-output causality. Once the system is orthogonalized, meaning that the instantaneous correlation between the Eurorates is removed, the moving-average representation of the vector error correc­ tion model (VECM) yields a zero long-run response of the German rate

6 Non-neutrality means nonzero terms in the impulse response function derived from the moving-average representation of a statiOnary process. In the bivariate case, without any instantaneous correla6on, it is equivalent to weak exogeneity. One advantage of the neutrality concept. based on the long-run effects, is that it is less likely to be affected by the m1sspecification of volatility, for example, omission of ARCH effects, than, for instance, Granger causality, which passes through the variables' changes, that is, the short-run linkages. 7 The neutrality of the French rate to the German rate, however, depends on the Jag length. When few lags, namely ten, are used, the French rate seems to be non-neutral to the German rate at the 10 percent significance level. This property disappears when the Jag length is increased. Moreover, no instantaneous correlation is found between the French rate's innovations, on the one hand, and the other two rates' innovations, on the other hand. On balance, it is very likely that the two stochastic trends come from the German and U .S. monetary policies. Figure 2 presents the findings for ten lags.

©International Monetary Fund. Not for Redistribution 900 JEROME HENRY and JENS WElDMANN

Figure 2. Non-Zero Long-Run Impulse Responses (Non-Neutrality)

'FRF 'FRF

I I I I I I I � ' 'DEM DEM ruso ruso I (414183- 11129/90) 2 ( 1213/90-12122193) Subperiod Subpcriod

continous (dashed) lines represent r jection neutrality at 5 percent Notes: The e of (10 percent). ' and denote German, French. and American rate, DEM, 'FRF , ruso the respectively.

to an impulse from the French one. In contrast, the French Eurorate is influenced by the orthogonal innovation of the German rate. The same neutrality results are obtained without the 1993 data when resorting a trivariate system.

Concluding Remarks m. Regarding the effect of German unification, we confirm the presence of a structural break occurring around the end of 1989 as already put forward by GP, but with a somewhat different methodology and implica­ tions. As for the causality linkages, the picture drawn by the empirical evidence presented above differs from GP's conclusion. For the preunification period, our findings of an asymmetric function­ ing of the ERM resemble those of GP, though we use a rather 'stronger' definition of asymmetry, since it involves the multipliers between the levels of interest rates. For the postunification period, however, our results indicate that the German leadership was not attenuated by Ger­ man unification as suggested by GP. On the contrary, Germany's dom­ inant position seems to have become even stronger, as the German rate appears quite independent from any other interest rate, including the U.S. one. We could interpret the conjunction of this autonomous German mon-

©International Monetary Fund. Not for Redistribution COMMENT ON GARDNER AND PERRAUDIN 901 etary policy after German unification with German dominance in the ERM as one possible reason for the 1992-1993 EMS crises. Moreover, it is worth noting that the establishment ofthe new ERM does not affect our results, and more particularly, does not result in a weaker German position.

REFERENCES

Bruneau, Catherine, and Jean-Paul Nicolai·, "Persistent Causality in a Multivari­ ate Non Stationary System," CDC Working Paper No. 1992-24!f (Paris: Caisse des Depots et Consignations, August 1992). Gardner, Edward H., and William R.M. Perraudin, "Asymmetry in the ERM: A Case Study of French and German Interest Rates Before and After German Unification," Staff Papers , International Monetary Fund, Vol. (June 1993), pp. 427-50. 40 Gonzalo, Jesus, "Five Alternative Methods of Estimating Long-Run Equilibrium Relationships," Journal of Ecorzometrics , Vol. 60 (January-February 1994), pp. 203-33. Granger, C. W.J., "Investigating Causal Relations by Econometric Models and Cross-Spectral Methods," Econometrica, Vo l. 37 (July 1969), pp. 424-38.

--- , "Some Recent Developments in a Concept of Causality," Journal of Econometrics, Vol . 39 (September-October 1988), pp. 199-211. Hansen, Henrik, and S0ren Johansen, "Recursive Estimation in Cointegrated VAR Models," Institute of Economics Discussion Paper No. 92-13 (Copen­ hagen: University of Copenhagen, 1992). Henry, J erome, and J ens Weidmann (1994a ), "Asymmetry in the EMS Revisited: Evidence from the Causality Analysis of Daily Eurorates," forthcoming in Annates d' Economie et de Statistique; also Banque de France , Note d'Etude et de Recherche, No. 29 (Paris: Banque de France, October 1994). --- (1994b), "The French-German Interest Rate Differential since German Unification: The Impact of the 1992-1993 EMS Crises." University of Bonn, Sonderforschungsbereich 303 Discussion Paper No. B-295 (Bonn: Univer­ sity of Bonn, 1994). Johansen, S0ren, "Statistical Analysis of Cointegration Vectors," Journal of Economic Dynamics and Control, Vol. 12 (June/September 1988), pp. 231-54.

--- , "Estimation and Hypothesis Te sting of Cointegrating Vectors in Gaus­ sian Vector Autoregressive Models," Econometrica, Vol. 59 (November 1991), pp. 1551-80. Mosconi, Rocco, and Carlo Giannini, "Non-Causality in Cointegrated Systems: Representation, Estimation and Testing,·· Oxfo rd Bulletin of Economics and Statistics , Vol . 54 (August 1992), pp. 399-417. Sims, Christopher, "Macroeconomics and Reality,·· Econometrica, Vol. 48 (Jan­ uary 1980), pp. 1-48.

©International Monetary Fund. Not for Redistribution 902 JEROME HENRY and JENS WEIOMANN

Stock, James H., and Mark W. Watson, "Interpreting tbe Evidence on Money­ Income Causality," Journal of Econometrics, Vol. 40 (January 1989), pp. 161-81. Toda, Hiro Y., and Peter C.B. Phillips, "Vector Auto regressions and Causality," Econometrica, Vol. 61 (November 1993), pp. 1367-93.

©International Monetary Fund. Not for Redistribution StaffPapers rMF Vol. 42. No. (December 1995) 4 1995 International Monetary Fund ©

IMF Working Papers

Staff Papers draws on F Working Papers, which are research studies by members of the Fund's/M sta ff . list of Working Papers issued in fo llows. A 1995:3

"Institutional Structure and Labor Market Outcomes: Western Lessons for European Countries in Transition," by Robert J. Flanagan [95/631 "Unemployment Benefits Versus Conditional Negative Income Taxes," by Dennis J. Snower [95/65] "Internal Migration, Center-State Grants and Economic Growth in the States of India," by Paul Cashin and Ratna Sahay [95/661 "Markets for Corporate Debt Securities," by R. Todd Smith [95/67] "Recent Turmoil in Emerging Markets and the Behavior of Country-Fund Discounts: Renewing the Puzzle of the Pricing of Closed-End Mutual Funds," by Charles Kramer and R. Todd Smith (95/68] "The Parallel Market for Foreign Exchange in an Oil Exporting Economy: The Case of Iran. 1978-1990," by Adnan Mazarei [95/69] "Employment and Wages in the Public Sector-A Cross-Country Study," by Aart Kraay and Caroline Van Rijckeghem [95/70] "Recording Interest Income in the Balance of Payments," by Peter Harper [95/71] "Recording Insurance Transactions in the Balance of Payments," by Peter Harper [95/72] "The Spanish Social Security: Prospects of the Pension and Health Care Sys­ tems," by Joaquim V. Levy [95/73] "Four Decades of Fund Arrangements: Macroeconomic Stylized Facts Before the Adjustment Programs," by Julio A. Santaella [951741 "Capacity Constraints. Inflation, and the Transmission Mechanism: Forward­ Looking Versus Myopic Policy Rules," by Peter Clark. Douglas Lax ton, and David Rose [95/75] "Asymmetry in the U.S. Output-Inflation Nexus: Issues and Evidence," by Peter Clark, Douglas Laxton, and David Rose [95/76] "Financial Transactions Taxes." by Parthasarathi Shome and Janet G. Stotsky [95177] "Exchange Market Reform, Inflation. and Fiscal Deficits," by Pierre-Richard Agenor and E. Murat U�er [95/78] "Do Taxes Matter for Long-Run Growth? Harberger's Superneutrality Con­ jecture," by Enrique G. Mendoza, Gian Maria Milesi-Ferretti and Patrick Asca [95/79] Working Paper Summaries. compiled by the Editorial Division [95/80] "The Fundamental Determinants of the Real Exchange Rate of the U .S. Dollar Relative to Other G-7 Currencies," by Jerome L. Stein [95/8 1] 903

©International Monetary Fund. Not for Redistribution 904 WORKING PAPERS

"Price Measurement and Mismeasurement in Central Asia,"' by Vincent Koen (95/82] "The Costs of Taxation and the Marginal Cost of Funds," by Joel Slemrod and Shlomo Yitzhaki [95/83] "The Distributional Effects of Public Expenditure-Update and Overview," by Gerd Schwartz and Teresa Te r-Minassian [95/84] "Are Exchange Rates Excessively Volatile? And What Does 'Excessively Volatile' Mean, Anyway?" by Leonardo Bartolini and Gordon M. Bodnar [95/85] "Inflation and Income Distribution: Further Evidence on Empirical Links," by Ales Bullr and Anne-Marie Guide [95/86] "Labor Market Representation in Quantitative Macroeconomic Models for Developing Countries: An Application to Cote d'Tvoire," by Vincent Bodart and Jean Le Dem (95/87] "Wage Dispersion in the 1980s: Resurrecting the Role of Trade Through the Effects of Durable Employment Changes," by Elaine Buckberg and Alun Thomas (95/88] "The Stability of the Gold Standard and the Evolution of the International Monetary System," by Tamim Bayoumi and Barry Eichengreen (95/89] "Financial Sector Reform in Jamaica During 1985-1992: Possible Lessons for the Caribbean," by David Marston [95/90] "Government Subsidies: Concepts, International Trends, and Reform Options," by Benedict Clements, Rejane Hugounenq, and Gerd Schwartz (95/9 1] "The Macroeconomic Effects of ESAF-Supported Programs: Revisiting Some Methodological Issues," by Louis Dicks-Mireaux, Mauro Mecagni, and Susan Schadler (95/92] "The Identification of Capital Transfers in the Balance of Payments," by Jack Bame [95/93] "The Measurement of Reinvested Earnings in the Balance of Payments," by Jack Bame (95/94] "Human Capital Accumulation and Public Sector Growth," by Vi to Tanzi and Howell H. Zee [95/95] "Dollarization in Transition Economies: Evidence and Policy Implications," by Ratna Sahay and Carlos A. Vegh (95/96] "Growth in East Asia: What We Can and What We Cannot Infer From It," by Michael Sarel [95/98)

An annual subscription to IMF Working Papers is available for US$200.00. It includes 12 monthly shipments and priority mail delivery. Requests should be made to: International Monetary Fund Publication Services Washington, D.C. 2043 1. U.S.A. Te lephone: (202) 623-7430 Te lefax: (202) 623-720 1

©International Monetary Fund. Not for Redistribution INTERNATIONAL MONETARY FUND

STAFF PAPERS

VOLUME 42, 1995

WA SHINGTON, D.C.

©International Monetary Fund. Not for Redistribution IMF Vol. StaffNo. Papers 4 (December 1995) 42, International Monetary Fund © 1995

Vo lume 42 Index

Volume 42 (1995) comprises four issues, as follows: March, pages 1-236 June, pages 237-436 September, pages 437-706 December, pages 707-918

Authors

Alesina, Alberto, and Roberto Perotti. The Political Economy of Budget

Deficits ...... Bayoumi, Tamim, and Barry Eichengreen. Restraining Yourself: The Implications of Fiscal Rules for Economic Stabilization ...... 32 Bayoumi, Tamim, and Ronald MacDonald. Consumption, Income, and International Capital Market Integration...... 552 Bayoumi, Tamim, and Alun Thomas. Relative Prices and Economic Adjustment in the United States and the European Union: A Real Story About EMU...... 108 Blejer, Mario I., Mohsin S. Khan, and Paul R. Masson. Early Contribu- tions of Staff Papers to International Economics...... 707

Cashin, Paul. Government Spending, Taxes, and Economic Growth . . .. 237 Cashin, Paul, and Norman Loayza. Paradise Lost? Growth, Convergence, and Migration in the South Pacific...... 608 Chadha, Bankim. Disequilibrium in the Labor Market in South Africa.. 642 Cottarelli, Carlo, Giovanni Ferri, and Andrea Generale. Bank Lending Rates and Financial Structure in Italy: A Case Study ...... 670 Eichengreen, Barry, and Ta mim Bayoumi. Restraining Yourself: The

Implications of Fiscal Rules for Economic Stabilization ...... 32 El-Erian, Mohamed A., and Manmohan S. Kumar. Emerging Equity Markets in Middle Eastern Countries...... 313 Faruquee, Hamid. Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Perspective ...... 80 Faruquee, Hamid. Pricing to Market and the Real Exchange Rate...... 855 Ferri, Giovanni, Carlo Cottarelli, and Andrea Generalc. Bank Lending Rates and Financial Structure in ltaly: A Case Study ...... 670 Flanagan, Robert J. Wage Structures in the Transition of the Czech Economy ...... 836 906

©International Monetary Fund. Not for Redistribution VOLUME INDEX 42 907

Folkerts-Landau, David, and Peter M. Garber. Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU ...... 134 Garber, Peter M., and David Folkerts-Landau. Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU ...... 134 Garber, Peter M., and Michael G. Spencer. Foreign Exchange Hedging and the Interest Rate Defense ...... 490 Generale, Andrea, Carlo Cottarelli, and Giovanni Ferri . Bank Lending Rates and Financial Structure in Italy: A Case Study ...... 670 Grilli, Vittorio, and Gian Maria Milesi-Ferretti. Economic Effects and Structural Determinants of Capital Controls...... 517 Guitian, Manuel. Conditionality: Past, Present, Future...... 792 Haas, Richard D., and Donald J. Mathieson. Establishing Monetary Control in Financial Systems with Insolvent Institutions...... 184 Haque, Nadeem Ul, and Se-Jik Kim. "Human Capital Flight": Impact of Migration on Income and Growth ...... 577 Henry, Jerome, and Jens Weidmann. Comment on Gardner and Perraudin...... 894 Hossain, Shahabuddin M. The Equity Impact of the Value-Added Tax in Bangladesh ...... 411 James, Harold. The Historical Development of the Principle of Surveil- lance...... 762 Khan, Mohsin S., Mario I. Blejer, and Paul R. Masson. Early Contribu- tions of Staff Papers to International Economics...... 707 Kim, Se-Jik and Nadeem Ul Haque. "Human Capital Flight": Impact of Migration on Income and Growth ...... 577 Kumar, Manmohan S., and Mohamed A. El-Erian. Emerging Equity Markets in Middle Eastern Countries...... 313 . Laxton, Douglas, Guy Meredith, and David Rose. Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implica- tions ...... 344 Levy, Joaquim, and Jonathan D. Ostry. Household Saving in France: Stochastic Income and Financial Deregulation ...... 375 Loayza, Norman, and Paul Cashin. Paradise Lost? Growth, Convergence, and Migration in the South Pacific...... 608 MacDonald, Ronald, and Ta mim Bayoumi. Consumption, Income, and International Capital Market Integration...... 552 MacDonald, Ronald. Long-Run Exchange Rate Modeling: A Survey of the Recent Evidence...... 437 Masson, Paul R., Mario Blejer, and Mohsin S. Khan. Early Contribu- tions of Staff Papers Lto International Economics...... 707 Mathieson, Donald J., and Richard D. Haas. Establishing Monetary Control in Financial Systems with Insolvent Institutions...... 184 Meredith, Guy, Douglas Lax ton, and David Rose. Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implica- tions ...... 344

©International Monetary Fund. Not for Redistribution VOLUME INDEX 908 42

Milesi-Ferretti, Gian Maria, and Vittorio Grilli. Economic Effects and Structural Determinants of Capital Controls ...... 517 Moghadam, Reza, and Caroline Van Rijckeghem. Unemployment Hys­ teresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium...... 882 Moser, Gary G. The Main Determinants of Inflation in Nigeria ...... 270 Ostry, Jonathan D., and Joaquim Levy. Household Saving in France: Stochastic Income and Financial Deregulation ...... 375 Perotti, Roberto, and Alberto Alesina. The Political Economy of Budget Deficits ...... 1 Polak, Jacques J. Fifty Years of Exchange Rate Policy and Research at the International Monetary Fund...... 734 Reinhart, Carmen M. Devaluation, Relative Prices, and International Trade: Evidence from Developing Countries...... 290 Rold6s, Jorge E. Supply-Side Effects of Disinflation Programs ...... 158 Rose , David, Douglas Laxton, and Guy Meredith. Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implications.. 344 Saal, Matthew I., and Lorena M. Zamalloa. Use of Central Bank Credit Auctions in Economies in Transition ...... 202 Sarel, Michael. Demographic Dynamics and the Empirics of Economic Growth ...... 398 Spencer, Michael G., and Peter M. Garber. Foreign Exchange Hedging and the Interest Rate Defense ...... 490 Thomas, Alun, and Tamim Bayoumi. Relative Prices and Economic Ad­ justment in the United States and the European Union: A Real Story About EMU...... 108 To karick, Stephen. External Shocks, the Real Exchange Rate, and Tax Policy...... 49 Van Rijckeghem, Caroline, and Reza Moghadam. Unemployment Hys­ teresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium...... 882 Weidmann, Jens, and Jerome Henry. Comment on Gardner and Per- raudin ...... 894 Zamalloa, Lorena M .. and Matthew I. Saal. Use of Central Bank Credit Auctions in Economies in Transition ...... 202

Titles

Asymmetric Effects of Economic Activity on fnflation: Evidence and Policy Implications. By Douglas Laxton, Guy Meredith, and David Rose...... 344 Bank Lending Rates and Financial Structure in Italy: A Case Study. By Carlo Cottarelli, Giovanni Ferri, and Andrea Generale ...... 670 Comment on Gardner and Perraudin. By Jer6me Henry and Jens Weidmann ...... 894

©International Monetary Fund. Not for Redistribution VOLUME INDEX 42 909

Conditionality: Past, Present, Future. By Manuel Guitian ...... 792 Consumption, Income, and International Capital Market Integration. By Tamim Bayoumi and Ronald MacDonald...... 552 Demographic Dynamics and the Empirics of Economic Growth. By

Michael Sarei ...... 00 00 .... 00 ••00 00 •• 00••••• 00 00.00 00...... 00 ..... 00. 00 ••00... 398 Determining the Value of a Financial Unit of Account Based on Com­ posite Currencies: The Case of the Private ECU. By David Fo lkerts-

Landau and Peter M. Garber 0000 0000 00 . . 00 00 00 00 .. 00 ..00 0000 00 00 ....00 .. 00... 134 Devaluation, Relative Prices, and International Trade: Evidence from

Developing Countries. By Carmen M. Rein hart ... . 00 ...... 00 00 0000 000 0 290 Disequilibrium in the Labor Market in South Africa. By Bankim

Chadha ...... oo ...... oo .... oooo ...... oo .. oo ...... oo.oo...... 642 Early Contributions of Staff Papers to International Economics. By Mario I. Blejer, Mohsin S. Khan, and Paul R. Massonoooooooooooooooo 707 Economic Effects and Structural Determinants of Capital Controls. By

Vittorio Grilli and Gian M aria Milesi-Ferretti 000 ...... 00 000000 00 00 0000. 517 Emerging Equity Markets in Middle Eastern Countries. By Mohamed A. EI-Erian and Manmohan S. Kumaroo 00 .. 00 00 00 .. 00 0000 00 00 0000 .. 0000 0000 00 00 313 The Equity Impact of the Value-Added Tax in Bangladesh. By

Shahabuddin M. Hossain oooooo.... oo .. ooooooooooooooooooooooooooooooooooooooooo 411 Establishing Monetary Control in Financial Systems with Insolvent Insti-

tutions. By Donald J. Mathieson and Richard D. Haas oooo.. oooooo... 184 External Shocks, the Real Exchange Rate, and Tax Policy. By Stephen

To karick 00 ... . 00 ...... 00 00 ...... 00••• 00 00 ••00 .... 00•• 00.00. 00 •• 00..• • .. • • • • • • • 49 Fifty Years of Exchange Rate Policy and Research at the International

Monetary Fund. By Jacques J. Polak oooooooooooooo ... oo ...oo .. oo...... 734 Foreign Exchange Hedging and the Interest Rate Defense . By Peter M.

Garber and Michael G. Spencer 0 0 ... 00 00...... 00 ...... 00...... 490 Government Spending, Taxes, and Economic Growth. By Paul Cashinoo 237 The Historical Development of the Principle of Surveillance. By Harold

James...... oo ...... oo.. oo .. oo...... oo.... oo ...... ooooooooo 762 Household Saving in France: Stochastic Income and Financial Deregula-

tion. By Jonathan D. Os try and Joaquim Levy 00 ...... 000000 0000 .. 00.. 375 "Human Capital Flight": Impact of Migration on Income and Growth. By

Nadeem Ul Haque and Se-Jik Kim .. 00 00 0000 .. 00 ..00 .... 00 ...... 00 00...... 577 Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Per-

spective. By Hamid Faruqee ...... oooooooooooooooooooooo...... oo ... oooo..... 80 Long-Run Exchange Rate Modeling: A Survey of the Recent Evidence.

By Ronald MacDonald ...... 00...... 437

The Main Determinants of Inflation in Nigeria. By Gary G. Moser 00 ... 270 Paradise Lost? Growth, Convergence. and Migration in the South Pacific.

By Paul Cash in and Norman Loayza ...... 00.. 00 .... 00 ....00 608 The Political Economy of Budget Deficits. By Alberta Alesina and

Roberto Perotti . . . 00 .. 00 ...... 00...... 1

Pricing to Market and the Real Exchange Rate. By Hamid Faruqee . . 00. 855 Relative Prices and Economic Adjustment in the United States and the

©International Monetary Fund. Not for Redistribution 910 VOLUME 42 INDEX

European Union: A Real Story About EMU. By Ta mim Bayoumi and Alun Thomas ...... 108 Restraining Yourself: The Implications of Fiscal Rules for Economic Stabilization. By Ta mim Bayoumi and Barry Eichengreen ...... 32 Supply-Side Effects of Disinflation Programs. By Jorge E. Rold6s ...... 158 Unemployment Hysteresis, Wage Determination, and Labor Market Flex­ ibility: The Case of Belgium. By Reza Moghadam and Caroline Van Rijckeghem ...... 882 Use of Central Bank Credit Auctions in Economies in Transition. By Matthew I. Saal and Lorena M. Zamalloa...... 202 Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836

Subjects

To facilitate electronic storage and retrieval of bibliographic data, Staff Papers has adopted the subject classification scheme of the Journal of Economic Literature (Nashville, Tennessee).

Methodology and History of Economic Thought B B20 History of Economic Thought Through 1925-General Early Contributions of Staff Papers to InternationalEconomics. By Mario I. Blejer, Mohsin S. Khan, and Paul R. Masson ...... 707

C Mathematical and Quantitative Methods C51 Model Construction and Estimation Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implications. By Douglas Laxton, Guy Meredith, and David Rose ...... 344 C68 Computable General Equilibrium Models External Shocks, the Real Exchange Rate, and Tax Policy. By Stephen To karick...... 49

E Macroeconomics and Monetary Economics El2 General Aggregate Models-Keynes, Keynesian; Post-Keynesian Pricing to Market and the Real Exchange Rate. By Hamid Faruquee...... 855 E21 Consumption; Saving Consumption, Income, and International Capital Market Integra- tion. By Tamim Bayoumi and Ronald MacDonald ...... 552

©International Monetary Fund. Not for Redistribution VOLUME 42 LNOEX 911

Household Saving in France: Stochastic Income and Financial Deregulation. By Jonathan D. Ostry and Joaquim Levy...... 375 E24 Employment; Unemployment; Wages Disequilibrium in the Labor Market in South Africa. By Bankim Chadha...... 642 E31 Price Level; Inflation; Deflation Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implications. By Douglas Laxton, Guy Meredith, and David Rose ...... 344 The Main Determinants of Inflation in Nigeria. By Gary G. Moser...... 270 E37 Prices, Business Fluctuations, and Cycles-Forecasting and Simu­ lation The Main Determinants of Inflation in Nigeria. By Gary G. Moser...... 270 E42 Monetary Standards and Regimes; Government and the Monetary System Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU. By David Folkerts-Landau and Peter M. Garber ...... 134 E43 Determination of Interest Rates; Term Structure of Interest Rates Bank Lending Rates and Financial Structure in Italy: A Case Study. By Carlo Cottarelli, Giovanni Ferri, and Andrea Generale .. 670 Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU. By David Folkerts-Landau and Peter M. Garber ...... 134 E44 Financial Markets and the Macroeconomy Bank Lending Rates and Financial Structure in Italy: A Case Study. By Carlo Cottarelli, Giovanni Ferri, and Andrea Genera le.. 670 ES Monetary Policy, Central Banking, and the Supply of Money and Credit Conditionality: Past, Present, Future. By Manuel Guitian ...... 792 E52 Monetary Policy (Targets, Instruments. and Effects) Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implications. By Douglas Laxton, Guy Meredith. and David Rose ...... 344

©International Monetary Fund. Not for Redistribution VOLUME INDEX 912 42

Bank Lending Rates and Financial Structure in Italy: A Case Study. By Carlo Cottarelli, Giovanni Ferri, and Andrea Generale .. 670 Establishing Monetary Control in Financial Systems with Insolvent Institutions. By Donald J. Mathieson and Richard D. Haas .. 184 Use of Central Bank Credit Auctions in Economies in Transition. By Matthew I. Saal and Lorena M. Zamalloa ...... 202 E58 Central Banks and Their Policies Bank Lending Rates and Financial Structure in Italy: A Case Study. By Carlo Cottarelli, Giovanni Ferri, and Andrea Generale .. 670 Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU. By David Folkerts-Landau and Peter M. Garber ...... 134 Establishing Monetary Control in Financial Systems with Insolvent Institutions. By Donald J. Mathieson and Richard D. Haas .. 184 Use of Central Bank Credit Auctions in Economies in Transition. By Matthew I. Saal and Lorena M. Zamalloa ...... 202 E61 Policy Objectives; Policy Designs and Consistency; Policy Coordi- nation Conditionality: Past, Present, Future. By Manuel Guitian ...... 792 The Historical Development of the Principle of Surveillance. By Harold James...... 762 E62 Fiscal Policy; Public Expenditures, Investment, and Finance; Taxation Government Spending, Taxes, and Economic Growth. By Paul

Cashin ...... 237 Restraining Yourself: The Implications of Fiscal Rules for Economic Stabilization. By Tamim Bayoumi and Barry Eichengreen ...... 32 E65 Macroeconomic Aspects of Public Finance, Macroeconomic Policy, and General Outlook-Studies of Particular Policy Episodes Unemployment Hysteresis, Wage Determination, and Labor Market Flexibility: The Case of Belgium. By Reza Moghadam and Caroline Van Rijckeghem...... 882

F International Economics Fll Neoclassical Models of Trade Devaluation, Relative Prices. and International Trade: Evidence from Developing Countries. By Carmen M. Rein hart ...... 290

©International Monetary Fund. Not for Redistribution VOLUME 42 INDEX 913

F12 Models of Trade with Imperfect Competition and Scale Economics Pricing to Market and the Real Exchange Rate. By Hamid Faruquee...... 855 F13 Commercial Policy; Protection; Promotion; Trade Negotiations External Shocks, the Real Exchange Rate, and Tax Policy. By Stephen To karick...... 49 F14 Country and Industry Studies of Trade

Devaluation, Relative Prices, and International Trade: Evidence from Developing Countries. By Carmen M. Reinhart ...... 290

F15 Economic Integration Relative Prices and Economic Adjustment in the United States and the European Union: A Real Story About EMU. By Tamim Bayoumi and Alun Thomas ...... 108 F21 International Investment; Long-term Capital Movements Economic Effects and Structural Determinants of Capital Controls. By Vittorio Grilli and Gian M aria Milesi-Ferretti ...... 517

F22 International Migration Paradise Lost? Growth, Convergence, and Migration in the South Pacific. By Paul Cash in and Norman Loayza...... 608

F3 International Finance Foreign Exchange Hedging and the Interest Rate Defense. By Peter M. Garber and Michael G. Spencer ·: ...... 490 F31 Foreign Exchange Fifty Years of Exchange Rate Policy and Research at the Interna- tional Monetary Fund. By Jacques J. Polak ...... 734 Long-Run Exchange Rate Modeling: A Survey of the Recent Evi- dence. By Ronald MacDonald ...... 437 Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Perspective. By Hamid Faruqee...... 80 Determining the Value of a Financial Unit of Account Based on Composite Currencies: The Case of the Private ECU. By David Folkerts-Landau and Peter M. Garber ...... 134 Devaluation. Relative Prices, and International Trade: Evidence from Developing Countries. By Carmen M. Rein hart ...... 290

©International Monetary Fund. Not for Redistribution 914 VOLUME 42 INDEX

F32 Current Account Adjustment; Short-term Capital Movements Devaluation, Relative Prices, and International Trade: Evidence from Developing Countries. By Carmen M. Rein hart...... 290 Economic Effects and Structural Determinants of Capital Controls. By Vittorio Grilli and Gian Maria Milesi-Ferretti ...... 517 F33 International Monetary Arrangements and Institutions Conditionality: Past, Present, Future. By Manuel Guitian ...... 792 Fifty Years of Exchange Rate Policy and Research at the Interna- tional Monetary Fund. By Jacques J. Polak ...... 734 Relative Prices and Economic Adjustment in the United States and the European Union: A Real Story About EMU. By Tamim Bayoumi and Alun Thomas ...... 108 F36 Financial Aspects of Economic Integration Consumption, Income, and International Capital Market Integra- tion. By Tamim Bayoumi and Ronald MacDonald ...... 552 F41 Open Economy Macroeconomics Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Perspective. By Hamid Faruqee...... 80 Pricing to Market and the Real Exchange Rate. By Hamid Faruquee...... 855 Supply-Side Effects of Disinflation Programs. By Jorge E. Rold6s...... 158 F42 Macroeconomic Aspects of International Trade and Finance­ International Policy Coordination The Historical Development of the Principle of Surveillance. By Harold James...... 762

G Financial Economics G14 Information and Market Efficiency; Event Studies Emerging Equity Markets in Middle Eastern Countries. By Mohamed A. EI-Erian and Manmohan S. Kumar ...... 313 G 15 International Financial Markets Emerging Equity Markets in Middle Eastern Countries. By Mohamed A. EI-Erian and Manmohan S. Kumar ...... 313

Public Economics H H20 Taxation and Subsidies-General "Human Capital Flight'': Impact of Migration on Income and Growth. By Nadeem Ul Haque and Se-Jik Kim ...... 577

©International Monetary Fund. Not for Redistribution VOLUME 42 INDEX 915

H22 Taxation and Subsidies-Incidence The Equity Impact of the Value-Added Tax in Bangladesh. By Shahabuddin M. Hossain ...... 411 H23 Taxation and Subsidies-Externalities; Redistributive Effects The Equity Impact of the Value-Added Tax in Bangladesh. By Shahabuddin M. Hossain ...... 411 H6 National Budget, Deficit, and Debt The Political Economy of Budget Deficits. By Alberto Alesina and Roberto Perotti ...... 1 H61 Budget; Budget Systems Restraining Yourself: The Implications of Fiscal Rules for Economic Stabilization. By Tamim Bayoumi and Barry Eichengreen ...... 32 H74 State and Local Borrowing Restraining Yourself: The Implications of Fiscal Rules for Economic Stabilization. By Tamim Bayoumi and Barry Eichengreen ...... 32

Health, Education, and Welfare I 130 Welfare and Poverty-General Unemployment Hysteresis, Wage Determination, and Labor Mar­ ket Flexibility: The Case of Belgium. By Reza Moghadam and Caroline Van Rijckeghem ...... 882

Labor and Demographic Economies J JOO General Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836 Jll Demographic Tre nds and Forecasts Demographic Dynamics and the Empirics of Economic Growth . By Michael Sarel...... 398 J21 Labor Force and Employment, Size, and Structure Demographic Dynamics and the Empirics of Economic Growth. By Michael Sarel ...... 398 J23 Employment Determination; Demand for Labor Disequilibrium in the Labor Market in South Africa. By Bankim Chadha...... 642

©International Monetary Fund. Not for Redistribution 916 VOLUME 42 INDEX

J3 Wages, Compensation, and Labor Costs Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836 J31 Wage Level and Structure; Wage Differentials by Skill, Training, Occupation, etc. Disequilibrium in the Labor Market in South Africa. By Bankim Chadha...... 642 14 Particular Labor Markets Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836 J5 Labor-Management Relations, Trade Unions, and Collective Bargaining Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836

Economic Development, Tec hnological Change, and Growth 0 015 Human Resources; Income Distribution; Migration "Human Capital Flight": Impact of Migration on Income and Growth. By Nadeem Ul Haque and Se-Jik Kim ...... 577 019 International Linkages to Development; Role of International Organizations The Historical Development of the History of Surveillance. By Harold James...... 762 040 Economic Growth and Aggregate Productivity-General "Human Capital Flight": Impact of Migration on Income and Growth. By Nadeem Ul Haque and Se-Jik Kim ...... 577 041 One, Two, and Multisector Growth Models Government Spending, Taxes, and Economic Growth. By Paul

Cashin ...... 237 047 Measurement of Economic Growth; Aggregate Productivity Demographic Dynamics and the Empirics of Economic Growth. By Michael Sarel...... 398 Paradise Lost? Growth , Convergence, and Migration in the South Pacific. By Paul Cashin and Norman Loayza...... 608 056 Economywide Country Studies-Oceania Paradise Lost? Growth, Convergence, and Migration in the South Pacific. By Paul Cash in and Norman Loayza...... 608

©International Monetary Fund. Not for Redistribution VOLUME 917 42 INDEX

Economic Systems P P2 Socialist Systems Wage Structures in the Transition of the Czech Economy. By

Robert J. Flanagan ...... 836 P21 Socialist Systems-Planning, Coordination, and Reform Unemployment Hysteresis, Wage Determination. and Labor Mar- ket Flexibility: The Case of Belgium. By Reza Moghadam and Caroline Van Rijckeghem ...... 882 PS Comparative Economic Systems Wage Structures in the Transition of the Czech Economy. By Robert J. Flanagan ...... 836 P52 Comparative Studies of Particular Economies Emerging Equity Markets in Middle Eastern Countries. By Mohamed A. El-Erian and Manmohan S. Kumar ...... 313

R Urban, Rural, and Regional Economics R11 Analysis of Growth, Development, and Changes Relative Prices and Economic Adjustment in the United States and the European Union: A Real Story About EMU. By Tamim Bayoumi and Alun Thomas ...... 108

©International Monetary Fund. Not for Redistribution In statistical matter throughout this issue,

dots ( ...) indicate that data are not av ailable;

a dash (-) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist;

a single dot (.) indicates decimals;

a comma (,) separates thousands and millions;

"billion" means a thousand million, and "trillion" means a thou­ sand billion;

a short dash (-) is used between years or months (for example, 1992-94 or January-October) to indicate a total of the years or months inclusive of the beginning and ending years or months;

a stroke (/) is used between years (for example, 1993/94) to indicate a fiscal year or a crop year;

a colon (:) is used between a year and the number indicating a quarter within that year (for example, ) 1994: 1 ;

components of tables may not add to totals shown because of rounding.

©International Monetary Fund. Not for Redistribution