Currency Convertibility in the Twentieth Century: The Gold Standard and Beyond (Routledge Explorations in Economic History, 3)

CURRENCY CONVERTIBILITY

The spread of currency convertibility is one of the most dramatic trends of the late twentieth century. It reflects the desire of policymakers to integrate their economies into the global trading system and to attract financial capital and direct investment from abroad. What implications does this have for economic growth and financial development? In this book leading international economists and economic historians look at parallel situations in the history of the international monetary system, focusing in particular on the gold standard. The concluding chapter uses a case study of Portugal to draw out implications for modern international monetary relations in Europe and the rest of the world. Contributors include Michael Bordo, Marcello de Cecco, Albert Fishlow, Alan Milward, Anna Schwartz and Gianni Toniolo. Jorge Braga de Macedo is Professor of Economics, Nova University at Lisbon, Research Associate of the National Bureau of Economic Research and Research Fellow of the Center for Economic Policy Research. As Portugal’s Minister of Finance, he was a signatory of the Maastricht Treaty and chaired the ECOFIN Council from January to June 1992. After leaving the Cabinet in December 1993 he chaired the Committee for European Affairs of the Portuguese Parliament until October 1995. Barry Eichengreen is John L.Simpson Professor of Economics and Professor of Political Science, University of California at Berkeley, Research Associate of the National Bureau of Economic Research, and Research Fellow of the Center for Economic Policy Research. Jaime Reis is Professor of Economic History at the European University Institute in Florence on leave as Senior Research Fellow at the Institute of Social Sciences of the University of Lisbon.

ROUTLEDGE EXPLORATIONS IN ECONOMIC HISTORY

1 Economic Ideas and Government Policy Contributions to Contemporary Economic History Sir Alec Cairncross 2 The Organization of Labour Markets Modernity, Culture and Governance in Germany, Sweden, Britain and Japan Bo Stråth 3 Currency Convertibility The Gold Standard and Beyond Edited by Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis

CURRENCY CONVERTIBILITY The Gold Standard and Beyond

Edited by Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis

London and New York

First published 1996 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 © 1996 Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalogue record for this book has been requested ISBN 0-203-98429-3 Master e-book ISBN

ISBN 0-415-14057-9 (Print Edition)

CONTENTS

List of figures

vii

List of tables

ix

Contributors

x

Preface

xii

Part 1 Overview 1

INTRODUCTION: CURRENCY CONVERTIBILITY IN HISTORICAL PERSPECTIVE—THE GOLD STANDARD AND BEYOND Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis

2

THE OPERATION OF THE SPECIE STANDARD— EVIDENCE FOR CORE AND PERIPHERAL COUNTRIES, 1880–1990 Michael D.Bordo and Anna J.Schwartz

3

11

Part II Myths and realities of the gold standard 3

THE ORIGINS OF THE GOLD STANDARD Alan S.Milward

85

4

SHORT-TERM CAPITAL MOVEMENTS UNDER THE GOLD STANDARD Marcello de Cecco

101

5

THE GEOGRAPHY OF THE GOLD STANDARD Barry EichengreenMarc Flandreau

111

COMMENT Angela Redish

143

COMMENT Albert Fishlow

151

vi

Part III Portuguese currency experience 6

FIRST TO JOIN THE GOLD STANDARD, 1854 Jaime Reis

159

7

LAST TO JOIN THE GOLD STANDARD, 1931 Fernando Teixeira dos Santos

183

8

MONETARY STABILITY, FISCAL DISCIPLINE AND ECONOMIC PERFORMANCE—THE EXPERIENCE OF PORTUGAL SINCE 1854 Eugénia MataNuno Valério

205

COMMENT Pablo Martin-Aceña

229

COMMENT Gianni Toniolo

235

Part IV Implications for Europe in the 1990s 9

CONVERGING TOWARDS A EUROPEAN CURRENCY STANDARD—CONVERTIBILITY AND STABILITY IN THE 1990s AND BEYOND Jorge Braga de Macedo

243

Index

269

FIGURES

2.1 Capital calls on new issues of Argentine securities in London, 1865– 1914 (current dollars) 2.2 Capital calls on new issues of U.S. securities in London, 1865–1914 (current dollars) 2.3 Inflation rate, 1881–1900 2.4 Money growth rate, 1881–1990 2.5 Government deficit as a percentage of nominal GNP, 1881–1990 2.6 Supply and demand shocks: country groups, 1881–1990 5.1 The geography of the gold standard, 1868 5.2 Income, convertibility and monetary standard in 1868 5.3 Sterling exchange rates, 1860–90 5.4 The geography of the gold standard, 1908 5.5 Exchange rate between London and Amsterdam, 1865–90 5.6 Exchange rate between London and Calcutta, 1870–1914 5.7 Exchange rate between London and Vienna, 1870–1914 5.8 Exchange rate between London and St.Petersburg, 1870–1914 5.9 Exchange rate between London and Madrid, 1870–1914 5.10 Reserves of the Bank of Denmark, 1852–85 5.11 Reserves of the Swedish State Bank, 1850–86 5.12 Reserves of the Bank of Norway, 1850–86 5.13 Reserves of the Bank of Belgium, 1870–86 5.14 Reserves of the Netherlands Bank, 1850–86 6.1 Gold-silver exchange rate in Lisbon, 1850–54 7.1 Escudo/sterling monthly exchange rate, 1910–1932 7.2 Government budget deficit as a percentage of GDP 7.3 Government debt to the Bank of Portugal and note issue 7.4 Price level: cost of living in Portugal and in the UK 7.5 Escudo/sterling monthly real exchange rate 7.6 Nominal exchange rate and relative prices: Portugal vs. UK 7.7 Government debt as a percentage of GDP: total and floating 7.8 Government domestic and external floating debt as a percentage of total debt 7.9 Escudo/sterling daily exchange rate May–October 1931 7.10 Escudo/dollar daily exchange rate May–October 1931 7.11 Government debt to the Bank of Portugal and note issue in 1931

40 43 48 54 62 68 114 116 117 120 122 123 123 124 124 126 127 127 128 129 167 185 185 187 187 188 188 190 190 192 193 194

viii

7.12 Bank of Portugal reserves in 1931: gold and foreign assets 7.13 Stock price indices, 1929–1932 7.14 Bank of Portugal reserves in 1931–1934: gold and foreign assets 9.1 Convergence diamond

195 195 196 263

TABLES

2.1 Pre-Bretton Woods specie convertibility and suspensions 2.2 Bretton Woods, 1946–1971, par values, suspensions and par charges 2.3 Descriptive statistics of new issues of securities in London (current and real values) 1865–1914 2.4 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: inflation 2.5 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: money growth 2.6 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: government deficit as a percentage of GNP 2.7 Supply (permanent) and demand (temporary) shocks: 21 countries, 1881–1990 5.1 Monetary systems of the world, 1868 5.2 Monetary systems of the world, 1908 8.1 Public accounts, monetary evolution and economic performance in Portugal 1854–1990: a summary 8.2 Statistical appendix 9.1 Convergence criteria 9.2 Phases of Portuguese currency experience and gradual regime changes towards convertibility 9.3 Relative duration of different currency regimes, 1834–1993 9.4 Convergence in the cohesion states

20 24 42 46 50 52 65 113 118 206 219 249 257 258 261

CONTRIBUTORS

Michael D.Bordo Center for Monetary and Financial History, Department of Economics, Rutgers University, and Visiting Research Economist, International Finance Section, Princeton University. Professor Bordo holds a PhD in Economics from the University of Chicago and taught previously at the University of South Carolina. Marcello de Cecco University of Rome ‘La Sapienza’. Professor de Cecco holds a PhD in Economics from the University of Chicago and taught previously at the European University Institute in Florence and at the University of Siena. Barry Eichengreen Center for European Studies and Departments of Economics and Political Science, University of California at Berkeley. Professor Eichengreen holds a PhD in Economics from Yale University and taught previously at Harvard University. Albert Fishlow Department of Economics, University of California at Berkeley. Professor Fishlow holds a PhD in Economics from Harvard University and taught previously at Yale University. Marc Flandreau Observatoire Français des Conjonctures Economiques. Professor Flandreau holds a PhD in Economics from the University of California at Berkeley and taught previously at Stanford University. Jorge Braga de Macedo Faculty of Economics, Nova University, Lisbon. Professor Braga de Macedo holds a PhD in Economics from Yale University and taught previously at Princeton University. He was also Director of National Economies at the European Commission in Brussels. Pablo Martin-Aceña Alcala University and Fundacion Empresa Publica, Madrid. Professor Martin-Aceña holds a PhD in Economics from the University of Toronto. Eugénia Mata Faculty of Economics, Nova University at Lisbon. Professor Mata holds a PhD in Economics from the Institute of Economics and Management at the Technical University of Lisbon. Alan S.Milward Department of Economic History, London School of Economics and Political Science. Professor Milward taught previously at

xi

Glasgow University and at the Manchester Institute of Science and Technology. Angela Redish Department of Economics, University of British Columbia. Professor Redish holds a PhD in Economics from the University of Western Ontario. Jaime Reis European University Institute in Florence and Institute for Social Science in Lisbon. Professor Reis holds a PhD in Modern History from Oxford University and taught previously at Glasgow and Leicester Universities and at the Faculty of Economics, Nova University, where he was also Dean. Fernando Teixeira dos Santos Faculty of Economics, University of Porto. Professor Santos holds a PhD in Economics from the University of South Carolina. He was appointed Secretary of State for Finance and Treasury in October 1995. Anna J.Schwartz New York office of the National Bureau of Economic Research, Inc. Dr Schwartz holds a PhD in Economics from Columbia University. She was Staff Director of the US Gold Commission in 1982. Gianni Toniolo University of Venice. Professor Toniolo taught previously at the University of California, Berkeley and at Oxford University. Nuno Valério Institute of Economics and Management, Technical University of Lisbon. Professor Valério holds a PhD in Economics from the Technical University.

PREFACE

The spread of currency convertibility is one of the most dramatic economic trends of the late twentieth century. It reflects the desire of policy-makers to integrate their economies into the global trading system and to attract financial capital and direct investment from abroad. The question is what implications it will have for economic growth and financial development. Can developing countries rely on capital inflows to finance their domestic capital requirements, or will international financial markets prove fickle, as they have in earlier periods? Will countries be able to peg their currencies to that of a major trading partner, as Estonia and Argentina are attempting to do, or will they suffer unmanageable speculative pressures, like those experienced by some members of the European Monetary System in 1992–93 and Mexico in 1994? While the countries and circumstances differ, the questions themselves are not new. In a sense, the situation represents a return to an earlier era, when countries were on commodity standards, legal tender was convertible into precious metal on demand, and individuals were free to import and export specie. The parallel encourages us to seek guidance in the history of the international monetary system. The contributors to this book attempt to present a rounded portrait of this historical experience and to draw out its implications for international monetary and financial developments in Europe and in other parts of the world. As in most discussions of the history of convertible currencies, the gold standard is central to their analyses, but the papers collected here portray the gold standard as a more complex and historically specific set of institutions than is typical of the literature. The gold standard, they show, was a short-lived international monetary arrangement that evolved out of the bimetallic and silver standards of the earlier nineteenth century. The price and exchange-rate stability that were its most praiseworthy features did not span the globe; they were largely limited to Western Europe and North America. Only for a short period did the gold standard actually embrace the better part of the world. The decision to join the system’s operation was conditioned by political as well as economic factors. And often the timing of the adoption of the gold standard profoundly shaped a country’s monetary experience for many years.

xiii

The contributors, who include both international economists and economic historians, develop these points using a combination of cross-country comparisons and case studies. Part III of the book, adopting the case study approach, considers the convertibility experience of Portugal, a country with a particularly eventful and informative international monetary history. The final chapter draws out the implications for modern international monetary relations in Europe and the rest of the world. It concludes that the successful introduction of a single European currency standard foreseen in the Maastricht Treaty will require that the multilateral surveillance procedures implemented since then be effective. These procedures may enforce a combination of exchange rate stability and of currency convertibility more lasting than that which prevailed under the gold standard. This will then mean that convergence and cohesion will reinforce each other, and that deepening and widening will be possible in the emerging European economy. On the eve of the revision of the Maastricht Treaty, this implication is as salient for the 15 member-states as it is for the 10 associated states in Central and Eastern Europe. The conference at which first drafts of these papers were presented, at the Convent of Arrabida, on June 3/4, 1994, was hosted by the National Commission for the Celebration of Portuguese Discoveries and sponsored jointly by the Bank of Portugal and the Luso-American Development Foundation. Their support is gratefully acknowledged, together with the assistance of Irene Lemos and Gregoria Soria, at the Bank and at Berkeley respectively. Initially conceived by the Minister of Finance of Portugal as a way to celebrate the 140th anniversary of Portugal joining the gold standard, the conference was attended by his successor and included as one of the contributors the current Treasury Secretary. Reviving the lessons from the past was closer to the design and implementation of convergence policies than anyone expected. Similar proximity may well be discovered in other case studies besides Portugal.

xiv

Part I OVERVIEW

2

1 INTRODUCTION Currency convertibility in historical perspective—the gold standard and beyond Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis

One of the most pronounced international economic trends of the late twentieth century is the adoption of convertible currencies. For many years after World War II, countries maintained restrictions on the freedom of residents to convert domestic currency into foreign exchange, Until the end of the 1950s the industrial economies, aside from the United States and Canada, controlled purchases and sales of foreign currency for purposes related to current account transactions (that is, purchases and sales of goods and services abroad). Developing countries were slower still to restore convertibility on current account. The centrally planned economies, for whom restrictions on convertibility were a corollary of their strategy of suppressing the market economy, never began the process. In recent years this situation has been transformed. Current account convertibility is now an economic fact of life throughout the industrial world. An increasing number of developing countries have relaxed exchange restrictions. Many of the previously-planned economies of Central and Eastern Europe, including several of the successor states of the former Soviet Union, have opted for convertibility after debating its merits relative to those of a payments union. The spread of convertibility reflects the desire of national policy-makers to integrate their economies into the global trading system. It is driven by the effort to attract financial capital and direct investment from abroad. It signals that governments have made significant progress in implementing a stable and sustainable macroeconomic strategy. Many countries have followed the establishment of current account convertibility by relaxing restrictions on international capital flows (by establishing convertibility on capital account, in other words). While the removal of controls on capital account transactions has been selective, in significant parts of the world—Western Europe for example—controls have all but disappeared. Members of the European Union participating in the Exchange Rate Mechanism of the European Monetary System peg their currencies to one another within multilateral bands without any restrictions on international capital flows. An increasing number of developing and transition countries, seeking to encourage the growth of their financial markets and to attract funds from abroad, have relaxed or removed controls on capital inflows and outflows.1

4 OVERVIEW

This trend toward currency convertibility raises a host of questions for economists and policy-makers. Can developing and transition economies rely on foreign capital to meet their financial needs, or will international financial markets prove fickle, as they have in earlier periods? Can the volatility of capital inflows and outflows be managed? Can central banks successfully peg their currencies to those of major trading partners, as countries like Estonia and Argentina are attempting to do, or will they suffer unmanageable speculative pressures, like those experienced by some members of the European Monetary System in 1992–93 and Mexico in 1994?2 Is there a compromise between the extremes of floating exchange rates on the one hand and monetary unification on the other? While the countries and circumstances differ, the questions are not new. In a sense, the current situation represents a return to an earlier era, when countries were on commodity standards, legal tender was convertible on demand, and individuals were free to import and export specie. In the cases of Estonia, Argentina and the small, open EU member states seeking to peg their currencies within the EMS, it resembles nothing so much as the gold standard of the late nineteenth century. The parallel encourages the attempt to seek lessons in the history of the international monetary system. This book attempts to do just that. It seeks to draw out the implications of gold standard history for modern international financial relations, focusing on the small, open economies at the fringes of the gold standard system whose experience resembles that of small, open economies in Southern Europe and other parts of the world today. The contributors are international economists and economic historians. They draw evidence from cross-country comparisons and national case studies. Much of the case study material is taken from the experience of one country, Portugal, with a particularly rich and informative—yet relatively neglected—international monetary history. Often controversy over the advisability of instituting convertibility and pegging the currency revolves around whether these policies can buttress the credibility of policy-makers’ commitment to price stability. Michael Bordo and Anna Schwartz, in Chapter 2, ask whether the gold standard can be understood as a credibility-enhancing rule. They present evidence, from 21 countries between 1880 and 1990, consistent with the notion that the gold standard can be understood as a contingent rule that enhanced the credibility of the commitment to price stability. The rule was contingent in the sense that it was binding under normal circumstances but could be relaxed in times of crisis. It worked smoothly, Bordo and Schwartz conclude, in the industrial countries that were at the core of the gold standard, but not in developing countries at the periphery of the system. But the authors caution that an exchange rate rule was not sufficient for credibility, although it could enhance a domestic commitment to stable policies.3 They emphasize that the effects and effectiveness of the gold standard rule depended on the economic and political context in which it operated. It was credibility-enhancing only when specific political and economic conditions were

THE GOLD STANDARD AND BEYOND 5

in place. Of such conditions, Bordo and Schwartz single out economic maturation and development, which cultivated support for the gold standard by creating a constituency of industrial and commercial interests that stood to benefit from monetary stability. This theme is taken up in Alan Milward’s chapter, which stresses the association of the gold standard with economic development and the constitutional rule of the middle class. The gold standards displacement of alternative monetary arrangements in the second half of the nineteenth century reflected the growing influence of the industrial and commercial classes, in Milward’s view. To assume that the movement to the gold standard reflected efficiency considerations alone is to ignore the over-arching role of politics. Political considerations are prominent also in Marcello de Cecco’s analysis of central banks and international capital flows. De Cecco shows how the success of an exchange rate policy under the gold standard hinged on the elusive factor of market confidence. He underscores the limitations of standard prescriptions of the steps the authorities could take to gain credibility in the markets. He dismisses as simplistic the view that monetary authorities who were officially independent in fact enjoyed freedom from political constraints. Central bank policies under the gold standard, he insists, were shaped from the start by the politics of banking and diplomacy.4 There is an analogy here with recent work in economics emphasizing the political prerequisites for maintenance of an exchange rate peg. Early models of speculative attacks on pegged currencies assumed that instability erupted when a government’s foreign exchange reserves fell to some arbitrary lower level.5 Currency instability in these models was driven by the economic determinants of the balance of payments (relative national inflation rates, budget deficits, productivity growth rates, etc.). Recent models, in contrast, acknowledge that countries can borrow from abroad, rendering the level of reserves irrelevant, and argue that speculative instability is likely to arise when the political costs of defending a currency peg become prohibitive (because, for example, overvaluation aggravates domestic unemployment, fanning opposition to the government’s policies).6 Under the gold standard these politics played themselves out in different ways in countries situated differently in the international system. The standard view, echoed in Bordo and Schwartz, is that the political conditions for a sustainable currency peg were in place in the industrial countries at the core of the gold standard system but not in the developing countries of the periphery. Barry Eichengreen and Marc Flandreau object that the geography of the gold standard is not adequately captured by this core-periphery distinction. They show that the international monetary system of the second half of the nineteenth century is more accurately described in terms of currency blocs. There existed several such blocs, each with a center and periphery, rather than a unified world system with a single core. Monetary arrangements in each bloc depended not just on factors common to countries at comparable stages of economic development but also on

6 OVERVIEW

the particular history and economic characteristics of the countries comprising the bloc. Many readers will detect echoes here of the current European debate over variable geometry. In that context the question is whether the members of the European Union should attempt to construct an evenly integrated economic, financial and political area, adopting a procedure in which the slowest ship is allowed to govern the speed of the entire convoy, or whether different member states should be allowed to proceed at different speeds, forming a series of (perhaps overlapping) commercial and financial blocs. Increasingly, the consensus view is that some form of variable geometry is inevitable because not all countries are prepared to proceed with further integration initiatives, particularly in the monetary domain. But if variable geometry is the answer, it is also a source of questions. For example, should variable geometry be ‘positive’ in the sense that any country that wishes to participate in a particular integration initiative should be entitled to do so, or ‘negative’ in that only countries which satisfy specific preconditions will be allowed to proceed? The gold standard can be seen as precisely a form of positive variable geometry. Not all countries participated. Those which joined, upon satisfying the minimal conditions for currency convertibility, did not have to seek the approval of the incumbent members of the gold standard club. All countries that wished to participate were permitted to do so. There is no evidence that countries which experienced financial difficulties subsequently—the Latin American countries, considered by Bordo and Schwartz, or the Southern European countries like Italy, considered by de Cecco—destabilized the common monetary standard. Positive variable geometry seems to have delivered tolerable results under the gold standard, in other words. Portugal was the first nation to join the United Kingdom on the gold standard in 1854, and the last to restore gold convertibility in 1931. She was the penultimate member state of the European Community to join the EMS in 1992. The papers by Jaime Reis, Fernando Santos and Jorge Braga de Macedo analyze the 1854, 1931 and 1992 decisions. In 1854 Portugal’s decision was a reflexive response to the monetary crisis of the 1840s. In 1931 the decision to join the gold standard had been in place for the better part of a decade; it was part of a grand policy design and was taken despite, not because of, the economic crisis of the 1930s. In both cases the timing of Portugal’s adoption of the gold standard differed from that of other European countries. And in both cases historical contingencies had important implications for the country’s subsequent international monetary experience. In 1854 Portugal adopted gold convertibility on the eve of a period when one country after another joined the gold standard club, reinforcing the advantages of membership and minimizing the strains of adopting the requisite policies. In 1931, when Portugal again joined the gold standard, the international monetary system was about to enter a period of crisis. It is no surprise that Portugal’s subsequent gold standard experience was tumultuous and abbreviated.

THE GOLD STANDARD AND BEYOND 7

The parallel with the 1990s is obvious. In 1992, when Portugal joined the EMS, German unification was placing upward pressure on European interest rates and subjecting the EMS to strains. It follows that Portugal’s experience was stormy and that the escudo naturally attracted the attention of speculators.7 Portugal’s monetary policy decisions were more than financial froth on the surface of deeper economic currents, however; they could themselves shape fundamental economic developments. Eugénia Mata and Nuno Valério, in their overview of 150 years of Portuguese currency experience, highlight the importance of international monetary policy decisions for the country’s economic development. Portugal ‘s suspension of convertibility in 1891, they observe, was costly in that subsequent lack of access to international capital markets implied policies of fiscal austerity and slow growth. The suspension of convertibility in 1931 again discouraged inward foreign investment, although it left a heritage of monetary stability that lasted for more than 40 years.8 As Jorge Braga de Macedo observes in his chapter, Portugal’s entry into the EMS in 1992 was in effect a third attempt to cement the escudo’s convertibility.9 The 1992 decision resembles that of 1854 in that it was designed to deal with specific problems of inflation and adjustment. It resembles that of 1931 insofar as it was part of an over-arching strategy to integrate Portugal into the European economy. The pay-off to that strategy will depend on the success with which the member states of the European Union navigate the final stages of their integration process. A key question for the 1996 Intergovernmental Conference is which member states will participate in the monetary union when Stage III of the Maastricht process commences, presumably at the beginning of 1999, and how the criteria determining whether countries qualify for participation will be applied. It is all but certain that monetary arrangements in Stage III will involve variable geometry: not all EU member states may wish to participate in the monetary union, and not all member states will be judged ready.10 But will the treaty’s ‘convergence criteria’ determining which countries qualify be interpreted in exclusionary fashion, resulting in a monetary union limited to the ‘hard core’ of Deutschmark-area countries and making it difficult for other countries to join subsequently? Or will the EU embrace a form of ‘positive variable geometry’ in which any state wishing to participate in the monetary union and satisfying the convergence criteria in a broad sense will be allowed to participate?11 If positive variable geometry was feasible under the gold standard, should the same not also be true in Europe today? Some observers would be skeptical that the analogy holds. Under the gold standard it was possible—if not always easy— for countries to unilaterally peg their currencies to gold. There was no question of the overriding commitment to the maintenance of the gold standard peg.12 There was no fully articulated theory linking interest rates to the state of the economy and hence little pressure for the government or the central bank to direct its monetary instruments to other ends. The extent of the franchise being limited, there was little pressure from workers suffering unemployment for the

8 OVERVIEW

authorities to stimulate output and employment rather than to pursue currency stability. In the more politicized environment in which monetary policy is made in the 1990s, questions inevitably arise about the credibility of a central bank’s commitment to the defense of its exchange rate target. This leads the markets to test governments’ resolve. Exchange rate pegs are therefore more fragile and prone to crisis. The liquidity of the markets and the scope for adverse speculation make it difficult for countries to defend those pegs unilaterally. A country attempting to stabilize its exchange rate will therefore require foreign support.13 Consequently, the pursuit of international monetary stability increasingly raises problems of collective action. Countries may want to reserve the right to decide which currencies they are obliged to support. Inflation-averse governments will not wish to be obligated to defend the currencies of inflation-prone, profligate partners. This has been the argument against positive variable geometry in the European context. The European Union has responded by building safeguards into the institutions of the European Monetary System. Multilateral surveillance of monetary and fiscal policies by the Committee of Central Bank Governors is designed to contain free-rider problems. In the run-up to monetary union, surveillance is buttressed by the ‘Excessive Deficit Procedure’ of the Maastricht Treaty. Such procedures are unlike anything witnessed during the gold standard years. While problems of collective action also arose under the gold standard and were met by international cooperation, cooperative arrangements were never institutionalized to the same extent as in Europe today.14 Thus, while there are parallels between the international monetary problems of small, open economies in the late nineteenth century and those of countries in an analogous position today, there are also important differences. The latenineteenth century gold standard was a ‘spontaneous order’ that arose unplanned out of the decisions of a large number of individual countries.15 The same can be said of the floating rate ‘non-system’ that emerged in the 1970s following the breakdown of Bretton Woods. The Maastricht process seeking to establish a European monetary union is an entirely different animal. For small, open economies like Portugal that must meet the convergence criteria but can also call on the support of their EU partner states, this is a source of both challenge and opportunity. NOTES 1 The Czech Republic is a good example of a Central European country whose currency is well on the road to full convertibility. 2 One can cite other examples as well, such as the devaluation of the currencies of the CFA franc zone against the French franc at the end of 1994. 3 Angela Redish, in her comment, suggests that a contingent rule may be difficult for investors to distinguish in practice from purely discretionary policy. Indeed, the

THE GOLD STANDARD AND BEYOND 9

4

5 6 7

8

9

10 11

12

13

14 15

EMS crisis of 1992–93 gave rise to renewed skepticism about the viability of such rules and about the notion that credibility can be imported from abroad rather than earned at home. In Macedo (1995), this objective is related to domestic stabilities. In his comment, Albert Fishlow concurs, but questions the generality of Italian experience. He suggests that de Cecco may exaggerate the relative importance of short- versus long-term capital movements as a result of his focus on the Italian case. This is the assumption of the seminal Krugman (1979) model. See, for example, Ozkan and Sutherland (1994). The point applies to Europe as a whole: the cyclical conjuncture can have important implications for the attempt to complete the transition to monetary union. It is no coincidence that the 1992 EMS crisis came in the throes of a European recession which made the defense of fixed exchange rates more burdensome. That exchange rates settled down and the outlook for European monetary unification brightened with the resumption of growth should again come as no surprise. Gianni Toniolo and Pablo Martin Aceña suggest the need to quantify this account of Portuguese currency experience in various respects. Both stress the limitations of the historical data upon which the authors’ analysis is based. Both underline the overriding importance of political factors in the country’s decision to join the gold standard in 1854 and 1931, stressed by Reis and Santos respectively. The author chaired the Council of the European Community’s finance ministers (Ecofin) when the escudo entered the EMS in 1992. He persuaded the central bank to dismantle exchange controls and led the Treasury to establish borrowing benchmarks in the international capital market. These and other measures are further discussed in Macedo (1996). Greece, for example, will not yet have a record of participation in the European Monetary System and therefore will not be judged ready. The committee for European Affairs of the Portuguese Parliament, chaired by Braga de Macedo after leaving the ministry in December 1993, unanimously adopted a resolution in favor of positive variable geometry in the period prior to the 1996 Intergovernmental Conference. De Cecco’s chapter provides a reminder that this position should not be overdrawn. The point here, however, is a comparative one vis-à-vis the 1990s. There seems no question that the priority attached to the defense of the currency peg was higher in the late nineteenth century than today, for reasons detailed below. Thus, Argentina required extensive support from the International Monetary Fund in the early months of 1995 to defend its peg to the dollar in the face of fall-out from the Mexican crisis. In the European Monetary System the provision of such support has been institutionalized through the operation of the Short-Term and Very-Short-Term Financing facilities. The role of international cooperation in management of the classic gold standard is stressed by Eichengreen (1992). The term is from Gallarotti (1995).

10 OVERVIEW

REFERENCES Eichengreen, Barry (1992), Golden Fetters: The Gold Standard and the Great Depression, 1919–1939, New York: Oxford University Press. Gallarotti, Giulio M. (1995), The Anatomy of an International Monetary Order: The Classical Gold Standard, 1880–1914, New York: Oxford University Press. Krugman, Paul (1979), ‘A Model of Balance of Payments Crises,’ Journal of Money, Credit and Banking 11, pp.311–325. Macedo, Jorge Braga (1996), ‘Portugal and European Monetary Union: Selling Stability at Home, Earning Credibility Abroad,’ in Francisco Torrey (ed.) Monetary Reform in Europe, Lisbon: Catholic University of Portugal. Ozkan, F.Gulcin and Alan Sutherland (1994), ‘A Model of the ERM Crisis,’ CEPR Working Paper No. 879 (January).

2 THE OPERATION OF THE SPECIE STANDARD Evidence for core and peripheral countries, 1880–1990* Michael D.Bordo and Anna J.Schwartz

INTRODUCTION The classical gold standard era from 1880 to 1914, when most countries of the world defined their currencies in terms of a fixed weight (which is equivalent to a fixed price) of gold and hence adhered to a fixed exchange rate standard, has been regarded by many observers as a most admirable monetary regime. They find that its benefits include long-run price level stability and predictability, stable and low long-run interest rates, stable exchange rates (McKinnon, 1988), and hence that it facilitated a massive flow of capital from the advanced countries of Europe to the world’s developing countries. Others have taken a less favorable view of the gold standard’s performance. Some criticize the record of relatively high real output and short-term price variability (Bordo, 1981; Cooper, 1982; Meltzer and Robinson, 1989), and some have faulted it for subordinating domestic stability to the maintenance of external convertibility (Keynes, 1930). A persistent critique of the gold standard is that it provided a favorable experience for the core countries (France, Germany, the United Kingdom and the United States), but a less favorable experience for the peripheral countries of the developing world (de Cecco, 1974). For the core countries the balance of payments adjustment mechanism was stable, so few crises occurred; the peripheral countries, by contrast, were subject to shocks imported under fixed exchange rates from abroad and frequently suffered exchange rate crises and a destabilized growth pattern. An alternative approach to these issues of gold standard history posits that adherence to the fixed price of specie, which characterized all convertible metallic regimes including the gold standard, served as a credible commitment mechanism to monetary and fiscal policies that otherwise would be time inconsistent (Bordo and Kydland, 1996; Giovannini, 1993). On this basis, adherence to the specie standard rule enabled many countries to avoid the problems of high inflation and stagflation that have troubled the late twentieth century. The specie standard that prevailed before 1914 was a contingent rule, or a rule with escape clauses. Under the rule, specie

12 OVERVIEW

convertibility could be suspended in the event of a well understood, exogenously produced emergency, such as a war, on the understanding that after the emergency had safely passed, convertibility would be restored at the original parity. Market agents would regard successful adherence as evidence of a credible commitment and would allow the authorities access to seigniorage and bond finance at favorable terms. On this view, the core countries were good players of the classical gold standard game—they adhered strictly to the rule, whereas many peripheral countries were not. Some never adhered to the rule. Others joined it when conditions were favorable to them, ostensibly to obtain access to capital from the core countries, but they quickly abandoned it when economic conditions deteriorated. The interwar gold standard can be regarded as an extension of the pre-1914 system because it was based on gold convertibility. However, it was less successful because the commitment to convertibility was often subordinated to other politically induced objectives. The Bretton Woods international monetary system can be regarded as a distant relative of the classical gold standard in that the center country, the United States, maintained gold convertibility. It was also based on a rule with an escape clause—parities could be changed in the event of a fundamental disequilibrium. However, it differed from the basic specie standard rule in that a credible commitment to the fixed parity was not of such primary importance. This chapter surveys the history of the specie standard as a contingent rule from the early nineteenth century, when most countries were still on a bimetallic or silver standard until the gold standards final collapse in the late 1930s. As a comparison we also briefly consider the Bretton Woods system and the recent managed floating regime. We then present some evidence on the economic performance of the core and a number of peripheral countries under the various regimes. This chapter first defines the contingent specie standard rule and discusses how the commitment to convertibility was maintained. It then presents and discusses a chronology of adherence to the rule by 21 countries under variants of the specie standard prevailing before World War II. A similar chronology is included for the Bretton Woods system. There is then offered some graphical evidence on capital flows from 1865 to 1914 from the United Kingdom to two countries of recent settlement—Argentina and the United States—during episodes of both suspension and adherence to convertibility. It suggests that adherence to the rule may have had some influence on the decision by British investors to invest abroad. Evidence is then presented on economic performance for 21 countries (both core and peripheral) across several regimes. Such evidence may shed light on whether differing economic performance can explain why some countries successfully adhered to the rule and others did not, or whether adherence/ nonadherence may have influenced performance. We examine the stability of both

OPERATION OF THE SPECIE STANDARD 13

nominal and real macro variables across four regimes (pre-1914 gold standard; interwar gold standard; Bretton Woods; the subsequent managed exchange rate float). We then present measures of both demand shocks (reflecting policy actions specific to the regime) and supply shocks (reflecting shocks to the environment independent of the regime). These measures allow us to determine whether adherents to the rule consistently pursued different policy actions from non-adherents, and whether persistent adverse shocks to the environment may, for some countries, have precluded adherence to the rule. The chapter concludes with answers to the questions: how successful was the specie standard rule as a contingent rule; why it was successful when it was; and why some countries adhered while others did not. THE SPECIE STANDARD AS A CONTINGENT RULE The domestic specie standard A specie standard has been traditionally viewed as a form of monetary rule or constraint over monetary policy actions. Following a rule, such as adherence to the specie standard, which would cause the money supply to vary automatically with the balance of payments, was viewed as superior to entrusting policy to the discretion of well-meaning and possibly well-informed monetary authorities (Simons, 1951).1 In contrast to the traditional view, which stresses both impersonality and automaticity, the recent literature on the time inconsistency of optimal government policy regards a rule as a credible commitment mechanism binding policy actions over time. The absence of a credible commitment mechanism leads governments, in pursuing stabilization policies, to produce an inflationary outcome (Kydland and Prescott, 1977; Barro and Gordon, 1983). In a closed economy environment, once the monetary authority has announced a given rate of monetary growth, which the public expects it to validate, the authority then has an incentive to create a monetary surprise to either reduce unemployment or capture seigniorage revenue. The public, with rational expectations, will come to anticipate the authorities’ perfidy, leading to an inflationary equilibrium. A credible commitment mechanism, by preventing the government from cheating, can preserve long-run price stability. Following a rule also allows a government to use debt to smooth distortionary taxes over time. In addition to choosing optimal taxes the government can also choose an optimal default rate on its outstanding debt. In a commitment regime the government can force itself to honor its outstanding debt and not default via inflation or suspension of payments. If the government cannot follow a binding commitment—in other words, if it follows a discretionary regime—rational bond-holders would expect the government to have an incentive to default on its

14 OVERVIEW

outstanding debt. Hence, in a discretionary equilibrium, bond-holders will be averse to purchasing government debt. Under the specie standard, the pledge to fix the price of a country’s currency in terms of gold, silver, or both, represents the basic rule. This involved, for instance in the case of a monometallic gold standard, defining a gold coin as a fixed weight of gold, called for example, one dollar. The monetary authority was then committed to keep the mint price of gold fixed through the purchase and sale of gold in unlimited amounts. Under the bimetallic system which prevailed in most countries until the third quarter of the nineteenth century, the monetary authorities would define the weight of both gold and silver coins, freely buying and selling them. Maintaining a fixed bimetallic ratio is a variant of the basic convertibility rule, since it is the fixed value of the unit of account that is the essence of the rule.2 The specie standard rule followed in the century before World War I can be viewed as a form of contingent rule or rule with escape clauses (Grossman and Van Huyck, 1988; DeKock and Grilli, 1989; Flood and Isard, 1989; Bordo and Kydland, 1996). The monetary authority maintains the standard—keeping the price of the currency in terms of specie fixed—except in the event of a well understood emergency such as a major war. In wartime it may suspend gold convertibility and issue paper money to finance its expenditures, and it can sell debt issues in terms of the nominal value of its currency, on the understanding that debt will eventually be paid off in specie. The rule is contingent in the sense that the public understands that the suspension will only last for the duration of the wartime emergency plus some period of adjustment. It assumes that afterwards the government will follow the deflationary policies necessary to resume payments at the original parity.3 Following such a rule will allow the government to smooth its revenue from different sources of finance: taxation, borrowing, and seigniorage (Lucas and Stokey, 1983; Mankiw, 1987).4 As we document later in this chapter, the gold standard contingent rule worked successfully for three core countries of the classical gold standard: Britain, France, and the United States. In all these countries the monetary authorities adhered faithfully to the fixed price of gold except during major wars. During the Napoleonic War and World War I for England, the Civil War for the United States, and the Franco-Prussian War for France, specie payments were suspended and paper money and debt were issued. But in each case, after the wartime emergency had passed, policies leading to resumption were adopted.5 Indeed, successful adherence to the rule may have enabled the belligerents to obtain access to debt finance more easily in subsequent wars.6 Examples of discretion—breaches of the rule—include postponement of resumption after the war and reasonable delay period have passed, and pegging to specie at a devalued parity. Under both situations, should there be another war within memory of the previous one, then the public’s willingness to absorb government debt would be quite different from that in the previous war, even if

OPERATION OF THE SPECIE STANDARD 15

the situation were otherwise similar and the government claimed to subscribe to a reasonable delay rule. It is crucial that the rule be transparent and simple and that only a limited number of contingencies be included. Transparency and simplicity would avoid the problems of moral hazard and incomplete information (Canzoneri, 1985; Obstfeld, 1992), i.e. prevent monetary authorities from engaging in discretionary policy under the guise of following the contingent rule. In this respect a second contingency—a temporary suspension in the face of a financial crisis, which in turn was not the result of the monetary authorities’ own actions, may also have been part of the rule. However, because of the greater difficulty of verifying the source of the contingency than in the case of war—invoking the escape clause under conditions of financial crisis, or in the case of a shock to the terms of trade (a third possible contingency) would be more likely to create suspicion that discretion was being followed. The specie standard rule may have been enforced by reputational considerations. Long-run adherence to the rule was based on the historical evolution of the standard itself, thus for example, gold was accepted as money because of its intrinsic value and desirable properties. Paper claims, developed to economize on the scarce resources tied up in a commodity money, became acceptable only because they were convertible into gold. In turn, the reputation of the specie standard would constrain the monetary authorities from breaching convertibility, except under well-understood contingencies. Thus, when an emergency occurred, the abandonment of the standard would be viewed by all to be a temporary event since, from their experience, only specie or specie-backed claims truly served as money. An alternative commitment mechanism was to guarantee gold convertibility in the constitution. This was the case for example in Sweden before 1914, when laws pertaining to the gold standard could be changed only by two identical parliamentary decisions with an election in between (Jonung, 1984, p. 368). Convertibility was also enshrined in the laws of a number of gold standard central banks (Giovannini, 1993). The international gold standard The specie standard rule originally evolved as a domestic commitment mechanism but its enduring fame is as an international rule. The classical gold standard emerged as a true international standard by 1880 following the switch by the majority of countries from bimetallism, silver monometallism, and paper to gold as the basis of their currencies (Eichengreen 1985). As an international standard, the key rule was maintenance of gold convertibility at the established par. Maintenance of a fixed price of gold by its adherents in turn ensured fixed exchange rates. The fixed price of domestic currency in terms of gold provided a nominal anchor to the international monetary system.

16 OVERVIEW

Recent evidence suggests that, indeed, exchange rates throughout the 1880 to 1914 period were characterized by a high degree of fixity in the principal countries. Although exchange rates frequently deviated from par, violations of the gold points were rare (Officer, 1986), as were devaluations (Eichengreen, 1985). According to the game theoretic literature, for an international monetary arrangement to be effective both between countries and within them, a time consistent credible commitment mechanism is required (Canzoneri and Henderson, 1991). Adherence to the gold convertibility rule provided such a mechanism. Indeed, Giovannini (1993) finds the variation of both exchange rates and short-term interest rates within the limits set by the gold points in the 1899– 1909 period consistent with market agents’ expectations of a credible commitment by the ‘core’ countries to the gold-standard rule in the sense of this paper.7 In addition to the reputation of the domestic gold standard and constitutional provisions which ensured domestic commitment, adherence to the international gold-standard rule may have been enforced by other mechanisms. These include: improved access to international capital markets; the operation of the rules of the game; the hegemonic power of the United Kingdom; and central bank cooperation. Support for the international gold standard likely grew because it provided improved access to the international capital markets of the core countries. Countries were eager to adhere to the standard because they believed that gold convertibility would be a signal to creditors of sound government finance and the future ability to service debt.8 This was the case both for developing countries seeking access to long-term capital, such as Austria-Hungary (Yeager, 1984) and Latin America (Fishlow, 1989), and for countries seeking short-term loans, such as Japan, which financed the Russo-Japanese war of 1905–06 with foreign loans seven years after joining the gold standard (Hayashi, 1989). Once on the gold standard, these countries feared the consequences of suspension (Eichengreen, 1992a, p.19; Fishlow 1987, 1989). The fact that England, the most successful country of the nineteenth century was on the gold standard, as well as other ‘progressive’ countries, was probably a powerful argument for joining (Friedman, 1990; Gallarotti, 1993). The operation of the ‘rules of the game,’ whereby the monetary authorities were supposed to alter the discount rate to speed up the adjustment to a change in external balance, may also have been an important part the commitment mechanism played under the international gold-standard rule. To the extent the ‘rules’ were followed and adjustment facilitated, the commitment to convertibility was strengthened and conditions conducive to abandonment were lessened. Evidence on the operation of the ‘rules of the game’ questions their validity. Bloomfield (1959), in a classic study, showed that, with the principal exception of the United Kingdom, the rules were frequently violated, in the sense that discount rates were not always changed in the required direction (or by sufficient

OPERATION OF THE SPECIE STANDARD 17

amounts) and in the sense that changes in domestic credit were often negatively correlated with changes in gold reserves. In addition, a number of countries used gold devices—practices to prevent gold outflows. One can reconcile the violation of the ‘rules of the game’ and the use of gold devices, with maintenance of credibility in the commitment to gold, by viewing the gold points as a form of target zone (Eichengreen, 1994). Belief that intervention would occur at the upper and lower gold points created a honeymoon effect whereby stabilizing capital flows caused the market exchange rate to revert towards parity before reaching the gold points (Krugman, 1991). Within the zone, the monetary authorities could alter discount rates to effect domestic objectives such as stabilizing real activity and smoothing interest rates (Svennson, 1994).9 Moreover, for the major countries, at least before 1914, such policies were not used extensively enough to threaten the convertibility into gold (Schwartz, 1984). An additional enforcement mechanism for the international gold-standard rule may have been the hegemonic power of England, the most important goldstandard country (Eichengreen, 1989). A persistent theme in the literature on the international gold standard is that the classical gold standard of 1880 to 1914 was a British-managed standard (Bordo, 1984). Because London was the center for the worlds principal gold, commodities, and capital markets, because of the extensive outstanding sterling-denominated assets, and because many countries used sterling as an international reserve currency (as a substitute for gold), it is argued that the Bank of England, by manipulating its bank rate, could attract whatever gold it needed and, furthermore, that other central banks would adjust their discount rates accordingly. Thus, the Bank of England could exert a powerful influence on the money supplies and price levels of other gold-standard countries. The evidence suggests that the Bank did have some influence on other European central banks (Lindert, 1969). Eichengreen (1987) treats the Bank of England as engaged in a leadership role in a Stackelberg strategic game with other central banks as followers. The other central banks accepted a passive role because of the benefits to them of using sterling as a reserve asset. According to this interpretation, the gold-standard rule may have been enforced by the Bank of England. Thus, the monetary authorities of many countries may have been constrained from following independent discretionary policies that would have threatened adherence to the gold-standard rule. Indeed, according to Giovannini (1989), the gold standard was an asymmetric system. England was the center country. It used its monetary policy (bank rate) to maintain gold convertibility. Other countries accepted the dictates of fixed parities and allowed their money supplies to passively respond. His regressions support this view—the French and German central banks adapted their domestic policies to external conditions, whereas the British did not. The benefits to England as leader of the gold standard—from seigniorage earned on foreign-held sterling balances, from returns to financial institutions

18 OVERVIEW

generated by its central position in the gold standard and from access to international capital markets in wartime—were substantial enough to make the costs of not following the rule extremely high. Finally, Eichengreen (1992b) argues that episodic central bank cooperation may have also strengthened the credibility of the gold standard. Lines of credit arranged between the Banque de France, other central banks and the Bank of England during incipient financial crises such as those of 1890 and 1907 may in turn have encouraged private stabilizing capital movements to offset threats to convertibility. The classical gold standard, the gold exchange standard and Bretton Woods Eichengreen (1994) posits three prerequisites for a successful international monetary arrangement: the capacity to undertake relative price adjustment; adherence to robust monetary rules; and ability to contain market pressures. According to him the classical gold standard contingent rule satisfied these criteria for the core countries because the credible commitment to maintain convertibility above all else allowed the escape clause to accommodate major shocks, and because central bank cooperation eased market pressures in the face of speculative attacks. By contrast, for peripheral countries, the credibility of commitment to the gold standard was considerably weaker, reflecting strong domestic political pressures to alter exchange rates (Frieden, 1993). Though gold convertibility was restored by 1926 by most countries, the interwar gold exchange standard was a much less successful application of the specie standard rule. The escape clause could not be invoked (lest it lead to destabilizing capital outflows), absent a credible commitment to maintain gold parity in the face of a politicized money supply process and, according to Eichengreen (1992b), the failure of cooperation. The Bretton Woods international monetary system can also be viewed within the context of the specie standard rule, although it is a distant variant of the original specie standard. Under the rules of Bretton Woods, only the United States, as central reserve country and provider of the nominal anchor, was required to peg its currency to gold; the other members were required to peg their currencies to the dollar (McKinnon, 1993). They also were encouraged to use domestic stabilization policy to offset temporary disturbances. The Bretton Woods system had an escape clause for its members—a change in parity was allowed in the face of a fundamental disequilibrium, which could encompass the contingencies under the specie standard rule—but it was not the same as under the specie standard because it did not require restoration of the original parity.10 The rule for members (other than the United States) was enforced, as under the gold standard, by access to U.S. capital and to the IMF’s resources. For the United States, there was no explicit enforcement mechanism other than

OPERATION OF THE SPECIE STANDARD 19

reputation and the commitment to gold convertibility. Capital controls were viewed as a method to contain market pressures. The system was successful as long as the United States maintained its commitment to convertibility (i.e. maintained price stability). But the escape clause mechanism quickly proved defective since the fundamental disequilibrium contingency was never spelled out and hence parity changes would be accompanied by speculative attacks which became more serious as capital controls became increasingly ineffective. Ultimately, by following highly expansionary monetary and fiscal policies beginning in the mid- 1960s, the United States attached greater importance to domestic concerns than to its role as the center of the international monetary system, and the system collapsed. Thus, although the Bretton Woods system can be interpreted as one based on rules, the system did not provide a credible commitment mechanism.11 The United States was unwilling to subsume domestic considerations to the responsibility of maintaining a nominal anchor. At the same time other Group of Seven (G-7) countries became increasingly unwilling to follow the dictates of the U.S.-imposed world inflation rate. CHRONOLOGY OF ADHERENCE TO AND SUSPENSION OF SPECIE RULES, PRE-AND POST-BRETTON WOODS Conforming to specie rules Tables 2.1 and 2.2 give a snapshot record of the conformity of 21 countries to specie rules—commitment to a fixed parity with an escape clause. Table 2.1 refers to the extended period from the nineteenth century through the post-World War I interwar years, when specie rules were acknowledged, whether or not observed; Table 2.2 refers to the Bretton Woods era, when specie rules were no longer acknowledged or observed (to a limited extent the United States was an exception) but countries submitted to the rule that only in the case of fundamental disequilibrium was a change in parity permissible. Countries did not lightly change the par values of their currencies. The countries are divided into two main groupings: four core countries and 17 peripheral countries. Over the extended period covered by Table 2.1, core countries were faithful to specie rules under the classical gold standard from 1880 to 1914, but not invariably so in the decades before and after. The peripheral countries, which are classified in the main according to geographical location, were intermittently faithful over the extended period. Table 2.1 separates experience under a bimetallic or silver standard, which prevailed before the last quarter of the nineteenth century, from experience under the gold standard that followed. For each standard and each country the table shows dates when a commitment was made to convert the national currency into specie, dates of suspension of the commitment, and the reasons for suspension. For

Table 2.1 Pre-Bretton Woods specie convertibility and suspensions

20 OVERVIEW

OPERATION OF THE SPECIE STANDARD 21

22 OVERVIEW

OPERATION OF THE SPECIE STANDARD 23

the gold standard an additional column indicates whether a change in parity was made on resumption of convertibility after suspension. The column is omitted for bimetallic or silver experience because in these cases we have not established the dates of devaluations or revaluations after resumptions. Table 2.2 dealing with Bretton Woods gives the dates each country (except for Switzerland, which was not a member) declared its par value to the IMF, dates of suspension, if any, dates of change of par value, and reasons for suspension or par change. Evaluating pre-Bretton Woods adherence to specie rules Core countries The four countries that we designate as the core include France, Germany, the United Kingdom, and the United States. We discuss first the record for these countries before they adopted the gold standard. For the bimetallic/silver standard period, there are no entries for Germany, which was not unified until 1871. The individual German states, however, were on a bimetallic standard, as also were the other three core countries. The year 1803 is the entry in the table for the date of convertibility of the French franc into gold or silver. A bimetallic system nevertheless predated that entry by centuries, but before 1803, France had endured devaluations, revaluations, John Laws inflationary inconvertible paper money experiment of 1716–20, and the revolutionary war assignat hyperinflation of 1789–95. So 1803 marks the beginning of a stable system, with only two interruptions until 1878, when France switched to gold. The two interruptions were suspensions in 1848– 50, following the overthrow of the July monarchy, and 1870–78, following the Franco-Prussian war. Both of these interruptions qualify as consistent with adherence to specie rules, since the suspensions were valid exercises of the escape clauses. Although the table shows 1694, the year the Bank of England was founded, as the date for convertibility of the British pound into silver, Britain was on a silver standard as far back as the thirteenth century. De facto the country was on a gold standard from 1717 on, owing to the overvaluation of gold by Sir Isaac Newton, the Master of the Mint; de jure the country adopted the gold standard in 1816, while suspension of convertibility was still in effect. There had been banking crises in 1763, 1772, and 1783, but no suspensions until the war with France ended convertibility from 1797 to 1821. This again we regard not as a breach of the rule but proper invocation of the escape clause not only for the duration of the war but for a period of adjustment thereafter. Resumption at the pre-war parity also respects the rule (Bordo and Kydland, 1996). Whether the United States is eligible for inclusion among core countries in the nineteenth century is the subject of debate. We discuss the issue when we examine the status of the United States under the classical gold standard. Having

24 OVERVIEW

Table 2.2 Bretton Woods, 1946–1971, par values, suspensions and par changes

OPERATION OF THE SPECIE STANDARD 25

Sources: Bordo (1993a); DeVries (1976); Horsefield (1969). Notes a Dual exchange rates. b Change from Australian Pound to Australian dollar. c Multiple exchange rates.

concluded for the later period that the United States belongs among core countries, we also do not exclude it from the core group when it was on a bimetallic standard. The U.S. Coinage Act of 1792 defined the bimetallic standard at a mint ratio of 15 to 1. In 1834 and again in 1837 the mint ratio was altered, remaining unchanged thereafter at 16 to 1. Banking panics in 1837 and 1857 led to temporary restriction of payments by banks, but no suspension of convertibility. The Civil War, however, occasioned suspension from 1862 to 1878. In 1873 there was a banking panic, like the earlier ones, in which the banks restricted payments of high-powered paper money. Despite contentious political opposition to deflation that resumption enforced, on January 1, 1879, resumption was achieved at the prewar parity, in line with the declaration of the Resumption Act of 1875. Under the classical gold standard, both France and Germany observed specie rules until the out-break of World War I. Each then suspended convertibility, and both devalued before resuming in the 1920s. Convertibility by France lasted for eight years, by Germany for seven years, and then both devalued after suspending in 1931. The public probably regarded suspension per se because of war and financial crisis as permissible under the escape clause. The change in parity, however, diluted the credibility of the countries’ attachment to specie rules. United Kingdom’s record before World War I is the epitome of proper conduct under the gold standard. As the country at the center of the system, operating with a small gold reserve, it nevertheless managed to serve both its domestic and international interests while maintaining convertibility. Three banking panics in 1847, 1857, and 1866 led to suspension of the Banking Act of 1844, which limited the Bank of England’s fiduciary issue, but did no damage to

26 OVERVIEW

the convertibility commitment. Thereafter the Bank acted to defuse panics before they emerged, as in 1890 and 1907. Convertibility was abandoned by the Bank in World War I (de facto in 1914 and de jure in 1919), taking advantage of the escape clause, and the return to the gold standard at the prewar parity was delayed until 1925, also consonant with the provisions of the escape clause. The convertibility commitment, however, lasted only for six years, and devaluation followed. The view, alluded to above, that would exclude the United States as a member of the group of core countries, would shift it to the peripheral country group that includes Australia and Canada. That view takes its cue in part from the silver agitation and legislation of 1878 and 1890 that threatened the convertibility of U.S. dollars into gold (Eichengreen, 1992b, 1994; Giovanini, 1993; Grilli, 1990). If lasting damage to U.S. commitment credibility as a result of the threat had resulted, we would concur. Since the threat was a temporary one, and convertibility was never suspended, we conclude that the United States, by the end of the nineteenth century a colossus on the world stage, belongs with the core. Another reason advanced for excluding the United States from the core is that before 1914 it was a net capital importer, and hence more like Australia and Canada than the core countries that provided the capital. This is a narrow dimension by which to judge the United States’ relative economic importance. It was wealthier and more populous than the United Kingdom under the classical gold standard, and certainly more so than France and Germany. The United States was a capital exporter as well as an importer in the nineteenth century, and by 1914 it was a net capital exporter. These considerations reinforce our conclusion that it is properly a core country. Apart from the silver threat, convertibility from 1879 to 1914 in the United States was never in doubt. It was preserved even during two banking panics in 1893 and 1907, when banks restricted payments. In World War I the United States embargoed gold exports, 1917–19, but did not otherwise attenuate the gold standard. Specie rules were, however, flouted by the devaluation of the dollar in 1933. The changed parity legislated in 1934 remained in effect until 1971. The record of commitment by the core countries to specie rules is unblemished under the pre-World War I gold standard. Neither France nor Germany played by those rules during the interwar period, having resumed convertibility with devalued gold content of their currencies. The United Kingdom reverted to its pre-war parity when it resumed convertibility in 1925 but by 1931 devalued and abandoned rules for discretion. The United States followed the United Kingdom in devaluing in 1933 and adopted a gold standard in 1934 that diverged in fundamental ways from the pre-World War I standard.

OPERATION OF THE SPECIE STANDARD 27

Australia and Canada Australia and Canada, countries that were settled by the United Kingdom and were part of the British Empire, initially used the British currency system. Silver was the metallic medium in Australia before it adopted the gold standard, but it is not clear that a silver standard prevailed. Convertibility at a fixed Australian price of gold dated from 1852. Despite severe banking problems in the 1890s, Australia did not suspend convertibility until July 1915 during World War I. It resumed, along with the United Kingdom in 1925, at its pre-war parity, and suspended at the end of 1929, when the world depression began. It devalued in March 1930 (Butlin, 1986). In Canada the first bank charters in 1821 required convertibility of bank notes into silver. A financial crisis and political instability in 1837 led to suspension. Resumption occurred in 1839. Canada adopted the gold standard in 1853 and, although it experienced a sharp cyclical downturn in 1907–08, it did not suspend convertibility until 1914. However, no change was made in the statutory price of gold, and the gold reserve requirement for Dominion notes was not suspended, hence expansionary domestic monetary policy was subject to gold limits. Since export of gold was embargoed, exchange rates were at a discount from pre-war parities. Canada restored legal convertibility at the prewar parity in July 1926, making the monetary adjustments to return to parity without a central bank. The de facto date of Canada’s suspension of gold convertibility was 1929 (Shearer and Clark, 1984, p. 300). Canadian banks could not ship gold abroad, but foreign holders of Canadian currency obligations could redeem them in gold. De jure suspension occurred in September 1931 when both internal and external gold convertibility ended. Both Australia and Canada were as faithful as the core countries in adhering to the gold standard before 1914, but devalued in the post-World War I period. Latin America The record of the three Latin American countries, Argentina, Brazil, and Chile does not match that of Australia and Canada. Before it adopted the gold standard, from 1822 to 1825 Argentina had a brief spell of convertibility of bank notes. The metallic medium was silver and gold. Convertibility ended at a time of large government expenditures related to a war with Brazil. For the following 35 years, a period of continuing fiscal improvidence, there was no convertibility. Between 1862 and 1865 contractionary monetary policy was in force. Gold convertibility in Argentina began in February 1867 after a failed attempt in 1863 (see pp. 38–43 below). Convertibility was suspended in May 1876 after several years of political unrest and rising government deficits. Although the exchange rate reached parity by 1881, resumption that year failed. Convertibility was restored in 1883 but lasted only until January 1885, at a time of financial

28 OVERVIEW

crisis in Europe and following a period of expansionary fiscal policy. Again inconvertibility thereafter until 1899 was associated with lax fiscal policy leading to debt default in 1890. In 1899 convertibility was restored at the original parity of 5 gold pesos to the pound with the return to fiscal orthodoxy in 1896 and the establishment of a form of currency board. However, paper pesos that had been circulating since 1885 at a large discount relative to gold were frozen at 2.27 per gold peso, in effect a substantial devaluation. Argentina suspended convertibility in 1914 on the outbreak of World War I. It resumed in August 1927 at a changed parity, and suspended again in 1929. Inconvertibility prevailed during the balance of the interwar period. From 1808 onwards Brazil followed a bimetallic standard at the colonial ratio of 16:1. From then until 1846 when it was altered to favor gold, the ratio was changed three times. Gold convertibility was suspended in November 1857 in the wake of a banking crisis, and resumed in 1858. It was susequently abandoned on several succeeding occasions (notably during the war with Paraguay) (Pelaez and Suzigan, 1976). It lasted for slightly more than a year in 1888–89. (1888 was the year slavery was abolished). In 1888–89 capital inflows were extraordinarily large. A republican revolution in November 1889 coincided with the ending of convertibility (Fritsch and Franco, 1992). The gold standard was identified with the deposed monarchy, and the new government introduced a system of regional banks to increase the money supply. The real exchange rate depreciated, and convertibility was suspended. As a condition for a large funding loan from London bankers, Brazil was required to shift to contractionary fiscal and monetary policies around the turn of the century. In 1906 Brazil restored convertibility to prevent continued appreciation of the milreis exchange rate that was harmful to coffee and rubber exporters. In addition it created a Conversion Office with a limit set to its issue of convertible notes at a newly established parity. Brazil’s external position deteriorated in 1913, owing to falling coffee and rubber prices and shrinking international capital flows following the Balkan wars. A cyclical decline lasted until the outbreak of World War I, when convertibility ended to preserve the gold holdings of the Conversion Office. As was the case in 1906, resumption in 1926 was sought to prevent appreciation of the exchange rate. It followed a program in 1925–26 to achieve monetary and fiscal discipline. As in the earlier case, a Stabilization Office, modeled on the Conversion Office, was created to issue notes at the new parity. The collapse of coffee prices in 1929 and the contraction in capital inflows led in late 1930 to an almost complete loss of gold reserves by the Stabilization Office. Convertibility was then abandoned for the duration of the remaining interwar years. Chile was on a bimetallic standard from 1818 to 1851; it then made a technical change in the mint ratio, continuing on the bimetallic standard until 1866, when it suspended. It resumed in 1870, but by the end of 1874, with the fall in the price of silver, it was on a de facto silver standard. Bad crops during the

OPERATION OF THE SPECIE STANDARD 29

next three years, and accompanying balance of payments deficits, were followed by bank runs in 1878. The authorities made bank notes inconvertible on July 23, 1878 (Llona-Rodriguez, 1993). For the next 17 years, Chile remained on a paper standard. In 1879 the War of the Pacific began, with Chile opposing Bolivia and Peru, and ended in 1883 with Chile the victor. The war was financed by government note issues. Thanks to its seizure of provinces in the losing countries, Chile became the worlds monopoly producer of nitrate. However, declining prices of nitrate and copper in world markets led to depreciation of the Chilean peso from 1883 to 1893, with the domestic inflation rate lower than exchange rate depreciation. The first attempt to return to a metallic standard was made in 1887, but it failed. To appreciate the exchange rate, the government was required to retire its peso note issues and burn them, until the total issue had been reduced from 25 to 18 million pesos. It was also required to establish a silver fund for the eventual redemption of the outstanding amount. Bank issues were to be reduced from 150 per cent to 100 per cent of net worth, but neither margin was a real restraint. Bank notes rose and so did government notes. An eight-month civil war from January to August 1891, resulted in further monetary expansion and exchange rate depreciation. A second conversion law in November 1892 was strictly implemented and the exchange rate appreciated, but again the government responded to political discontent by issuing notes. The exchange rate thereupon depreciated. A new conversion law of February 11, 1895, set June 1 as the day for redemption of government notes, devalued the gold content of the peso, and authorized loans and sales of nitrate fields to accumulate a gold reserve. Bank notes, with limits on the authorized total, had to be backed by gold or bonds to be acceptable for taxes. As a result, the banks contracted and the money supply shrank. Following rumors of war with Argentina and a run on the banks in July 1898, the legislature ended convertibility and, to deal with the panic, bank notes were declared government obligations. Chile did not resume until 1925, when it again devalued, and in 1931 it abandoned the gold standard. Common elements in the experiences of the Latin American ABC countries that made their adherence to the gold standard chancy were war and threats of war and fiscal and monetary policies incompatible with fixed exchange rates. It is difficult, however, to isolate policy from the balance of payments problems that were their lot as exporters of primary products whose prices were set in world markets. Whereas for core countries war was a contingency that justified abandonment of the standard, it was a temporary abandonment with the commitment to return to it; for the Latin American peripheral countries, not only the aftermath of war but deflation generally were reasons for absence of commitment to convertibility.

30 OVERVIEW

Southern Europe Except for Greece and Portugal, the record of adherence to specie rules is mainly blank for the Southern European countries included in Tables 2.1 and 2.2. For 42 out of the 52 years between 1833 and 1885, when it adopted the gold standard, Greece was on a bimetallic standard. Until 1828 it had no national currency; Turkish coins were the medium of exchange. A commercial bank, established in 1842, operated de facto as a central bank (Lazaretou, 1994). Convertibility prevailed from February 1833 to March 1848, when suspension was declared for the balance of the year in response to panic worldwide. Resumption in January 1849 lasted through to December 1868. Although Greece signed on as a member of the Latin Monetary union in April 1867, it did not formally participate until November 1882, when it defined 1 drachma as equivalent to 1 French franc. Greece suspended for a year and a half until July 1870 because of revolution in Crete, which remained under Turkish occupation until 1899. It resumed in August 1870 until May 1877, which marked the end of its bimetallic experience. From June 1877 through to December 1884, suspension was associated with the Russian-Turkish War of 1877–78. After the war ended, Greece made several attempts to resume, cutting back on monetary growth, while the government increased indirect taxes to raise revenues. From 1879 to 1884 it borrowed 360 million gold French francs. In 1882, it devalued the drachma. Despite gold outflows because of high interest payments and a trade crisis at the end of 1884, Greece adopted the gold standard at that time. Convertibility, however, failed as gold outflows persisted, and by September 1885 Greece reverted to a paper money standard and floating exchange rates. Continued borrowing from France until 1891 and low tax revenues led to debt default in December 1893. The defeat of Greece in 1897 in the war with Turkey, which saddled it with a huge war indemnity payable in funds convertible into gold, was the spur for the appointment in 1898 of an International Committee for Greek debt management. The Committee imposed fiscal prudence on the government, and a loan of 150 million gold French francs was arranged to enable Greece to pay the war indemnity to Turkey. These measures restored confidence in Greek monetary and fiscal policies. A law of March 1910 required note circulation above a statutory ceiling to be backed by gold or foreign exchange. Bank notes were to be convertible into French francs at parity, and official reserves of the National Bank of Greece were stipulated as mainly interest-bearing deposits denominated in foreign currencies. In April 1910 Greece resumed convertibility on the gold exchange standard. The new standard was successful until December 1914. Money creation then financed wartime spending. Exchange rate parity was maintained until reserves were depleted in August 1919. Exchange rates floated until May 1928, when Greece returned to the gold exchange standard. It instituted foreign exchange

OPERATION OF THE SPECIE STANDARD 31

controls in September 1931 and devalued in April 1932, when convertibility ended. Italy, unlike Greece, adhered to a specie standard for only 14 years during the 52 years before World War I but operated a paper standard during most of the rest of the period as if subject to specie constraints. In 1862 it adopted the bimetallic standard, although de facto the standard was gold. In 1865 Italy joined the Latin Monetary union. Fiscal improvidence and war against Austria in 1866, however, ended convertibility (Fratianni and Spinelli, 1984). Fiscal and monetary discipline was achieved by 1874, and exchange rate parity was restored. The government announced on March 1, 1883, that it would restore convertibility on April 12, 1884, but convertibility took place only in silver because silver was overvalued at the mint. Public finances then deteriorated and unlawful bank issues indicated an absence of monetary discipline. By 1894 Italy was back on a paper standard, and floating exchange rates. Inconvertibility lasted until 1913. After periods of laxity, the government embraced fiscal and monetary rectitude as if it were on a gold standard. Italy did not return to the gold standard until December 1927. It resorted to foreign exchange controls in May 1934, and devalued in October 1936. Portugal was runner-up to Greece in the number of years it adhered to specie rules (Reis, 1992). It had been on a bimetallic standard since the 1680s with de facto gold predominance alternating with de facto silver predominance. In 1846 it was a weak country facing a civil war, and in no position to mint its own coin to any great extent. Instead, Portugal legalized silver coinage from other countries and set a new gold parity for the milreis at 4.5 to the pound sterling that effectively ensured that British money would mainly be the foreign inflow. England was Portugal’s chief trading partner and creditor, to whom it shipped British coin to settle its accounts. Furthermore, the mint ratio that the law established favored gold. The decision to shift to a gold standard in 1854 was made by the government as the most convenient for Portugal, since gold circulation was ample, and Bank of Lisbon paper that had been circulating at a discount was virtually back to par. The parity with the pound was unchanged from 1854 until 1891, during which period there were no convertibility crises. Gold coins circulated, notes and deposits constituting a minor proportion of the money supply. Yet Portugal’s balance of trade, except for one year, was in deficit, and it was a net capital importer. Moreover, the government budget was typically short of revenue, but until 1890 Portugal succeeded in borrowing long-term funds at home or abroad to cover the shortfall. In addition to borrowings, Portugal offset the negative elements in its balance of payments by remittances from Brazil and by earnings on Portuguese foreign investments. Thus Portugal, a debtor nation, was a regular importer of gold, unlike other peripheral nations. All this came to a halt in 1891, however after which it was no longer able to raise foreign loans. An increase in the ratio of its debt service payments to revenues, and government support of failing Portuguese

32 OVERVIEW

enterprises clouded its reputation as a creditworthy nation. The finance minister in office at this juncture was a soft-money silver supporter, which did not help Portugal’s credit standing. Portugal’s suspension of convertibility in 1891 lasted until after World War I. It returned to gold in July 1931 at a devalued parity and suspended two months later with England. Although Spain adopted a bimetallic regime in April 1848, it was not until the currency reform of 1868 that established the peseta as the monetary unit that the regime was fully operative (Martin-Aceña, 1993). In 1868 the gold-silver ratio was set at 15.5:1, as in the Latin Monetary union (which Spain did not join), whereas the 16:1 ratio set in 1848 was followed by six reductions in the intervening years. The Bank of Spain in 1874 became the monopoly issuer of bank notes that were freely convertible into both gold and silver. With the fall in the market price of silver in the 1870s, the 15.5:1 ratio undervalued gold. Gold was driven out of circulation, and the gold reserves of the Bank of Spain declined, but until mid-1883 trade surpluses and capital imports sustained convertibility. Because foreign holders of Spanish bonds refused to accept the terms of a conversion the Treasury was engaged in at the time, there was a capital outflow and a fall in inflows. In addition, the trade balance declined sharply from 1881 to 1883. To avoid deflation, Spain ended convertibility. Between 1888 and 1900 the peseta exchange rate depreciated, a budget deficit arose in every year but three from 1884 to 1899, the war with Cuba in 1898–99 was financed largely by money creation, and Spanish prices until 1905 fell much less than world prices—all factors hostile to resumption. These factors mainly after 1900 turned favorable to resumption, but it did not take place. Efforts by finance ministers to restore convertibility and adopt the gold standard before World War I foundered on the opposition of the Bank of Spain. Unlike other countries, Spain did not even briefly during the interwar period turn to gold convertibility. It adopted foreign exchange controls in May 1931. Scandinavia The Scandinavian countries were as faithful adherents to the classical gold standard as the core countries. Sweden and Denmark were independent countries throughout the period. Only Sweden of the group has a monetary history available to us for the early nineteenth century, and its record of respect for specie rules during that period is not inferior to that of the core. Sweden had a silver standard from 1803 to 1809. Large budget deficits to finance a war with Russia in 1808 bloated the money supply. In 1809 convertibility was suspended and a silver standard was not restored until 1834. This lasted until 1873, when Sweden adopted gold. The four Scandinavian countries show a common pattern of adoption of the gold standard between 1872 and 1877 and adherence to the standard until 1914.

OPERATION OF THE SPECIE STANDARD 33

Even during the period when the four countries adopted the gold standard, Finland and Norway were not independent. The former was an autonomous grand duchy of the Russian empire until 1917, the latter part of Sweden until 1905. The Swedish constitution guaranteed the convertibility of the central bank’s notes into gold, as noted earlier. For a change in gold standard arrangements to be adopted Parliament had to give its assent at two different dates with an election intervening. The central bank’s decision in 1914 to make the notes inconvertible was unconstitutional, since the bank disregarded the provision for Parliamentary approval. Only Finland devalued on resumption in 1926. The others resumed at their prewar parities. Sweden returned to gold de jure in March 1924, but de facto the prewar par rate of the krona in gold was restored in 1922. What delayed the de jure return was the authorities’ opposition to the krona being the sole convertible European currency. All the Scandinavian countries suspended in 1931 and devalued. In June 1933 the Swedish krona was fixed to to the British pound, and the exchange rate was unchanged until after the start of World War II. Western Europe The three Western European countries listed in Part 6 of the table, Belgium, the Netherlands, and Switzerland, adopted the gold standard in the second half of the 1870s and adhered to it until World War I. A note-issuing bank was established in Belgium in 1822 before the country became independent in 1832, when it was a bimetallic adherent. It suspended convertibility in 1848 in the face of French political and financial problems. The Netherlands was on a bimetallic standard in 1847, Switzerland in 1850. All three countries suspended convertibility in 1914. Belgium lost the right of note issue following the German occupation. Switzerland declared bank notes legal tender in 1914. The Netherlands prohibited gold export in 1914. Belgium devalued when it returned to gold in October 1926, and devalued again in March 1935. The Netherlands returned to gold at the pre-war parity in April 1925 and devalued in October 1936. Switzerland devalued in 1929 when it returned to gold and when it left the gold standard in 1936. Japan Convertibility was not firmly established in Japan until 1885, when it was on a silver standard (Shinjo, 1962). In 1868 Japan introduced a new monetary unit, the yen, defined as the same weight and fineness as the Mexican silver dollar. In an act of 1871 Japan prescribed a gold yen, but silver still remained the preferred metal for foreign trade. Bimetallism was in effect as judged by the government’s metallic reserve. The government issued notes beginning in 1868 that were

34 OVERVIEW

redeemable in silver, with an 1880 expiration date of redemption. In addition, government-issued currency notes were inconvertible.12 New national banks were created late in 1872 to issue bank notes against specie reserves of not less than two-thirds the amount emitted. The notes, however, quickly returned to the banks for redemption in silver. The petition in 1875 of four existing national banks to change from silver to government currency note convertibility was granted the following year. National banks multiplied and their issues sharply increased.13 From 1868 until 1878 the paper currency depreciated against both silver and gold. The government issue ceased in 1879. It was recognized that a budget surplus was essential to decrease the outstanding note issue and to accumulate specie. By 1881, after keeping government expenditure constant for three years, the government budget was in surplus, and this was used partly to destroy existing notes and the rest as specie reserve. In October 1882 the Bank of Japan was founded, but it did not issue convertible notes, payable in silver on demand, until 1885. These notes, initially limited in amount, replaced government currency notes. In 1888 the bank was authorized to issue a substantial fiduciary circulation backed by government and commercial paper, any amounts in excess to be backed by gold and silver. National banks lost the right of issue after the expiration of their charters. Deposit banks replaced them. The premium on silver disappeared once convertibility was established but, since the price of silver against gold was declining, the exchange rate against gold standard countries was unstable. By 1893 Japan recognized that it was desirable to adopt the gold standard, but the reform was not introduced until the indemnity China paid in gold for losing the Sino-Japanese war of 1894–95 enabled the government to acquire an adequate reserve. The Coinage Act of 1897 established the gold standard. The act governing the Bank of Japan was revised to require convertibility of its notes into gold instead of silver. The government contributed the gold indemnity to the bank’s reserve. Silver was limited to one-fourth of the specie reserve. Japan adhered to the gold standard until September 1917, without interruption despite runs on banks in 1901 and 1907–08, war with Russia in 1904–05, extraordinary government expenditures financed by foreign loans, and an unfavorable balance of payments during most of the period. The decrease in the specie reserve prompted flotation of foreign bonds in London and Paris in every year from 1906 to 1915. Foreign capital maintained Japan’s gold standard. Japan was a beneficiary of the demand for its goods and services by World War I belligerents. It used the surplus in its balance of payments to increase the Bank of Japan’s and the government’s gold reserves, to replenish its foreign exchange balances abroad, to pay back foreign loans, and to increase its foreign investments. Japan became a creditor country. In September 1917 Japan followed the United States in embargoing gold and silver export. Bank of Japan notes became inconvertible. Wartime prosperity

OPERATION OF THE SPECIE STANDARD 35

ended in 1920, with failures of firms and runs on banks. A deflationary policy in 1921–22 provoked further runs on banks, and the policy was discontinued. An earthquake in September 1923 led to increased government expenditures. Again there was a move to deflation and a readiness to follow England’s return to gold in April 1925. Financial panic in 1926 halted that step. A new government in 1929 adopted a program of lifting the gold embargo and returning to gold at the pre-war parity, as it did in January 1930. Speculative transactions to sell yen and buy dollars once the gold standard was restored reflected the market’s belief that the yen would have to be devalued. The suspension of the gold standard in December 1931 came after huge gold losses by Japan. In March 1932 Japan began a series of devaluations of the gold content of the yen as the exchange rate of the yen declined. Foreign exchange controls were introduced in May 1933. The limit on the Bank of Japan’s fiduciary note issue was repeatedly expanded, and ultimately eliminated. Evaluating adherence to rules during Bretton Woods Core countries Under Bretton Woods, the rule for countries other than the United States was that a change in par value was permissible to correct fundamental disequilibrium. Though undefined, the term was intended to refer to disturbances other than government policies that justified a change in par. Examples of such disturbances were a change in the terms of trade and in productivity trends, and other contingencies similar to those the gold standard escape clauses encompassed. Obtaining the advance agreement of the IMF to a change in par was a way of insuring that such were the disturbances that prompted the action (Eichengreen, 1994). As is well known, the Bretton Woods arrangements required countries other than the United States to peg their currencies to the dollar, and the United States to peg the dollar to gold. The system became fully operative in 1959 and broke down in 1971. The rules for neither the center nor the other countries were successful (Bordo, 1993a; Giovannini, 1993). Problems raised by the rule for countries other than the United States were apparent from the start of the system. We review the problems as evidenced by core country performance. When France devalued in January 1948, it did not seek IMF authorization. By creating multiple exchange rates, it took a discretionary action, contrary to the rule. It then restored unified exchange rates in the devaluation of 1949. The United Kingdom announced convertibility for current account transactions in July 1947, but suspended the next month, making sterling subject to exchange controls. It declared a par value in 1948, but devalued in September 1949 by a larger per cent than it had indicated to the IMF.

36 OVERVIEW

After the general realignment of September 1949, in which Germany joined, changes in par of the core countries were rare, possibly due to reluctance to alter their parities after the experience of the speculative attacks that occurred after the 1949 changes. Keynesian full-employment policies that many countries adopted conflicted with Bretton Woods obligations to maintain fixed exchange rates. In consequence, exchange rate crises erupted as market participants anticipated policy-induced pressures to devalue or revalue. The countries whose currencies the market targeted for attack resisted changing par values in the belief that domestic concerns should not yield to external concerns. They tried to buy time by imposing capital controls and deploying international reserves to preserve existing par values, but in the end failed. The meaning of fundamental disequilibrium, however, was altered to refer to government monetary and fiscal policies that were inconsistent with those par values. As events unfolded, it was difficult to distinguish the original correct use from this incorrect use of the escape clause. The core country par values that changed reflected this shift in the meaning of fundamental disequilibrium. Two examples are France, under inflationary conditions, which devalued in 1957 and 1958, and the U.K. Chronic United Kingdom balance of payments deficits during the 1960s led to devaluation of sterling in November 1967, despite rescue packages. France suspended the par value of the franc in November 1968 because of speculative attacks on its currency at a time of social unrest and inflation. Capital controls and massive international loans were not effective in preventing devaluation in August 1969. In contrast, Germany revalued in 1961, in response to a persistent balance of payments surplus reflecting higher productivity than her partners. In 1968 it resorted to border taxes and restricted capital inflows, but ultimately revalued again in September 1969. During 1969–71 a persistent outflow of funds from the United States overwhelmed foreign exchange markets. In May 1971 Germany suspended dealings in Deutschmarks and allowed its currency to float, since it could not maintain exchange rates within the established margins. France introduced dual exchange rates in August 1971, and the United Kingdom suspended the sterling par value that month, and allowed sterling to float in June 1972. If countries other than the United States did not observe the Bretton Woods rule on par value change as it was conceived, neither did the United States, the center country, comply with the gold convertibility rule. Faced with balance of payments deficits after 1957 that increased dollar liabilities while the monetary gold stock was shrinking, the problem for the United States was how to preserve convertibility. Many stratagems were devised to induce holders of dollars to refrain from cashing them in for gold. In addition, the elimination of the gold reserve requirement against Federal Reserve notes in 1968 betokened a weakening of the commitment mechanism to maintain stable money—an obligation of the reserve country. Moreover, by engaging in expansionary monetary policy after the mid-1960s, the United States exacerbated the

OPERATION OF THE SPECIE STANDARD 37

convertibility problem, since the incentive to hold dollars declined as inflation rose. The closing of the gold window in August 1971 marked the end of the convertibility rule and the readjustment of currency parities at the Smithsonian meeting in December 1971 marked the first dollar devaluation since 1934. Peripheral countries Like the core countries, peripheral countries under the Bretton Woods system differed in the extent to which they adopted expansionary monetary and fiscal policies. High-inflation countries repeatedly devalued; low inflation countries did not. Australia, at the start of the Bretton Woods era, was part of the sterling area. Along with the United Kingdom it declared the par value of its pound in 1947 and devalued in 1949. In 1966 it changed its currency unit to the Australian dollar, as it converted its currency to the decimal system. No appreciation or depreciation of the exchange rate accompanied the new monetary unit. Australia did not devalue with the United Kingdom in 1967. It suspended its par value in August 1971, but did not change it, so its currency appreciated relative to the U.S. dollar from the par value that existed before the closing of the U.S. gold window. Canada declared its par value at the end of 1946 and devalued with the United Kingdom in 1949. It acquired a special status under Bretton Woods when the IMF did not actively oppose its decision to float its dollar in September 1950. Capital inflows from the United States were increasing Canada’s reserves, with expansionary consequences for its money supply. Under fixed exchange rates Canada found it difficult to resist the inflationary results. It did not revert to fixed exchange rates until May 1962. In May 1970 Canada again decided to float for the same reason as it had two decades earlier. Its foreign exchange reserves were accelerating, and the situation that was created thereby was deemed unmanageable under fixed rates. Canada continued to float until the collapse of Bretton Woods. The Latin American countries all had high inflation experiences during the Bretton Woods era. Before Argentina declared its par value in 1957 it had in 1955 introduced multiple exchange rates. It devalued in 1959, 1962, and 1970. Brazil was plagued by inflation from World War II on, and especially after 1958. It devalued in 1967, and in August 1968 introduced a flexible exchange rate policy which involved devaluation of the currency by small amounts at frequent irregular intervals. Similarly, Chile, which declared its par value in 1946, introduced multiple exchange rates in 1953 and devalued in 1962. Of the four Southern European countries, only Spain devalued before 1971. It did so in November 1967. The IMF regarded the devaluation as a correction of a previously existing fundamental disequilibrium. In 1971 Spain maintained its par value unchanged, but widened the margin to up to 2¼ per cent.

38 OVERVIEW

Only Sweden of the Scandinavian countries did not devalue between 1951, the date of par declaration, and 1971. Denmark devalued in 1949 and 1967. Finland devalued in 1957 and 1967. Norway devalued in 1949. In Western Europe Belgium and the Netherlands realigned with Germany in 1949 and the latter revalued with Germany in 1961. Japan did not devalue between 1949, when it declared its par value, and 1971. Conclusion The chronology of adherence to rules before and after Bretton Woods reveals a decay of respect for rules over the century Tables 2.1 and 2.2 cover. Rules were not universally honored even during the classical gold standard era. A core group of countries was usually faithful to specie rules, but countries in Western Europe, the new Anglo-Saxon settlement countries, and the Scandinavian countries also conducted their financial affairs so that the fixed price of gold that defined their currencies was unchanged for extended periods. Monetary and fiscal policies in the remaining countries in the table were such that suspensions of the specie rule were not exceptional, and they were followed by changes in the former parity. Only extraordinary events like wars occasioned departures from the standard among the core and their cohorts, and until World War I resumption took place at the pre-war parity. In the interwar period a return to the gold standard was sometimes at the earlier parity, but often at a devalued rate. During most of the period floating exchange rates were common. The attempt under Bretton Woods to impose a rule that the par value of its currency with the dollar that each country member declared would be changed only under extraordinary circumstances, as under the classical gold standard, failed. Domestic economic objectives proved to be paramount to international obligations. CAPITAL FLOWS AND SPECIE STANDARD ADHERENCE: ARGENTINA AND THE UNITED STATES, 1865–1914 One of the enforcement mechanisms of the specie standard rule for peripheral countries was presumably access to the capital needed for their economic development from the core countries. Adherence to the convertibility rule would be viewed by lenders as evidence of financial probity—i.e. membership in the international gold standard would be like a ‘good housekeeping’ seal of approval. It would signal that a country followed prudent fiscal and monetary policies and would only temporarily run large fiscal deficits in well understood emergencies. Moreover, the monetary authorities would be willing to go to considerable lengths to avoid defaulting on externally held debt. It would also presumably be a signal to the lenders in London and other metropolitan areas that the groups in power observed similar standards of financial rectitude.

OPERATION OF THE SPECIE STANDARD 39

This suggests that adherence to the specie standard rule, ceteris paribus, would make a difference in the volume of capital a country attracted from abroad. Presumably loans would only be made with gold clauses (or be sterling denominated), so that currency risk would not matter. But there still would be a risk of abrogation of the gold clauses or of total default on the debt. That eventuality would be reflected in a risk premium on the loan. In that case it would be attractive to a potential borrower to adhere to the specie standard rule as a signal of financial responsibility, to induce the lender to lower the risk premium.14 But a more fundamental problem could arise in a world of asymmetric information with the possibility of a ‘lemons premium’ (Akerlof, 1970; Stiglitz and Weiss, 1981). In that case, charging a high interest rate might attract borrowers willing to engage in unduly risky projects. Lenders faced with imperfect information on borrowers’ likely actions would then be reluctant to lend at any price. A credible commitment to the specie standard rule, as evidenced by the holding of substantial gold reserves, would provide a signal to lenders of the costs borrowers would be willing to bear to avoid default, and hence would circumvent the aversion to lending imposed by asymmetric information. As a tentative step in the direction of examining the connection between capital flows and adherence to the specie standard rule, we briefly focus on the experience of two major borrowers of British capital in the late nineteenth century, each of which had a record of suspension and of specie standard adherence over the period 1865–1914—Argentina and the United States. In that fifty-year period, Argentina was off gold in three episodes of suspension totalling twenty-four years (excluding the general breakdown of the international gold standard in 1914). The United States was off gold for seventeen years and the gold standard subsequently was under threat of suspension for another seven years. In Figures 2.1 and 2.2 we present annual data on capital calls on new issues of securities—a measure of access to new capital in London for the two countries. The data, kindly supplied by Lance Davis, underlying Davis and Huttenback’s (1986) study of the economics of British imperialism,15 are expressed in millions of current U.S. dollars.16 The periods of suspension (and for the United States the episode of silver threat to the gold standard) are marked off in the figures by shaded areas. An ideal analysis of the influence of adherence to the specie standard rule would be based on a model of the determinants of capital flows, including such variables as: the expected real rates of return in both countries, the levels of real activity, the terms of trade, and the phase of the business cycle (see Ford, 1962; Abramowitz, 1973; Edelstein, 1982). One could then test for the marginal influence of adherence/non-adherence to the rule. Here our aims are much more modest—to simply compare the annual capital calls on new issues of securities with a chronology of events related to adherence to the rule. Any connection revealed should be treated merely as highly suggestive.

40 OVERVIEW

Figure 2.1 Capital calls on new issues of Argentine securities in London, 1865–1914 (current dollars)

Bertalomé Mitre became president of Argentina in 1862 and succeeded in unifying the country after five decades of intermittent civil strife, external wars and highly unstable monetary and fiscal policies (Cortés-Condé, 1989). Under Mitre, contractionary monetary and fiscal policies were successful in achieving specie convertibility in 1867 (following a failed attempt in 1863). Argentina then began five decades of extraordinary economic growth, with rapid development in agriculture, transportation, and commerce. It also was the start of a wave of immigration from Europe and of the massive inflow of capital (Cortés-Condé, 1989). The pattern for Argentina in Figure 2.1 reveals little change in capital flows from a low and stable level when the country joined the gold standard in 1867 until 1876. The period 1870–75 is characterized by a mild boom in capital calls, ending just before suspension of convertibility in 1876. It would be difficult to disentangle the effect of establishing convertibility in 1867 from that of restoring political stability as the key determinants of the beginning of capital flows from Europe. Increasing civil strife in 1873–76, rising government deficits, and a down-turn in the world business cycle account for the suspension of convertibility in 1876. It was followed by several years of negligible capital inflows. Thanks to contractionary fiscal policy in the late 1870s, the exchange rate reached parity in 1881 but it took until 1883 to restore convertibility (Cortés-Condé, 1989). The commitment to restore convertibility may have led to the observed resumption of capital inflows. Convertibility was short-lived, however. Expansionary fiscal

OPERATION OF THE SPECIE STANDARD 41

policy in 1884 led to a crisis and suspension on January 1, 1885. Again suspension is associated with a decline in capital flows but the significant rebound from 1886 to 1889 under inconvertible money suggests that British investors placed more weight on the long-run economic prosperity of the Argentine economy than currency stability. The boom ended with a crash in 1890, following several years of exceedingly loose fiscal policy and the creation in 1887 of a free banking system which produced a plethora of bank money (Eichengreen 1992b). A revolution in that year, followed by default on external debt, precipitated the Barings crisis in London. Capital inflows then plummeted until the authorities again instituted monetary and fiscal austerity. It took four years for convertibility to be restored in 1899. Restoration of convertibility along with the creation of a quasi currency board (the Caja de Conversion), which in essence tied the hands of the monetary authorities, succeeded in creating a climate conducive to the resumption of significant capital movements until World War I. This narrative suggests that adherence to the rule by Argentina may have had some marginal influence on capital calls on new issues of securities in London before 1890 (see Table 2.3 which suggests that the mean of capital calls in current and real dollars was higher during periods of adherence than of suspension), but that the key determinant was the opening up of the country’s vast resources to economic development once unification and a modicum of political stability were achieved. The 1890 crisis was a major shock to investor confidence and it took years of austerity, the restoration of convertibility, and the establishment of a currency board before British investors’ confidence was restored. The U.S. experience under suspension has been well studied by others (Sharkey, 1959; Unger, 1964; Friedman and Schwartz, 1963; Roll, 1972; Calomiris, 1988). As can be seen in Figure 2.2 capital calls on new issues were low in the first five years after the Civil War. The debate over resumption in these years seems to have had little impact on the decision by investors to purchase securities in London destined for the United States, but then purchases picked up significantly in 1870 and 1871. This phenomenon may be explained by the Public Credit Act of 1869 which guaranteed that the principal on U.S government bonds would be payable in gold (Calomiris, 1988). The subsequent decline may reflect adverse news of the likelihood of resumption with the reversal in 1871 of an earlier Supreme Court decision declaring the issue of greenbacks unconstitutional, as well as the Treasury’s expansionary fiscal policy (Calomiris, 1993). Capital calls increase until they are reversed by the Panic of 1873, a decline in economic activity, and two years of soft money victories (the reissue of retired greenbacks in 1873 and the Inflation Bill of 1874). The Resumption Act of 1875 is then followed in the next two years by the largest increase in capital inflows over the whole fifty-year span. Resumption of specie payments on January 1, 1879, is followed by a rising but variable trend in capital

Source: Davis and Huttenback (1986).

Table 2.3 Descriptive statistics of new issues of securities in London (current and real values): 1865–1914

42 OVERVIEW

OPERATION OF THE SPECIE STANDARD 43

Figure 2.2 Capital calls on new issues of U.S. securities in London, 1865–1914 (current dollars)

calls on new issues but the volume is well below the average of the suspension period (see Table 2.3). The period 1890–96 is important in the history of U.S. adherence to the rule. Passage of the Sherman Silver Purchase Act in 1890 led to a six-year period of uncertainty surrounding the nation’s ability to remain on the gold standard (Friedman and Schwartz, 1963; Calomiris, 1993; Grilli, 1990). This episode is reflected in the capital calls on new issue series in Figure 2.2. They peak in 1890 and though they may have rallied following the repeal of the silver purchase part of the Sherman Silver Act in 1893, and the Belmont Morgan syndicate replenishment of the Treasury’s gold reserves in 1895, they rebound significantly only after the 1900 Gold Standard Act. In sum, for the United States as for Argentina, events suggesting the restoration of convertibility during suspension and threats to convertibility during adherence seem to be associated with increases and declines in capital calls on new issues of securities in London. More systematic research is required to distinguish these influences from the fundamental determinants of capital flows. THE ECONOMIC PERFORMANCE OF CORE AND PERIPHERAL COUNTRIES UNDER ALTERNATIVE MONETARY REGIMES In this section we present some evidence based on annual data over the past 110 years on the macroeconomic performance of the four core countries and

44 OVERVIEW

seventeen peripheral countries during the classical gold standard (1880–1914) and three successive monetary regimes: the interwar period (1919–39); the Bretton Woods international monetary system (1946–70); and the recent managed float (1974–90). The Bretton Woods system, as a variant of the contingent specie standard rule, is directly comparable to the classical gold standard. The recent managed float, a regime not based on the rule, and the interwar period, which comprises episodes of free floating, adherence to the gold standard, and managed floating, are presented as contrasts to the two rule-based regimes.17 The data are organized in the seven groupings shown in Tables 2.1 and 2.2, as well as broader aggregates. Such evidence for the classical gold standard regime and for Bretton Woods may shed light on whether differing economic performance can explain why some countries successfully adhered to the convertibility rule and others did not, or whether adherence/non-adherence to the rule may have influenced performance. By contrast, under the recent floating regime we seek to determine whether the observed differences between countries’ performance under convertible regimes persist in the absence of rules. Stability and convergence Tables 2.4 through 2.6 present descriptive statistics on three macro variables for each country pertinent to the issue of adherence to convertibility rules, the data for each variable converted to a continuous annual series from 1880 to 1989. The three variables are: the rate of inflation (GNP deflators), money growth, and the ratio of government expenditure less government revenues to GNP.18 The definition of the variable used, e.g., M1 versus M2, was dictated by the availability of data over the entire period. For each variable, and each country we present two summary statistics: the mean and standard deviation. For each of the seven country groupings from Table 2.1, and four aggregate groupings (all countries, all except the four core countries; the G-10 countries plus Switzerland (G-11); and all except the G-11) we show as a summary statistic: the grand mean. We comment on the statistical results for each variable. Inflation (Table 2.4) The classical gold standard had the lowest rate of inflation of any monetary regime for all 21 countries, and the interwar period displayed mild deflation for all except Latin America (see Figure 2.3). Within the classical gold standard regime, the inflation rate was lowest in countries identified in Table 2.1 as following the convertibility rule: the four core countries; some of the different European groupings and the Anglo-Saxon countries of new settlement, see Figure 2.3.19 It was considerably higher in Latin America and Japan. This pattern can also be seen in a comparison of the G-10 plus Switzerland aggregate (an

OPERATION OF THE SPECIE STANDARD 45

expanded core group) with the other peripheral countries. The former grouping contains nine countries which followed the rules; the latter only three. Under Bretton Woods, like the gold standard, a distinct difference can be observed between the core countries and a number of other groupings (AngloSaxon new settlement, other Western Europe, Japan), which had low inflation; and a set of countries with higher inflation (Latin America, to a lesser extent countries in Southern Europe and Scandinavia). This observed difference between country groupings also is found under the recent float. The evidence on inflation suggests that if the specie standard rule or its variant did provide a credible commitment mechanism for low inflation, it was strongest in the gold standard period followed by Bretton Woods. Within these regimes observance of the rule clearly demarcates inflation performance between countries. The gold standard period had the most stable inflation rate of any regime (across all countries) judged by the standard deviation. This was followed by Bretton Woods, the interwar and then the float. For the G-11 countries, the recent float is the most stable period. Within the gold standard regime, core countries and countries following the specie standard rule exhibited greater price stability than the others. For Bretton Woods a similar difference between country groupings is observed, with Japan joining the stable inflation group, and Southern Europe the unstable inflation group. These differences persist into the recent float. The evidence of a high degree of price stability under the gold standard (and to a lesser extent under Bretton Woods) and of greater price stability during those periods in countries following the rules compared to those that did not is consistent with the traditional view that commodity-money-based regimes provide a stable nominal anchor; however, the price stability observed may also reflect the absence of major shocks. Money growth (M2) (Table 2.5) Money growth was considerably more rapid across all countries post-World War II than before the war (see Figure 2.4). For the core countries and most of the G-11 countries there is not much difference between Bretton Woods and the subsequent floating regime. Southern Europe, Latin America and Japan exhibited considerably higher money growth under Bretton Woods than the others. For the Latin American countries money growth rates accelerated over the entire postwar period, reaching their highest levels under the float. By contrast with postwar, the gold standard exhibited lower money growth in both core and peripheral countries alike with the principal exception of Japan; however, it was still higher in core than peripheral countries. The observation of lower money growth in both core and peripheral countries may be a reflection of the omnipresence of the specie standard rule. Across all countries, money growth was least variable in the interwar and most variable in the recent float. However, for the core and G-11 countries it was

Table 2.4 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: inflation. Annual data: means and standard deviationa

46 OVERVIEW

a Mean growth rate calculated as the time coefficient from a regression of the natural logarithm of the variable on a constant and a time trend. Data sources: See Data Appendix.

OPERATION OF THE SPECIE STANDARD 47

48 OVERVIEW

Figure 2.3 Inflation rate, 1881–1990

least variable under the gold standard. Under that regime money growth variability was higher in the periphery than the core. Under Bretton Woods and the recent float the difference between core and periphery money growth

OPERATION OF THE SPECIE STANDARD 49

variability is less obvious, again with the principal exception of Latin America, which exhibits considerably greater money growth variability than all other

50 OVERVIEW

countries. For these countries the increase in money growth variability reflects the breakdown of any linkage to a commitment regime.

OPERATION OF THE SPECIE STANDARD 51

Government deficit (Table 2.6) For the underlying data see Figure 2.5. For all 21 countries the average ratio of the government deficit to GNP is lowest during the Bretton Woods period, followed by the gold standard. The highest ratio is for the recent float. In all regimes, average deficit ratios are higher in peripheral than core countries. (This also holds comparing the G-11 with the rest.) Under the gold standard, the highest average ratios were in the Anglo-Saxon countries of new settlement and Latin America; under Bretton Woods they occurred in Latin America and under the float in Latin America and Southern Europe. The standard deviations in fiscal policy followed a pattern similar to the means—they are generally lower in core than peripheral countries. Thus, adherence to convertibility rules may have constrained fiscal policy in the same way as it did monetary policy. On the other hand, more limited fiscal needs during these regimes may have made it easier to adhere to the convertibility rule. One group of countries stands out in the cross-country comparison. For the Latin American countries the fiscal deficit as a share of GNP increases dramatically between the pre-World War II and post-World War II regimes. In the post-war period it increases between regimes, reaching a peak with the float. Indeed a closer correlation between the fiscal deficit, money growth and inflation can be observed across regimes for these countries than is the case for most countries in the G-11. Such a correlation may be evidence of lack of credibility in the commitment of the monetary regime.

Table 2.5 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: money growth. Annual data : means and standard deviation a

52 OVERVIEW

a

notes and sources: See Table 2.3

OPERATION OF THE SPECIE STANDARD 53

54 OVERVIEW

Figure 2.4 Money growth rate, 1881–1990

OPERATION OF THE SPECIE STANDARD 55

Summary In summary, the gold standard regime had the lowest and least variable inflation performance for all countries. However, within that regime peripheral countries performed worse than the core (and expanded core). Bretton Woods exhibited a similar pattern with somewhat higher overall inflation. By contrast the recent

56 OVERVIEW

float displayed both higher and more variable inflation than both regimes. This evidence may reflect a favorable influence of regime adherence on performance,

OPERATION OF THE SPECIE STANDARD 57

but on the other hand adherence may have been possible because of greater stability. Money growth was generally lowest and most stable under the gold standard across all countries, followed by Bretton Woods. Core countries within each regime followed more prudent and stable monetary policies than the periphery. By contrast, under the float, money growth was considerably higher, less stable across all countries. Lower and more stable money growth by core countries under the gold standard compared to the periphery may reflect better adherence to the commitment mechanism, or alternatively that there was less pressure to gear monetary policy for domestic purposes (Eichengreen, 1992b). The fact that peripheral countries’ money growth was still relatively low and stable compared to later regimes may reflect their intention to adhere to the rule, when conditions were favorable. A similar but more muted pattern is observed in comparing Bretton Woods to the subsequent float. Finally, the fiscal deficit, like monetary policy, is lowest and most stable under the gold standard, followed by Bretton Woods and in sharp contrast to the recent float. Within these regimes, however, weaker performance is observed for peripheral countries. Again like monetary policy, intended adherence to the regime, ceteris paribus, may have restrained policy-makers during these periods compared to the recent float. Thus, this statistical evidence reveals substantial differences in economic performance across regimes as well as differences between countries’ and groupings of countries’ performances within regimes. It is not clear, however, how much the different performance reflects adherence or non-adherence to rules

Table 2.6 Descriptive statistics of selected open economy variables, 21 countries, 1881–1990: government deficit as a percentage of GNP. Annual data: means and standard deviation

58 OVERVIEW

a

notes and sources: See Table 2.3

OPERATION OF THE SPECIE STANDARD 59

60 OVERVIEW

Figure 2.5 Government deficit as a percentage of nominal GNP, 1881–1990

and vice versa. Analysis of the shocks facing different countries may shed more light on this issue.

OPERATION OF THE SPECIE STANDARD 61

Demand and supply disturbances An important issue is the extent to which the performance of alternative monetary regimes, as revealed by the data in the preceding tables, reflects the

62 OVERVIEW

operation of the monetary regime in constraining policy actions or the presence or absence of shocks to the underlying environment. One way to shed light on this issue, following earlier work by Bayoumi and Eichengreen (1992, 1993,

OPERATION OF THE SPECIE STANDARD 63

1994a, 1994b), is to identify underlying shocks to aggregate supply and demand. According to them, aggregate supply shocks reflect shocks to the environment and are independent of the regime, but aggregate demand shocks likely reflect policy actions and are specific to the regime. The approach used to calculate aggregate supply and demand shocks is an extension of the bivariate structural vector autoregression (VAR) methodology developed by Blanchard and Quah (1989). Following Bayoumi and Eichengreen (1994a), we estimated a two-variable VAR in the rate of change of the price level and output.20 Restrictions on the VAR identify an aggregate demand disturbance, which is assumed to have only a temporary impact on output and a permanent impact on the price level, and an aggregate supply disturbance, which is assumed to have a permanent impact on both prices and output.21 Overidentifying restrictions, namely, that demand shocks are positively correlated and supply shocks are negatively correlated with prices, can be tested by examining the impulse response functions to the shocks. The methodology has important limitations which suggest that the results should be viewed with caution. The key limitation is that one can easily imagine frameworks in which demand shocks have permanent effects on output while supply shocks have only temporary effects.22 We estimated supply (permanent) and demand (temporary) shocks, using annual data for each of the 21 countries, over alternative regimes in the period 1880–1989. The VARs are based on data for three separate regime periods (to the extent available): 1880–1913, 1919–39, and 1946–89, omitting the war years because complete data on them were only available for a few of the countries.

64 OVERVIEW

The VARs have two lags. We also did the estimation for aggregated price and output data for the seven country groupings and for the four broader aggregates of countries. Table 2.7 presents the standard deviations of supply and demand shocks for the 21 countries by regimes. We also present aggregate shocks for the seven country groupings and for the four broad aggregates of countries. In addition we show, following Bayoumi and Eichengreen (1994a b), the weighted average of the individual country shocks. Figure 2.6 shows the shocks for the seven country groupings and aggregate shocks for the four broad aggregates of countries.23 Table 2.7 shows for the seven country groupings, with the principal exception of Latin America, that the recent float regime was the most tranquil, with the lowest demand and supply shocks. The interwar period, again with the exception of Latin America, was the most volatile.24 Bretton Woods experienced relatively high demand shocks in most of the country groupings with the exception of Japan. This likely reflects the widespread use of Keynesian demand management policies in this period. By contrast supply shocks were quite low across all groupings, not much different from the float. The gold standard in general exhibited fairly sizeable supply shocks. Indeed, for the core countries, the AngloSaxon countries and Japan, they were more than twice as high as in the postWorld War II period.25 In Latin America demand shocks exceeded supply shocks in all regimes except the interwar. In the post-World War II period, for these countries, demand shocks considerably exceeded supply shocks, especially under the float. For these countries the constraints of the convertible regime appear to be much weaker than for the others, although one could argue that the much greater instances of supply shocks in these countries may in part account for the greater use of discretionary policy. Across country groupings, the difference between the ‘expanded core’ of countries and the periphery observed in Tables 2.4 to 2.6 is also apparent. Under the gold standard regime most of these countries had substantially lower supply and demand shocks than the others. This can be seen in the comparison between the G-11 aggregate and the aggregate of all countries except the G-11. The pattern also holds up in both post-World War II regimes. In sum, the evidence on demand and supply shocks complements the preceding evidence in Tables 2.4 to 2.6. For the ‘expanded core’ countries, represented roughly by the G-11, the gold standard was characterized by higher demand and especially higher supply shocks than in the post-World War II regimes, and within the post-war period both the Bretton Woods regime and the float were relatively stable.26 This evidence suggests that, for these countries, it is unlikely that the convertible regimes prevailed because of the absence of supply shocks, since the size of demand and supply shocks was quite similar across both types of regimes. The durability or fragility of past convertible regimes likely had more to do with regime design (Bordo, 1993b). By contrast, for the peripheral countries, especially Latin America and Southern Europe, demand shocks exceeded supply shocks across all regimes, and

Table 2.7 Supply (permanent) and demand (temporary) shocks: 21 countries, 1881–1990. Annual data: standard deviations of shocks (per cent)

OPERATION OF THE SPECIE STANDARD 65

66 OVERVIEW

Notes a Aggregate data. b Weighted average of individual country shocks. The weights are calculated as the share of each country’s national income in the total income in the group of countries in 1970, where the GNP data are converted to dollars using the actual exchange rate. Data sources: See Data Appendix.

68 OVERVIEW

Figure 2.6 Supply and demand shocks: country groups, 1881–1990

especially under the post-war float. The constraints of the convertible regime appear to be much weaker than for the G-11, although one could argue that the

OPERATION OF THE SPECIE STANDARD 69

much greater incidence of supply shocks in peripheral countries may in part account for the greater use of discretionary policy.

70 OVERVIEW

CONCLUSION In conclusion, we suggest answers to three questions. First, what is the evidence for the specie standard as a contingent rule? Second, why was the rule successful

OPERATION OF THE SPECIE STANDARD 71

when it was? Third, why did some countries successfully adhere while others did not?

72 OVERVIEW

Our historical survey reveals that the contingent rule over the entire period was strictly followed by a relatively small number of countries—the core countries (with the exception of France and Germany after World War I); the

OPERATION OF THE SPECIE STANDARD 73

Anglo-Saxon countries of new settlement, most of the Scandinavian and smaller Western European countries, and Japan. For the rest it was violated. If we focus only on the classical gold standard period, the basic convertibility rule was followed by a larger group of countries. For Bretton Woods, the rule was much less clearly defined, hence it is difficult to distinguish adherence and nonadherence. For that regime clearly, as statistical evidence of superior performance measured by nominal variables compared to other regimes shows, convertibility was important. Second, the rule prevailed when it did for a number of possible reasons. The classical gold standard era was one of stable economic growth, with few impediments to the free allocation of labor, capital and goods both within and across countries. It was also a period characterized by relative political stability and, for many of the ‘expanded core’ countries by the absence of populist pressure for demand management (Eichengreen, 1992a). In addition, it was an era characterized by the coincidence of beliefs in free trade and exchange rate stability by the dominant industrial commercial groups in different countries (Gallarotti, 1993). The dominance of England as a commercial power also was important. England’s clear commitment to the rule and the benefits of access to her finance markets was a key determinant of other countries’ adherence. Third, the different experience of adherence to the rule by the expanded core countries and the peripheral countries may, we suggest, be explained by their different stages of economic development. By the beginning of the classical gold standard era the ‘expanded core’ countries were industrialized, had experienced decades of rising per capita income, had stable polities and were dominated by

74 OVERVIEW

groups who perceived the benefits to them of monetary stability. In addition, as maturing diversified economies they were less subject to the disruptions of massive swings in the price of primary products. These conditions were absent in the peripheral countries. Faced with frequent supply shocks, for them adherence to the rule was more difficult. NOTES * For able research assistance we would like to thank Alexandre Hohmann. For supplying us with data and valuable information we are indebted to the following: Pablo Martin-Aceña, Lance Davis, Marc Flandreau, Lars Jonung, Sophie Lazaretou, Augustin Llona Rodriguez, Gerardo della Paollera, David Pope, Angela Redish, Jaime Reis, Georg Rich, Fernando Santos. The usual disclaimer holds. 1 The Currency School in England in the early nineteenth century made the case for the Bank of England’s fiduciary note issue to vary automatically with the level of the Bank’s gold reserve (the currency principle). Following such a rule was viewed as preferable (for providing price-level stability) to allowing the note issue to be altered at the discretion of the well-meaning and possibly well-informed directors of the Bank (the position taken by the opposing Banking School). For a discussion of the Currency Banking School debate, see Viner (1937), Fetter (1965), and Schwartz (1987). 2 Viewed, however, as a rule in the traditional sense—as an automatic mechanism to ensure price stability—bimetallism may have had greater scope for automaticity than the gold standard, because of the additional cushion of a switch from one metal to the other. See Friedman (1990). 3 This description is consistent with a result from a model of Lucas and Stokey (1983), in which financing of wars is a contingency rule that is optimal. In their example, where the occurrence and duration of the war are uncertain, the optimal plan for the debt is not to service it during the war. Under this policy, people realize when they purchase the debt that effectively it will be defaulted on in the event the war continues. 4 A case study comparing British and French finances during the Napoleonic Wars shows that Britain was able to finance its wartime expenditures by a combination of taxes, debt, and paper money issue—to smooth revenue; whereas France had to rely primarily on taxation. France had to rely on a less efficient mix of finance than Britain because she had used up her credibility by defaulting on outstanding debt at the end of the American Revolutionary War and by hyper-inflating during the Revolution. Napoleon ultimately returned France to the bimetallic standard in 1803 as part of a policy to restore fiscal probity, but because of the previous loss of reputation, France was unable to take advantage of the contingent aspect of the bimetallic standard rule. See Bordo and White (1993). 5 The behavior of asset prices (exchange rates and interest rates) during suspension periods suggests that market agents viewed the commitment to gold as credible. For the United States see Roll (1972) and Calomiris (1988), who present evidence of expected appreciation of the greenback during the American Civil War based on a negative interest differential between bonds that were paid in greenbacks and

OPERATION OF THE SPECIE STANDARD 75

6 7

8

9

10 11

12

13

14

15

those paid in gold. Also, see Smith and Smith (1993) who demonstrate that movements in the premium on gold from the Resumption Act of 1875 until resumption was established in 1878 were driven by a credible belief that resumption would occur. For the case of Britain’s return to gold in 1925, see Smith and Smith (1990), Miller and Sutherland (1992) and (1994). An application of the stochastic process switching literature suggests that the increasing likelihood that resumption would occur at the original parity gradually altered the path of the dollarsterling exchange rate towards the new ceiling, several months in advance. For suggestive evidence see Bordo and Kydland (1996). Also see Officer (1993). His calculations of speculative bands (bands within which uncovered interest arbitrage prevails consistent with gold point arbitrage efficiency) for the interwar dollar-sterling exchange rate show serious violations only in 1931, at the very end of the gold exchange standard. A case study of Canada during the Great Depression provides evidence for the importance of the credible commitment mechanism of adherence to gold. Canada suspended the gold standard in 1929 but did not allow the Canadian dollar to depreciate nor the price level to rise for two years. Canada did not take advantage of the suspension to emerge from the depression, because of concern for its credibility with foreign lenders. See Bordo and Redish (1990). Eschweiler and Bordo (1993) provide evidence for this interpretation for Germany over the period 1883–1913 based on an estimation of the Reichsbank’s reaction function. They find that the central bank’s pursuit of an interest rate smoothing policy (an obvious violation of ‘the rules of the game’) was subordinate to its commitment to keep the exchange rate within the gold points. The United States could change the dollar price of gold if a majority of members (and every member with 10 per cent or more of the total quotas) agreed. Indeed, Giovannini’s (1993) calculations show that during the Bretton Woods convertible period credibility bounds on interest rates for the major currencies, in contrast to the classical gold standard, were frequently violated. In 1869 institutions to supply capital to new firms and industries by issuing notes convertible into government notes were established in eight commercial cities; only the one that issued notes convertible into silver coin found public acceptance. Despite the monetary disarray, the Meiji government succeeded in floating two foreign loans in London in 1871 and 1873, the first for £1 million 9 per cent, for construction of a railroad between Tokyo and Yokohama, the second for £3.4 million pounds at 7 per cent, to pay pensions to feudal lords and soldiers, the feudal clans having been abolished in 1870 by the new national government. During the same years, England, France, and Russia sold their bonds in Japan, paid for by gold exports from Japan. Although Chile’s experience of borrowing funds in sterling at prevailing rates in a period when the currency was inconvertible does not seem to support this statement (see Llona-Rodriguez, 1993, n. 9). Use of capital calls on the new securities issued is superior to the new foreign lending series used in earlier studies (e.g. Edelstein, 1982). This is because the capital calls give the amount of funds actually available to send to the receiving countries, whereas the new foreign lending series represents the face value of the bonds. The amounts actually available were only a fraction of the total issue which

76 OVERVIEW

16 17 18 19 20 21

22 23 24

25 26

the investors committed themselves to provide over a period of years during which calls were made (see Davis and Huttenback, 1986, ch.1). A real series deflated by the British Sauerbeck-Statist wholesale price index is very similar to the current dollar series presented here. For earlier applications of similar regime comparisons in a different context see Bordo (1993a), (1993b) and Bordo and Jonung (1996). In Bordo and Schwartz (1994) we also present data for real per capita growth, and the absolute rates of change of nominal and real exchange rates. The data sources for Figure 2.3 and all subsequent figures are listed in the Data appendix. Both variables were rendered stationary by first differencing. Specifically, four restrictions are placed on the matrix of the shocks: two are simple normalizations, which define the variances of the shocks to aggregate demand and aggregate supply; the third assumes that demand and supply shocks are orthogonal; the fourth is that demand shocks have only temporary effects on output, i.e., that the cumulative effect of demand shocks on the rate of change in output must be zero. See Keating and Nye (1991). Figures of the shocks for each of the 21 countries are available upon request. The results for most country groupings in the interwar period figures are similar to those reported for the United States by Cecchetti and Karras (1992), who estimated a three-variable VAR with monthly data. The late 1920s and early 1930s reveal a major negative demand shock consistent with Friedman and Schwartz’s (1963) attribution of the onset of the Great Depression to monetary forces. After 1931, negative supply shocks predominate, consistent with Bernanke’s (1983) and Bernanke and James’ (1991) explanation for the severity of the Great Depression that stresses the collapse of the financial system. The rankings by regime for the weighted average of individual country shocks are similar to the group aggregates. These results are very similar to those presented for the G-7 in Bordo (1993b) and by Bayoumi and Eichengreen (1994a b).

DATA APPENDIX For the G-10 countries plus Switzerland, see data appendix in Bordo and Jonung (1996) and data appendix in Bordo (1993a) except for the following countries: Australia

(1) Money 1880–1990, provided by David Pope, Australian National University. (2) Real GDP 1880–1990, David Pope. (3) GDP deflator 1880–1990, David Pope. (4) Exchange rate 1880–1990, David Pope. (5) Government expenditures and revenues 1880–1990, David Pope. Denmark

(1) Money 1880–1990, provided by Lars Jonung. (2) Real GDP 1880–1990, Lars Jonung. (3) GDP deflator 1880–1988, B.R.Mitchell (1992). (4) Exchange rate

OPERATION OF THE SPECIE STANDARD 77

1880–1990, Lars Jonung. (5) Government expenditures and revenues 1880–1988, B.R.Mitchell (1992). Finland

(1) Money 1880–1990, Lars Jonung. (2) Real GDP 1880–1990, Lars Jonung. (3) GDP deflator 1880–1988, B.R.Mitchell (1992). (4) Exchange rate 1880–1990, Lars Jonung. (5) Government expenditures and revenues 1880–1988, B.R.Mitchell (1992). Greece

(1) Money 1880–1914, Sophia Lazaretou; 1962–90, International Financial Statistics (1992). (2) Real GDP 1927–88, B.R.Mitchell (1992). (3) GDP Deflator 1927–88, B.R.Mitchell (1992). (4) Exchange rate 1880–1914, Sophia Lazaretou; 1962–90, International Financial Statistics (1992). (5) Government expenditures and revenues 1880–1936, Sophia Lazaretou; 1937–39, B.R.Mitchell (1992); 1956–86, B.R.Mitchell (1992). Norway

(1) Money 1880–1990, Lars Jonung. (2) Real GDP 1880–1990, Lars Jonung. (3) GDP deflator 1880–1988, B.R.Mitchell (1992). (4) Exchange rate 1880–1990, Lars Jonung. (5) Government expenditures and revenues 1880–1988, B.R.Mitchell (1992). Portugal

(1) Money 1880–1990, provided by Fernando Santos, Universidade do Porto. (2) Real GDP 1880–1990, Lars Jonung. (3) GDP deflator 1880–1988, B.R.Mitchell (1992). (4) Exchange rate 1890–1990, Fernando Santos. (5) Government expenditures and revenues 1880–1988, B.R.Mitchell (1992). Spain

(1) Money 1880–1990, Estadisticas Historicas de Espana; 1981–88, B.R. Mitchell (1992). (2) Real GDP 1901–53, Estadisticas Historicas de Espana; 1954–88, B.R.Mitchell (1992). (3) GDP deflator, 1901–53, B.R.Mitchell (1992). (4) Exchange rate 1880–1980. Estadisticas Historicas de Espana; 1981–90, IFS (1992). Government expenditures and revenues 1880–1988, B.R.Mitchell (1992). REFERENCES Abramowitz, Moses. 1973. The Monetary Side of Long Swings in U.S. Economic Growth. Stanford University Center for Research on Economic Growth. Memorandum No. 146 (mimeo). Barro, Robert J. and David B.Gordon. 1983. Rules, Discretion and Reputation in a Model of Monetary Policy. Journal of Monetary Economics 12:101–21.

78 OVERVIEW

Bayoumi, Tamin and Barry Eichengreen. 1992b. Is There a Conflict Between EC Enlargement and European Monetary Unification? NBER Working Paper No. 3950. January. Greek Economic Review (forthcoming). Bayoumi, Tamin and Barry Eichengreen. 1993. Shocking Aspects of European Monetary Unification. In Francesco Torres and Francesco Giavazzi (eds) Adjustment and Growth in European Monetary Union. Cambridge: Cambridge University Press. Bayoumi, Tamin and Barry Eichengreen. 1994a. Economic Performance Under Alternative Exchange Rate Regimes: Some Historical Evidence. In Peter B.Kenen, Francesco Papodia and Fabrizio Saccomani (eds) The International Monetary System. Cambridge: Cambridge University Press. Bayoumi, Tamin and Barry Eichengreen. 1994b. Monetary and Exchange Rate Arrangements for NAFTA. Journal of Development Economics 43:125–65. Bernanke, Benjamin. 1983. Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression. American Economic Review 73:259–76. Bernanke, Benjamin and Harold James. 1991. The Gold Standard, Deflation and Financial Crisis in the Great Depression: An International Comparison. In Financial Markets and Financial Crisis (ed.) R.Glenn Hubbard, Chicago: University of Chicago Press: 33–68. Blanchard, Olivier and Danny Quah. 1989. The Dynamic Effects of Aggregate Demand and Aggregate Supply Disturbances. American Economic Review (September): 655– 73. Bloomfield, Arthur. 1959. Monetary Policy Under the International Gold Standard, 1800– 1914. New York: Federal Reserve Bank of New York. Bordo, Michael D. 1981. The Classical Gold Standard: Some Lessons for Today. Federal Reserve Bank of St. Louis Review 63 (May): 2–17. Bordo, Michael D. 1984. The Gold Standard: The Traditional Approach. In A Retrospective on the Classical Gold Standard, 1821–1931 Michael D.Bordo and Anna J.Schwartz. (eds) Chicago: University of Chicago Press. Bordo, Michael D. 1993a. The Bretton Woods International Monetary System: An Historical Overview. In Michael D.Bordo and Barry Eichengreen (eds) A Retrospective on the Bretton Woods System: Lessons for International Monetary Reform. Chicago: University of Chicago Press. Bordo, Michael D. 1993b. The Gold Standard, Bretton Woods and Other Monetary Regimes: A Historical Appraisal. Federal Reserve Bank of St. Louis Review 75–2 (March/April): 123–91. Bordo, Michael D. and Lars Jonung. 1996. Monetary Regimes, Inflation and Monetary Reform. In D.F.Vaz and K.Vellupillai (eds) Inflation, Institutions and Information. Essays in Honor of Axel Leijonhufvud. London: Macmillan. Bordo, Michael D. and Finn E.Kydland. 1996. The Gold Standard As a Commitment Mechanism. In Tamin Bayoumi, Barry Eichengreen and Mark Taylor (eds) Modern Perspectives on the Gold Standard. Cambridge: Cambridge University Press. Bordo, Michael D. and Angela Redish. 1990. Credible Commitment and Exchange Rate Stability: Canada’s Interwar Experience. Canadian Journal of Economics 23(2) 357– 80. Bordo, Michael D. and Anna J.Schwartz. 1994. The Specie Standard As a Contingent Rule: Some Evidence for Core and Peripheral Countries, 1880–1990. Rutgers University Working Paper No. 94–11.

OPERATION OF THE SPECIE STANDARD 79

Bordo, Michael D. and Eugene N.White. 1993. British and French Finance During the Napoleonic Wars. In Michael D.Bordo and Forrest Capie (eds) Monetary Regimes in Transition. Cambridge: Cambridge University Press. Butlin, S.J. 1986. The Australian Monetary System 1851–1914. Sydney Australian Reserve Bank. Calomiris, Charles W. 1988. Price and Exchange Rate Determination During the Greenback Suspension, Oxford Economic Papers December. Calomiris, Charles. 1993. Greenback Resumption and Silver Risk: The Economics and Politics of Monetary Regime Change in the United States, 1862–1900. In Michael D.Bordo and Forrest Capie (eds) Monetary Regimes in Transition Cambridge: Cambridge University Press. Canzoneri, Matthew. 1985. Monetary Policy Games and the Role of Private Information. American Economic Review 75:1056–70. Canzoneri, Matthew B. and Dale W.Henderson. 1991. Monetary Policy in Interdependent Economies. Cambridge: Massachusetts Institute of Technology Press. Cecchetti, Stephen G. and Georgios Karras. 1992. Sources of Output Fluctuations During the Interwar Period: Further Evidence on the Causes of the Great Depression. NBER Working Paper No. 4049 April. Cooper, Richard. 1982. The Gold Standard: Historical Facts and Future Prospects. Brookings Papers on Economic Activity 1:1–45. Cortés-Condé, Roberto. 1989. Dinero, Deuda y Crisis: Evolución Fiscal y Monetaria en la Argentina. Editorial Sudamericana, Instituto Torcuato Di Tella: Buenos Aires. Davis, Lance E. and Robert A.Huttenback. 1986. Mammon and the Pursuit of Empire: The Political Economy of British Imperialism, 1860–1912. Cambridge: Cambridge University Press. de Cecco, Marcello. 1974. Money and Empire: The International Gold Standard, 1890– 1914. New Jersey: Rowman and Littlefield. DeKock, Gabriel and Vittorio Grilli. 1989. Endogenous Exchange Rate Regime Switches. NBER Working Paper No. 3066 August. DeVries, Margaret G. 1976. The International Monetary Fund 1966–1971: The System Under Stress, Vol. 1: Narrative. Washington, DC: International Monetary Fund. Edelstein, Michael. 1982. Overseas Investment in the Age of High Imperialism: The United Kingdom, 1850–1914. New York: Columbia University Press. Eichengreen, Barry. 1985. Editor’s Introduction. In Barry Eichengreen (ed.) The Gold Standard in Theory and History. London: Methuen. Eichengreen, Barry. 1987. Conducting the International Orchestra: Bank of England Leadership Under the Classical Gold Standard. Journal of International Money and Finance (6): 5–29. Eichengreen, Barry. 1989. Hegemonic Stability Theories. In Richard Cooper et. al. (eds) Can Nations Agree? Washington, DC: Brookings Institution. Eichengreen, Barry. 1992a. The Gold Standard Since Alec Ford. In S.N.Broadberry and N.F.R.Crafts (eds) Britain in the International Economy: 1870–1939 Cambridge: Cambridge University Press. Eichengreen, Barry. 1992b. Golden Fetters: The Gold Standard and the Great Depression, 1919–1939. Oxford University. Eichengreen, Barry. 1993. Three Perspectives on the Bretton Woods System. In Michael D.Bordo and Barry Eichengreen (eds) A Retrospective on the Bretton Woods System. Chicago: University of Chicago Press and NBER.

80 OVERVIEW

Eichengreen, Barry. 1994. International Monetary Arrangements for the 21st Century. Washington, DC: Brookings Institution. Eschweiler, Bernhard and Michael D.Bordo. 1993. Rules, Discretion and Central Bank Independence: The German Experience 1880–1989. NBER Working Paper No. 4547. Fetter, F. 1965. Development of British Monetary Orthodoxy, 1797–1875. Cambridge: Harvard University Press. Fishlow, Albert. 1987. Market Forces or Group Interests: Inconvertible Currency in Pre-1914 Latin America. University of California at Berkeley (mimeo). Fishlow, Albert. 1989. Conditionality and Willingness to Pay: Some Parallels from the 1989s. In Barry Eichengreen and Peter Lindert (eds) The International Debt Crisis in Historical Perspective. Cambridge: Massachusetts Institute of Technology Press. Flood, Robert P. and Peter Isard. 1989. Simple Rules, Discretion and Monetary Policy. NBER Working Paper No. 2934. Ford, A.G. 1962. The Gold Standard 1880–1914: Britain and Argentina. Oxford: Clarendon Press. Fratianni, M. and Spinelli, F. 1984. Italy in the Gold Standard Period, 1861–1914. In Michael D.Bordo and Anna J.Schwartz (eds) A Retrospective on the Classical Gold Standard, 1921–1931. Chicago: University of Chicago Press. Frieden, Jeffrey A. 1993. The Dynamics of International Monetary Systems: International and Domestic Factors in the Rise, Reign, and Demise of the Classical Gold Standard. In Jack Snyder and Robert Jervis (eds) Coping with Complexivity in the International System. Colorado: Westview Press. Friedman, Milton. 1990. Bimetallism Revisited. Journal of Economic Perspectives 4(4): 85–104. Friedman, Milton and Schwartz, Anna J. 1963. A Monetary History of the United States, 1867–1960. Princeton: Princeton University Press. Fritsch, Winston and Gustavo H.B.Franco. 1992. Aspects of the Brazilian Experience Under the Gold Standard. PUC Rio de Janeiro (mimeo). Gallarotti, Giulio, M. 1993. The Scramble for Gold: Monetary Regime Transformation in the 1870s. In Michael D.Bordo and Forrest Capie (eds) Monetary Regimes in Transition, Cambridge: Cambridge Univerity Press. Giovannini, Alberto. 1989. How Do Fixed Exchange-Rate Regimes Work: The Evidence From the Gold Standard, Bretton Woods and the EMS. In Marvin Miller, Barry Eichengreen and Richard Portes (eds) Blueprints for Exchange Rate Management. London: Centre for Economic Policy Research: 13–46. Giovannini, Alberto. 1993. Bretton Woods and Its Precursors: Rules Versus Discretion in the History of International Monetary Regimes. In Michael D.Bordo and Barry Eichengreen (eds) A Retrospective on the Bretton Woods System. Chicago: University of Chicago Press. Grilli, Vittorio. 1990. Managing Exchange Rate Crises: Evidence from the 1890’s. Journal of International Money and Finance 9:258–275. Grossman, Herschel J. and John B.Van Huyck. 1988. Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation, and Reputation. American Economic Review 78:1088–97. Haavisto, Tarmo. 1992. Money and Economic Activity in Finland 1866–1985. Lund, Sweden. Lund Economic Studies. Hayashi, Fumio. 1989. Japan’s Saving Rate: New Data and Reflections. NBER Working Paper No. 3205.

OPERATION OF THE SPECIE STANDARD 81

Horsefield, Keith. 1969. The International Monetary Fund, 1945–1965: Twenty Years of International Monetary Co-operation, Vol. 1, Chronicle. Washington, DC: International Monetary Fund. Jonung, Lars. 1984. Swedish Experience Under the Classical Gold Standard, 1873–1914. In Michael D.Bordo and Anna J.Schwartz (eds) A Retrospective on the Classical Gold Standard, 1821–1931. Chicago: University of Chicago Press. Keating, John W. and John V.Nye. 1991. Permanent and Transitory Shocks in Real Output: Estimates from Nineteenth Century and Postwar Economies. St. Louis: Washington University Working Paper No. 160. Keynes, John Maynard. [1930] 1971. The Applied Theory of Money: A Treatise on Money, Vol. 6 of The Collected Writings of John Maynard Keynes. Reprint London: Macmillan and New York: Cambridge University Press for the Royal Economic Society. Krugman, Paul. 1991. Target Zones and Exchange Rate Dynamics. Quarterly Journal of Economics 56:669–82. Kydland, Finn E. and Prescott, Edward. 1977. Rules Rather than Discretion: The Inconsistency of Optimal Plans. Journal of Political Economy 85:473–91. Lazaretou, Sophia. 1994. Government Spending, Monetary Policies and Exchange Rate Regime Switches: The Drachma in the Gold Standard Period. Explorations in Economic History, October. Lindert, Peter. 1969. Key Currencies and Gold, 1900–1913. Princeton Studies in International Finance, No. 24. Princeton: Princeton University Press. Llona-Rodriguez, Augustine. 1993. Chile During the Gold Standard: A Successful Paper Money Experience. Instituto Torcuato Di Tella (mimeo). Lucas, Robert E. Jr. and Nancy L.Stokey. 1983. Optimal Fiscal and Monetary Policy in an Economy Without Capital. Journal of Monetary Economics 12: (1)55–93. McKinnon, Ronald I. 1988. An International Gold Standard Without Gold. Cato Journal 8 (Fall):351–73. McKinnon, Ronald I. 1993. International Money in Historical Perspective. Journal of Economic Literature 31(1): 1–44. Mankiw, Gregory. 1987. The Optimal Collection of Seigniorage—Theory and Evidence. Journal of Monetary Economics 20: (2)327–41. Martin-Aceña, Pablo. 1993. Spain During the Classical Gold Standard Years, 1880–1914. In Michael D.Bordo and Forrest Capie (eds) Monetary Regimes in Transition. Cambridge: Cambridge University Press. Meltzer, Allan H. and Saranna Robinson. 1989. Stability Under the Gold Standard in Practice. In Michael D.Bordo (ed.) Monetary, History and International Finance: Essays in Honor of Anna J.Schwartz. Chicago: University of Chicago Press: 163–95. Miller, Marcus and Alan Sutherland. 1992. Britain’s Return to Gold and Entry into the ERM. In Paul Krugman and Marcus Miller (eds) Exchange Rate Targets and Currency Banks. Cambridge: Cambridge University Press. Miller, Marcus and Alan Sutherland. 1994. Speculative Anticipations of Sterling’s Return to Gold: Was Keynes Wrong? Economic Journal. July. Mitchell, Brian R. 1992. International Historical Statistics: Europe 1750–1988. New York: Stockton Press. Obstfeld, Maurice. 1992. Destabilizing Effects of Exchange Rate Escape Clauses. NBER Working Paper No. 3606.

82 OVERVIEW

Officer, Lawrence. 1986. The Efficiency of the Dollar-Sterling Gold Standard, 1980– 1908. Journal of Political Economy 94 (October): 1038–73. Officer, Lawrence. 1993. Gold-Point Arbitrage and Uncovered Interest Arbitrage Under the 1925–1931 Dollar—Sterling Gold Standard. Explorations in Economic History 30 (1): 98–127. Pelaez, Carlos M. and Wilson Suzigan. 1976. Historia Monetaria do Brazil. Editoria Universidade de Brazilia. Reis, Jaime. 1992. The Gold Standard in Portugal, 1854–1891. Universidale Nova de Lisbon (mimeo). Roll, Richard. 1972. Interest Rates and Price Expectations During the Civil War. Journal of Economic History 32 (June): 476–98. Schwartz, Anna J. 1984. Introduction. In Michael D.Bordo and Anna J.Schwartz (eds) A Retrospective on the Classical Gold Standard, 1821–1931. Chicago: University of Chicago Press. Schwartz, Anna J 1987. ‘Banking School, Currency School, Free Banking School.’ In New Palgrave Dictionary of Economics. London: Macmillan Sharkey, R.D. 1959. Money, Class, and Party. Baltimore: Johns Hopkins Press. Shearer, Ronald A. and Carolyn Clark. 1984. Canada and the Interwar Gold Standard, 1920–1935: Monetary Policy Without a Central Bank. In Michael D. Bordo and Anna J.Schwartz (eds) A Retrospective on the Classical Gold Standard 1821–1931. Chicago: University of Chicago Press. Shinjo, Hiroshi. 1962. History of the Yen. Research Institute for Economics and Business Administration. Japan: Kobe University. Simons, Henry C. 1951. Rules Versus Authorities in Monetary Policy. In Richard D.Irwin (ed.) Readings in Monetary Theory. Homewood, Illinois. Smith, Gregor and Todd Smith. 1993. Wesley Mitchell and Irving Fisher and the Greenback Gold Reforms 1865–1879. Queens University (mimeo). Smith, W.S. and Smith, R.T. 1990. Stochastic Process Switching and the Return to Gold. Economic Journal 100 (March): 164–75. Stiglitz, Joseph and Andrew Weiss. 1981. Credit Rationing in Markets with Imperfect Information. American Economic Review 71 (6): 393–410. Svennson, Lars E.O. 1994. Why Exchange Rate Bands? Monetary Independence In Spite of Fixed Exchange Rates. Journal of Monetary Economics 33 (1): 157–99. Unger, I. 1964. The Greenback Era: A Social and Political History of American Finance, 1865–1879. New Jersey: Princeton University Press. Viner, Jacob. 1937. Studies in Theory of International Trade. Chicago: University of Chicago Press. Yeager, Leland B. 1984. The Image of the Gold Standard. In Michael D.Bordo and Anna J.Schwartz (eds) A Retrospective on the Classical Gold Standard, 1821–1931. Chicago: University of Chicago Press.

Part II MYTHS AND REALITIES OF THE GOLD STANDARD

84

3 THE ORIGINS OF THE GOLD STANDARD Alan S.Milward

It is a puzzling fact that from the mid-1870s an increasing number of countries valued their national currencies by one method or another against gold; puzzling, because convincing rational explanations are lacking. In the 1920s an allegedly rational explanation temporarily prevailed: that fixing a gold value for the currency maintained stable exchange rates and through the primacy which it gave to the currency ‘s external value ensured the stability of internal prices. It is true that one characteristic of the gold standard period was that a prime aim of national government policy in the more developed economies was to preserve the value of the currency, because that was what their narrow electorates demanded, at least before 1900. But there was much more political dispute about this in those less developed economies which also went on the gold standard and, furthermore, the tendency of historical research has been to show that even in the developed economies, central banks did not consistently make a priority of preserving stable exchange rates (Ford, 1989). Compared to today’s institutions, they were also relatively indifferent to problems of managing the international economy or of its impact on the domestic economy; there was very little central bank cooperation. The preferred rational explanation of the 1920s can now be seen mainly as romantic nostalgia for an epoch of relatively stable exchange rates and much more stable domestic prices than the wildly inflationary aftermath of World War I. Three general categories of explanation for this move to a gold standard now seem to exist. Separating them, however, is an arbitrary process. They overlap, and the most authoritative work on the subject, that of Mertens (1944), achieves its authority by combining elements of each of them. Indeed, the less complex the explanations are, the more they seem to reflect the theoretical or historical prejudices of their authors. Nevertheless, a predisposition in favor of one primary motivation as the explanation for going on the gold standard can usually be discerned even in the more complicated accounts, so that the categorization which follows is not merely arbitrary. The oldest explanation, particularly favored by monetary theorists, relates the origins of the gold standard to changes in the relative supply of the two precious metals, gold and silver. An increase in the supply of silver relative to gold

86 MYTHS AND REALITIES OF THE GOLD STANDARD

starting in the 1860s and continuing until 1900 forced, it is argued, countries interested in stable prices and currency values to switch from a silver standard to a gold standard, or where bimetallic standards were operated, to tie the currency effectively to gold. However, evidence about changes in the relative supplies of the two metals, particularly when juxtaposed with the timing of the move to gold by developed economies, does not do much to support this argument. It is more suggestive indeed of the counter-argument, that the relative prices of silver and gold were determined more by demand than supply. Of course, any argument about the relative prices of the two metals also has its basis in a political economy of the choice between them. Whether the fall in the price of silver relative to gold was supply or demand induced, the choice of gold was presumably in the context of the argument made by political groups who desired a stable currency. Only the rather absurd argument that silver fell so far in relative value as to be no longer functional as a means of payment, because large payments required too much of it, can avoid the issue of political choice. Some authors, most notably de Cecco (1974), have explained the origins of the gold standard by constructing a political economy of choice which tries to answer this question for all the countries involved. Exporters of primary products, he argues, preferred silver because falling prices benefited their trade without reducing their income, not least because they negated most of the effect of the ad valorem tariffs imposed on foodstuff imports into Western European countries from the mid-1870s onwards. Gold, by contrast, became the choice of the manufacturing and financial sectors, and where that predominated politically the gold standard was imposed. Certainly, historians would have no quarrel with the basis of de Cecco’s political economy of choice, and in its essentials it has stood the test of research. But as evidence accumulates about the choice of gold in less developed economies it appears, as no doubt do all political economies, too simplistic to explain individual cases, and also tied too closely to a materialistic calculus of group interest. Where the relative power of group interests was ultimately determined by general elections, as in Italy and the United States, this type of analysis remains the indispensable starting-point. If, however, we turn to Russia or Japan, something more complex is required from the outset. De Cecco’s work, however, had the additional merit of insisting on the importance of the particular set of international political circumstances in which the gold standard became predominant. A mere political economy of metallic choice was not enough; the gold standard was also formed out of the interplay of the European great powers after 1870. A third category of explanation sees this specific diplomatic set of circumstances as the primary, rather than an important subsidiary, origin of the gold standard. Thus Flandreau’s (1993) sophisticated analysis finally locates the origins of the gold standard in the defeat of France in the Franco-Prussian war and the German Empire’s use of the war indemnity to create a gold-based German national currency. Far from being a rational choice either from the stand-point of monetary stability or changing balances of

ORIGINS OF THE GOLD STANDARD 87

domestic political power, he implies, the origins of the gold standard are to be found firstly in a series of irrational confrontational acts of policy in France and Germany after 1870. In one sense therefore the spectrum of choice is very wide between the gold standard as the defining mark of a historical epoch, with its causes deeply rooted in ineluctable patterns of economic and social change, and the gold standard as an accident. Much of the economic literature avoids these issues by dealing with the gold standard as an economic convenience, with implicitly rational origins related to efficiency choices, and testing it for efficiency in the hope of refining the nature of the efficiency choice. As soon, however, as the question is raised of what sort of an economic convenience it was, the historical specifics of the period force themselves back into the discussion as ‘externalities’. Did its advantage lie in the fact that it was technically an easy form of convertibility, that it facilitated commodity trade, that it seemed the surest guarantee of a fixed rate, or that for whatever reason, it was feared that silver would become ‘bad’ money? All of these issues were indeed contingent to the decision to adopt a gold standard in those countries where its adoption has been more thoroughly studied. But thorough studies are rare, and for less developed economies almost entirely lacking. On the present state of the evidence it is not possible to do more than eliminate a certain measure of false historical argument and suggest more promising lines of enquiry. In the first place, the evidence that nineteenth-century governments so deluded themselves as to suppose that any metallic standard of value by itself would lead to long-run price or exchange-rate stability is lacking. The myth of gold as a stable unit of value probably arose because in so many countries emerging industrial and financial interest groups wanted to combine rapid development with stable monetary values and hoped, rather than believed, that gold had certain temporary advantages in this respect. The illusion that this had actually proved possible was, however, peculiar to their offspring after 1918. Although the falling value of silver relative to gold did at times strongly influence rhetoric on the respective advantages of different metallic standards, the origins of the gold standard in most countries are to be found in two other more realistic motivations. One was the increasing advantages which a gold standard offered for foreign commercial policy. The other lay in its economic and political relationship to economic development; it made it easier to secure development capital and it satisfied at the same time, partly symbolically, the strong, if vague, aspirations of ‘modernization.’ Contrary to much of what was written later, the choice of gold was not inspired by the constraints on domestic monetary policy which it theoretically implied, but by the promise of increased foreign trade, and in less developed economies the increase in foreign capital inflows which it seemed to offer. That, no doubt, is the chief reason why studies of the operation of the gold standard in single countries usually end up by acknowledging that central banks did very little, sometimes nothing, to enforce the rules by which economists and

88 MYTHS AND REALITIES OF THE GOLD STANDARD

historians explain that it operated (Goodhart, 1972). While government was concerned to maintain the value of money, other and often conflicting pressures also operated on it from traders and investors. In the case of laggard economies in the development process these other pressures were more important, and because they represented not only the hope of financial gain but also psychological attitudes closely linked to national and nationalist aspirations, they could become irresistible. The choice of the gold standard was not, in short, in the international economy as a whole, a conservative but a dynamic one. This does not mean that the gold standard was necessarily the choice of exporters because of its advantages as a system of currency convertibility in a period when gold did retain a relative stability of value. Exporters in laggard economies were, as de Cecco percipiently pointed out, as likely as not to be conservative landowners, whose exports and incomes might seem to have better prospects through the devaluationary impact of remaining on a silver standard when silver was falling in value against gold and thus against the currencies of major importers, like the United Kingdom and Germany. Rather, the gold standard was the choice of those two wished to shift the export structure towards what was thought of as modernity, manufactures which rivalled those of the United Kingdom and of France, and this in turn was but part of a general political aspiration towards a society which resembled those models. Developmental aspirations, and the desire to copy those societies where they seemed to have been realized, could overcome, partly through their emotional appeal, political power structures dominated by conservative elites, whose preference as international traders was for silver as the basis of international convertibility. It is only the powerful emotional symbolism of these aspirations and the way they became linked to gold which can fully explain the explosive and prolonged nature of the arguments over bimetallism, whose vehemence seems so strange less than a century later. Fully considered, the adoption of the gold standard is therefore less a story of precious metals and their relative values than of the intellectual, social and political change within nations brought about by nineteenth-century economic development. The first category of explanation of the gold standard ‘s origins seems therefore altogether unsatisfactory as an aid to understanding. Widely coined since its massive importation into Europe in the sixteenth century, silver was still in the early nineteenth century the money of international trade within Europe and between Europe and Asia. With what precise timing gold came to replace it for many of the trades in question remains a matter of some obscurity, but the change was directly related to the pivotal position which the United Kingdom came to occupy by the 1860s in world trade, world shipping, and the financing of both on the London money market. The United Kingdom’s currency was on a gold standard and, at least when times were smooth, was redeemable in gold at the central bank. The obvious importance of worldwide foreign trade to the modernization and industrialization of the British economy after 1815 made the gold standard seem an effective choice, rather than the eccentricity which it had

ORIGINS OF THE GOLD STANDARD 89

seemed earlier. But this phenomenon has to be seen in the wider political perspective in which from the 1840s onwards the United Kingdom also began to replace France as the model of a politically developed society. Aspirations to emulate the United Kingdom were strongly influential in determining the relative demand for the two precious metals. The fall in the price of silver in the nineteenth century was attributable not so much to excess supply as to falling demand. The increase in the supply of silver between 1815 and 1880 was in fact somewhat less than that of gold. Indeed, the gold strike in California in 1849 induced a greater, although brief, downward fluctuation in the price of gold than any similar movement in silver prices before 1859, the year in which silver reached its maximum price for the century in terms of gold. It was the instability of gold which seems to have played a part in inducing some governments—Belgium, the Neapolitan Kingdom, Spain and Switzerland—to bow before pressure from French imperialism to standardize their currencies on the French silver franc and even to demonetize gold. In the 1850s, when the French economy developed so rapidly and when France was the main source of development capital for the rest of the continent, emulation of the United Kingdom’s gold standard was confined to Portugal, and as Reis shows in Chapter 6 the motivation for Portugal’s choice was very complex. From 1859 a clear, although at first only gradual, downward trend of the price of silver both absolutely and relative to that of gold can be observed. This accelerated sharply from 1872 to 1876, and in the latter year silver fell relative to gold by almost 8 per cent. It held its value at that low plateau until the end of the decade, fell again over the next decade at a rate comparable to that of 1872–76, and then between 1890 and 1900 fell by more than 40 per cent. By that last decade most major Western countries had begun to hoard gold in their reserves and few, the major exception being Spain, remained on a silver standard. No one has doubted that the shortfall in demand explains the steep fall in price after 1890, the reason for the fall in demand being so publicly evident. What is more complex is to explain the fall in demand for silver before 1876 and the growing preference for gold. The fall in the price of silver between 1872 and 1876 did of course play its part in the choice of a metallic standard. These, however, were the years when Germany and Scandinavia opted for a gold standard and the fall in the demand for silver was mainly caused by the contemporary public clamor for its replacement in Germany. The origins of this clamor are to be found in a mixture of political emotions relating to national unification and pressure for constitutional rule by the middle class, a constitutional demand with strong echoes in Sweden and elsewhere. The resonance of this debate was so loud because everyone recognized the momentous change in Europe’s balance of power signified by the fact that the constitutional unity of Germany was erected on the defeat of French imperialism. It was, furthermore, French imperialism which over the previous two decades had imposed the bimetallic silver standard on much of Western Europe. The resonance was also loud because of the

90 MYTHS AND REALITIES OF THE GOLD STANDARD

remarkable volume of publications on the subject of metallic standards— pamphlets, speeches, treatises and even sermons—which accompanied the debate in Germany. What was being debated over the period 1870–5 was, however, essentially the question of what sort of constitution the German Empire would have. Who would rule it was the issue, and the choice of a metallic standard was seen by most pamphleteers and by most politicians in the Reichstag as contributory to that greater question. The choice of the gold standard was seen in terms of adherence to the worldwide liberal trading system, of which free-trade Britain was the central pillar, and thus also as a guarantee in some sense of a liberal middle-class constitutional and legal order. It marked also the renunciation by liberal constitutionalists of their earlier hopes of finding this guarantee in the construction of a united Western Europe. The expansion of Germany’s international commerce, the huge migration of Germans around the world, and the beginnings of German capital export were all moving German liberal politicians away from a narrow concentration on domestic issues towards defining their position on a world scale. The Napoleonic revolutionary idea of a united middle-class Europe, led and defended by French imperialism, so appealing in the 1850s, now not only seemed archaic, but its last emperor was living out his days in a peaceful London suburb entirely built and occupied by London’s new commercial middle class. It was the collapse of the Second Empire before German guns which made evident not only the defeat of Europe’s greatest military power, but also of its currency, the franc of 5 grammes of silver of 28 March 1803 (7 germinal ano XI), the U.S. dollar bill of its day. ‘In fact,’ as Mertens wrote (1944, p. 360) ‘the world adopted the gold-standard because it did not wish or know how to impose on itself the effort of organising bimetallism on an international basis. The gold-standard owed its success less to an objective and reasoned preference for the yellow metal than to restrictive measures against silver.’ Given the great difficulties of maintaining a gold standard, not the least of which was the inability of most countries to prevent an outflow of gold of such dimensions that the convertibility of the domestic currency into the gold against which it was denominated was always in question, a bimetallic standard might indeed have seemed a more rational solution. Not only that, but because of the massive political pressures from farmers in the United States and Europe for retaining some form of silver standard as an inducement to agricultural exports, it could be argued that a political consensus which would sustain the gold standard would have been untenable. In the 1880s two former governors of the Bank of England were in fact to attain a certain quiet notoriety by arguing that the survival of free trade depended on making a bimetallic standard work, because otherwise European agricultural tariffs would remain necessary (Gibbs and Greenfell, 1986). But these were the viewpoints which were swept aside in the German constitutional debate after 1870. Bimetallism had come to stand for the last attempt of French imperialism to impose a fixed exchange rate system on its European satellites; the gold standard

ORIGINS OF THE GOLD STANDARD 91

for a parliamentary monarchy, a middle-class legal order, national liberty including the liberty of exchange-rate choice, and ultimately worldwide economic power. The French nineteenth-century bimetallic standard saw little gold coin in circulation but did involve a fixed exchange rate between the silver franc and gold. With the downward trend in gold prices after 1849 the stream of French capital exports to Western Europe, which began to flow with the railwaybuilding boom in 1852, tended to be valued for repayment purposes against silver. In the 1850s the railway systems of Holland, Italy, Spain and Switzerland were laid down by French capital and engineering projectors, while much of the dynamic development of Rhineland banking, coal mining and iron production in the same decade also attracted heavy French investment. France was the source of development capital for the rapid European industrialization of the period and the extension of the continental transport network, as the United Kingdom was for the developing non-European world. While the French government looked benignly on these capital flows the competition was, nevertheless, to secure private capital from French banks, which raised the capital by retailing foreign state bonds and shares in foreign joint-stock companies to their depositors. In the competition to obtain capital, apparent security of repayment of interest was decisive and it was this above all which led the borrowers to align their own silver currencies on the French franc. This did not initially imply a bimetallic standard. The metallic standard was usually the official silver content of the 1803 franc. Thus Switzerland after its civil war decided on a nominally equivalent franc as its national currency. Belgium issued new silver franc coins in 1850 also of equivalent official silver content. Several Italian states followed suit and unified Italy did the same in 1862, except that the official silver content of the coin was marginally lower. This was going far beyond anything later attempted under the gold standard. It implied a fixed exchange rate system between the countries concerned, and when the price of silver began to fall relative to gold, it also implied a joint attempt to fix a silver-gold exchange rate. From the outset this Napoleonic system ran into difficulties whose resolution demanded a degree of common action, effectively subservience to the Bank of France, which sometimes went beyond anything the recipients of the loans would have wished. Much French coin was old and was, for example, exported to Belgium to purchase newer Belgian coin whose actual weight in silver was higher. When the Italian national currency was issued at a lower silver content it was exported to France to purchase French coin, The persistence of these problems, particularly the massive inflow of Italian silver coin into France, made it evident that a genuinely common monetary standard must be imposed. This was first attempted by the Latin Monetary Convention, signed between France, Belgium, Italy and Switzerland in 1865. Like all great powers, France was not prepared to alter its own monetary system to accommodate others. It therefore followed that the common standard had, like that of France, to be bimetallic. The Convention not only agreed on the standard

92 MYTHS AND REALITIES OF THE GOLD STANDARD

weight of the basic silver coin, but also to fix and keep a gold-silver exchange rate. It even specified the ratio of silver coinage to population in each country for the next 15 years, a concession to the monetarist argument that the fall in the price of silver was due to excess issue of money. This first attempt to manage a Western European fixed exchange rate system could not have come at a less propitious time. In developing Western European transport systems so extensively, French capital and technology had increased the commerce in goods and services between Western Europe and the world. With the sole exception of the 1950s there are no decades in which world trade expanded so rapidly as during the period 1850–70. The biggest part in the expansion was played by the growth of continental European exports, in the 1850s within the continent, in the 1860s outside it. The downfall of the silver franc as an easy unit of convertibility was thus the almost immediate result of the economic expansion of the 1860s, because the increasing commercial links with the extra-European world depended on London. It was in the London money market that extra-European trade was financed and, above all, in that market that the currencies were convertible. It followed that the London market set the rates, and it set them in gold. It was the rate against sterling, the worlds one major goldbacked currency, that mattered in extra-European trade, and it was in sterling that the bills of exchange which financed it in the 1860s were most typically inscribed. Ironically, on no member of the Latin Monetary Convention did this have greater effect than on France itself, and the Convention is best understood as a hopeless attempt to preserve a narrowly European monetary system which had already become pointless and unsustainable. It was French trade surpluses with the United Kingdom which allowed an international payments system to support the expansion of world commerce by the device of incorporating most-favorednation clauses in tariff agreements. Without France’s sterling earnings it would have been impossible for the rest of developing Europe to earn sufficient sterling to pay for its increasing imports from Britain and the extra-European world. In the 1860s the rise of the German states as an international trading entity disturbed this balance, while it still remained the case that for Germany, Paris was an important source of sterling acquired by exports to the French market. The assertively self-confident group of financiers around Napoleon III were well aware of the importance of the London market and its gold-standard to their own operations, and French financial commercial interests in general in the 1860s became increasingly doubtful about and lukewarm in their support for a foreign policy which sought to impose a bimetallic standard on a continental West European trading bloc. The close association of the Credit Mobilier with the foreign policy of the regime was not typical of the other finance houses by the mid-1860s. London maintained an open exchange market, with no attempt to fix the rate between silver and gold. As the increasing power of commercial and industrial interests in developing Western European countries pushed them away from

ORIGINS OF THE GOLD STANDARD 93

bimetallism and towards the gold standard as an instrument of convertibility, governments which resisted this on foreign policy grounds would be branded as conservative, something made easier by the growing general perception of the United Kingdom’s constitutional arrangements as the supreme example of a legal democratic order based on property rights and allowing access to power to the world of business and the professions. The linkage of European currencies to the franc germinal in the 1850s had been a dynamic one; in the 1860s French attempts to fix it more securely were seen as conservative. In all the member states of the Latin Monetary Convention, business and financial circles opposed its signature, so that, setting aside the difficulties of maintaining a fixed exchange rate system, the bimetallic standard was probably doomed by the time the Convention was signed. From being the rallying cry of bourgeois Europe, silver standards and bimetallism declined to being the rallying cry of debtors, protectionists, and agricultural exporters, who were often the same people. When in Switzerland in 1860 the Committee of the National Council appointed to examine the status of the silver standard accepted after the civil war and the question of adhesion to the bimetallic standard officially reported, it accepted the existing position only as a temporary one. Bimetallism was, the Committee wrote, ‘a transitional measure, a pathway to the single gold standard’ (Paillard, 1909, p. 267). Belgium and Italy both declared themselves in favor of a gold standard during the negotiations for the Latin Monetary Convention. The difficulty was in how to get there, and French diplomacy owed its hollow triumph to the particular impossibility of Italy’s going on the gold standard except after a long period of financial reform. Italy was by far the biggest capital borrower in the Latin Monetary union, financing in that way the infrastructure of the new nation state and a huge deficit on commodity trade with France. The price of Italian state bonds fell so relentlessly on the Paris market from 1860 to 1866 that one year after the Latin Monetary Convention Italy was in fact obliged to declare the inconvertibility of the lira. From the moment of that declaration onwards, France and Italy found themselves in a position of mutual monetary blackmail. Depreciating Italian government bonds had provided an additional motive for the inflow of Italian silver into France. They could be bought there with Italian coin and repatriated to Italy. There was no chance that Italy could redeem its silver in France at the fixed rates of the Latin Monetary union. When as a result of defeat in the Franco-Prussian war France too was obliged to declare the inconvertibility of its currency, its holdings of Italian silver then became in turn an extremely effective threat against all Italian attempts to get on the gold standard, unless it were by mutual agreement between the two countries. No matter how strong the pressure from commercial interests, neither country was in fact in a position to make the transition. Nor could one act independently of the other, for France would always have the support of the other member states of the convention in demanding that Italy redeem the silver in circulation outside its frontiers, while Italy could always threaten to coin more silver, at depreciating

94 MYTHS AND REALITIES OF THE GOLD STANDARD

rates against gold, which the others would have to take. Italy did in fact issue a further silver coinage in 1878, but only after a commitment to the other member states to redeem all earlier silver which they held. On France’s side the bimetallic standard of the Latin Monetary union was only maintained because it provided a support until the convertibility of the franc in London could be fully restored and on Italy’s side by the great political difficulty of establishing a ‘sound’ currency by what would in effect be a prolonged deflation. In the lingering deadlock of the Latin Monetary union, events in Germany became even more decisive for the adoption of the gold standard. Throughout the German world commercial interests pressed with increasing vigor in the 1860s for a coinage linked through gold to sterling. But the problem was greatly complicated by the different state currencies. The only common currency, which did not circulate and which played a role rather less than that of the current ecu in the European Union, was in silver, the Vereinsmünze, with a fixed rate against both the north German thaler and the south German kreuzer. When under Austrian pressure in 1857 a common currency which actually circulated throughout the German area, although in very limited volume, was created, this too was silver. But the revealing aspect of those negotiations was the Austrian demand that gold coin should also be minted, not necessarily as legal tender but as a commercial money for use in relations with the United Kingdom. Little coinage was in fact struck, but the idea of a money instantly convertible at favorable rates into sterling had been deliberately advanced as a move in a political argument to make an appeal to the German middle class over the heads of their many rulers, including the Prussian monarch. From the early 1860s petitions from Chambers of Commerce, who were joined across the many frontiers in a common association, for a monetary reform which would introduce a common currency on the gold standard, were an increasingly common expression of the nationalist demand for unification. There was, however, a sharp division as to whether such a coinage should be equivalent to French coinage. French investment in western Germany and the links of manufacturing to France were as powerful as the attraction of the northern ports to London and the wider world. ‘Modernization’ also favored France because it embodied demands for the decimal system, rather than the archaic divisionary system of sterling which also existed in most of the German states. The decimal basis of the Napoleonic French franc had of course been an integral aspect of its initial European-wide appeal to a modernizing middle class. In debate in Switzerland after the civil war it was often cited as one reason for adopting it rather than the non-decimal south-German kreuzer as a standard. The Franco-Prussian war was to emancipate Germany from France, but it did not by any means at first commit the government of the new empire to the gold standard. That commitment came only from the sustained constitutional pressure of liberal interests in the parliament. Bismarck, whose family’s economic interest was landed estate in the eastern provinces, remained privately in favor of bimetallism throughout his Chancellorship. The Prussian decree stopping the free

ORIGINS OF THE GOLD STANDARD 95

purchase of silver by the Prussian mint on 3 July 1871 was a defensive measure against the large stock of depreciated silver coin in France. It would not have done to have the larger war indemnity paid from that stock. The indemnity was transferred in a mixture of bonds, exchanged for gold on the London market and elsewhere, and direct gold transfers. The first transfers, however, were in silver, at a price set below the official French rate. It was only when the deliberately lowered rate did not deter such transfers that the stop on silver minting was imposed at first in Prussia and then later copied elsewhere. At that stage the new all-German government would still have preferred a bimetallic standard for the Empire. The law of October 1871 which permitted the minting of gold coins was a concession to public pressure, but only a small one and intended as a barrier against a gold standard. It had to be modified in the Bundesrat under pressure from the Permanent Committee of Chambers of Commerce to stipulate that if gold coins were minted they should be legal tender. The debate on the law in the upper house, the Reichsrat, then showed conclusively that there was a clear parliamentary majority for the gold standard. It was only this hard political reality which led to the prohibitions on minting silver coinage in the other German states. The law was then further amended under strong pressure from both houses of parliament to authorize the withdrawal of silver coinage, although in the absence of any imperial currency this could have no immediate effect. The main force of the amendment was to drive the imperial government towards a speedier replacement of state currencies by a national coinage, and it did not imply that the new coinage should not be silver. Indeed it offered no opinion on whether, if a silver coinage was minted, it should have a fixed ratio to the gold coinage that was clearly preferred. The provisional coinage law of July 9, 1873 first proclaimed the objective of an imperial monetary reform to be an Imperial gold coinage. But there was still a delay before this was achieved. It cannot be said that the German Empire went on to the gold standard until July 1875, when the Bank of Prussia, which was to become as the Reichsbank the central bank of the new Empire, declared that it would henceforward only redeem notes in gold. The new imperial currency, the gold Reichsmark, was only issued on January 1, 1876 and from that date onwards was the currency unit in which all accounts had by law to be settled. Any new silver money would be purely divisionary (Helfferich, 1898). The long drawn out German political battle was decisive, for the weight of Germany’s foreign trade in the international system was little less than that of France, and on West European markets growing so rapidly that the German Empire would soon become the main trading partner of many smaller European states. The Swedish coinage provides a striking example of the impact of the Franco-Prussian war and the German Empires choice of the gold standard. The ‘Carolin’ issued in July 1868, was indicated on its face as equivalent to ten French francs; it was withdrawn in 1873 when the provisional coinage law

96 MYTHS AND REALITIES OF THE GOLD STANDARD

passed in Germany. The Scandinavian Monetary Convention in 1872 opted for a gold standard, although Norway delayed its acceptance until 1875 (Janssen, 1911, p. 110 ff). There was no realistic alternative once the German decision meant that both Scandinavia’s two major trading partners were on gold. As in Britain, paper money was used and gold had little circulation. The Netherlands stopped the free coinage of silver in May 1873. The ten guilder gold coin was issued in June 1875. Nevertheless, the Dutch gold coin was not officially tied to any metallic standard, its exchange rate was left to be settled solely by purchasing power parity. Unlike the Bank of England, the Nederlandsche Bank had no legal obligation to provide gold, although it did in fact do so. Silver coinage remained also in practice convertible into gold. Commercial pragmatism and practicality had dictated the decision, and as elsewhere the choice of gold was a rejection of the Latin Monetary unions commitment to fixed rates. Austria only suspended the minting of silver coin in 1879. Until then the domestic money supply consisted of depreciating paper currency and silver coins, which from 1876 depreciated more rapidly. The Ausgleich of 1867 had created a common currency for the union between Austria and a Hungary which depended on agricultural exports. As Austrian modernizers envisaged a gold standard which would guarantee monetary stability for a long enough period of time for capital lenders to recover loans and interest without depreciation, so did Hungarian agricultural exporters prefer to stay with depreciating convertible silver. They were supported by many of the nationalist movements elsewhere in the Empire. It was only the violent oscillations, not the fall, in the price of silver between 1889 and 1892, including a steep upward movement in 1890, which finally caused the Habsburg Empire to change to the gold standard (Reidl, 1904). The best-known case of highly organized pressure from the agricultural sector for the preservation of a silver standard is in the United States. But even the plea of a presidential candidate that Western farmers whose interest was in exports to Europe should not be crucified on a cross of gold did not prevent America’s movement to a gold standard in 1879, after the suspension of convertibility during the civil war. It did, however, prolong a vigorous international agitation for a workable bimetallic standard and so played its part also in prolonging the Latin Monetary Convention until 1885. The Paris international conference in 1878 on the future of bimetallism was called at American instigation and with French compliance, not with any serious intention on the part of either government of halting the move to gold, but as a dramatic international gesture to appease a dramatic domestic opposition to the move in both countries. At the International Monetary Conference of 1881, the member states of the Latin Monetary union argued as one against silver, chiefly against Austria-Hungary and the United States, while the Société des Agriculteurs de France publicly opposed their governments stance (Baas, 1984). Even while volubly advocating the coinage of silver in other countries, and officially committing itself to silver purchases, the American government was in

ORIGINS OF THE GOLD STANDARD 97

fact avidly acquiring gold for its reserves. Under the Third Republic the essential characteristic of French policy was, likewise, its pragmatic ‘precious metallism’, the maximization of the amount of both precious metals in the country. Reserves were increasingly kept in gold—1,323 million francs out of 1,820 million francs in spring 1876, and the accumulation of gold in France and the USA provoked the final collapse of the silver price. At the same time, partly because of the earlier unredeemed inflow from the Latin Monetary union, as much silver as would be accepted by the public without driving out gold was allowed to circulate in France unsterilized (Baas, 1984, p. 267 ff). It was probably the quarrels in North Africa between Italy and France which eventually put an end to the Latin Monetary unions half-life in 1885. In 1881 a bill for the abolition of convertibility in Italy for foreign silver coins passed the Italian parliament, tantamount to a statement that the Union would not be renewed by any further agreement when the current one expired in 1885. At the same time customs duties began to be collected in gold. From September 1883 the state was legally bound to hold two-thirds of its reserves in gold, and banks were forbidden to accept silver for notes. The union was no longer of any value to French foreign policy and once Italy, as it became increasingly dependent on German rather than French capital inflows, also challenged France in North Africa there was no point in French governments losing domestic support in the pursuit of objectives which were unattainable. To the most ambitious developing powers, going on the gold standard then became a policy imperative. Japan’s banking commission pronounced in favor of the standard in 1893. Victory over China and payment of the Chinese war indemnity increased the gold stock over the years 1895–8 and brought the goal within reach. The minting of silver and its convertibility were terminated in 1897 (Matsuoka, 1937). Russia had been acquiring gold and limiting the amount of silver currency from the time that Germany and Scandinavia moved to gold. The transition took until 1896 when it was completed by Witte, than whom no European statesman was more representative of the modernizing middle class in power. To complete the panoply of domestic differences between states on the gold standard, and to re-emphasize the central importance of foreign commerce and convertibility on the London market as the motivation, Russia did not issue a gold currency. The Paris international conference in 1878 had presented a clear indication to the world that the divide between gold and silver was becoming a divide between the rich and the poor nations. The United Kingdom cast only a supercilious observer’s eye over the proceedings, having made it explicit before attendance that it was not intending to fetter sterling to any foreign or international rates, especially a fixed silver-gold ratio. Convertibility, for foreign trade purposes, in the London market; the financial stability which would facilitate foreign capital inflows and which was believed to be a consequence of that convertibility; the attraction of the United Kingdom as an economic and a political model of development to the rest of the industrializing world; all these elements now

98 MYTHS AND REALITIES OF THE GOLD STANDARD

appeared to justify Britain’s stance. The gold standard had become a fait accompli. Dogmatism would be silly when much of the story is still to be discovered, but the origins of the gold standard, it can be said, lie in some combination of the second and third categories of explanation which were set out at the beginning. What weight is respectively attributed to them as general causes is probably a matter of individual preference. But the argument here is that both remain insufficient and to them should be added something which links them closely, the desire for modernization and development. The aspirations that awoke overcame more easily reckoned short-term political calculuses of interest and drove political societies towards a choice of metallic standard which may have been both sub-optimal and irrational. The mobilization of these aspirations into effective political action would have been impossible without the existence of a model of development which it seemed feasible to emulate. In Europe this model changed in the 1860s from France to the United Kingdom, and from bimetallism to the gold standard. The timing of origins of the gold standard was crucially determined by France’s defeat by Prussia in 1870, but it had never been in France’s power to impose a bimetallic standard as the basis of a worldwide trade and payments mechanism, and the fact that France became a pillar of the gold standard in the 1890s was the entirely predictable outcome of the path of economic and social change there in the 1850s. In the long run of history, however, the issues were not finally decided in the 1870s. The adoption of the international gold standard coincided with the onset of the steady relative loss of British economic power. The ecu has now replaced the franc germinal, the Treaty of European unions clauses on a European Monetary Union have replaced the Latin Monetary union, although perhaps with no greater chances of success. The London market still struggles, in more difficult circumstances, to stand for free worldwide trade and uncontrolled rates, but the constitutional model which now increasingly attracts the world of business and industry has reverted to a unified European market and currency, managed from Paris, Bonn and Brussels. This is current ‘modernization’ and ironically for London, the power of its supporters may prove greater than that of their precursors of the 1850s. REFERENCES Baas, N.W.J. (1984) Die Doppelwahrungspolitik Frankreichs 1850–1885, doctoral thesis, European University Institute. de Cecco, M. (1974) Money and Empire. The International Gold Standard, 1890–1914, Oxford. Ford, A.G. (1989) ‘International Financial Policy and the Gold Standard, 1870–1914’, in P.Mathias and S.Pollard (eds), The Cambridge Economic History of Europe, vol. viii, The Industrial Economies: The Development of Economic and Social Policies Cambridge.

ORIGINS OF THE GOLD STANDARD 99

Frandreau, M. (1993) Cette faim sacrée d’or: la France, le bimétallisme, et la stabilité du système monétaire international, 1848–1873, doctoral thesis, Paris. Gibbs H.H. and H.R.Grenfell (1986) The Bimetallic Controversy; A Collection of Pamphlets, Papers, Speeches and Letters, London. Goodhart, C.A.E. (1972) The Business of Banking 1891–1914, London. Helfferich, K. (1898) Geschichte der deutschen Geldreform, vol.1 of Die Reform des deutschen Geldwesens nach Gründung des Reiches, Leipzig. Janssen, A.E. (1911) Les Conventions monétaires, Paris. Matsuoka, K. (1937) La genèse d’un étalon monétaire fondé sur le change-or: le cas du Japon, Annales d’histoire économique et sociale, vol. ix. Mertens, J.E. (1944) La Naissance et le développement de l’étalon-or 1696–1922, Paris. Paillard, G. (1909) La Suisse et l’Union Monétaire Latine. Etude économique et juridique, Paris, Lausanne. Riedl, R. (1904) Die Wahrungsreform in Osterreich-Ungarn Jahrbuch für Gesetzgebung, vol. XXVIII.

100

4 SHORT-TERM CAPITAL MOVEMENTS UNDER THE GOLD STANDARD Marcello de Cecco

Jacob Viner stated in 1937 that international capital movements had not yet found their historian. Things have progressed since then, but not substantially. In particular, there do not seem to have appeared, since the publication of Arthur Bloomfield’s Princeton paper, ‘Short-term capital movements under the pre-1914 gold standard’ many contributions which try to shed more light on the subject, on the basis of research conducted in the archives of banks and central banks. By default, therefore, we seem to remain inclined to accept the version of the story of how the pre-1914 gold standard worked which assigns short-term capital movements a largely equilibrating role, and sees the central banks of the core countries intent in following the Bank of England, the conductor of an ideal international monetary orchestra, or the leader of an international oligopoly over monetary affairs. The Old Lady would signal policy changes by changing her discount rate, money would be called to London as a differential appeared between interest rates there and elsewhere, and other central banks would raise their own rates. This is a way of giving back neoclassical virtue to institutions, like pre-1914 central banks, which on the contrary gave contemporary observers the distinct impression of being created exactly for the purpose of putting obstacles to international short-term capital movements which national economic authorities considered as endangering the economic policies which they were trying to enforce in their respective countries. Contemporary observers, in other words, thought that, far from forming an international orchestra expertly led by the Bank of England for the benefit of the world economy as a whole and of its component parts, central banks were increasingly used, in the course of the twenty years leading to World War II, as institutions whose role was to increase the national authorities’ control over their national economic systems, on the assumption that it was not always for the benefit of each individual economy that the international cycle moved and international capital movements, especially short-term ones, occurred. It was Albert Hirschman who defined central banks as honorary members of the liberal economic system, because money would not manage itself, either nationally or internationally, and it took centralized control of monetary policy to allow free trade and laissez faire to be maintained in the long run.

102 SHORT-TERM CAPITAL MOVEMENTS

This would involve very activist policies to prevent short-term capital movements from playing havoc with national economic policies aiming at fast growth and stability. And central banks, either by themselves or by mustering the help of their own commercial banks, were the carriers of such policies. This view of pre-1914 central banking is warranted by the observation of the actions of several central banks. But there are as many observations pointing to a completely opposite diagnosis of central bank policy. Central banks were institutions invented to increase, rather than decrease, financial repression. They accumulated gold and foreign exchange reserves and thus distorted the gold and foreign exchange markets from working according to the rule of unfettered profit maximization, the only rule consistent with global economic efficiency. Central banks did, on many occasions, and increasingly so as the fateful year 1914 drew nearer, operate to prevent gold from leaving their reserves. They also kept large foreign exchange balances at foreign banking centers not only in order to use them as intervention munitions, but as strategic balances to be used in the furtherance of foreign policy goals. This view of pre-1914 central banking has come to be discarded in favor of the more benign view, as the belle epoque has become more distant in time and people have elected to recall it as a golden age, as the good old days when there were clear rules and they were followed by well-intentioned gentlemen. Arthur Bloomfield and Jacob Viner never much believed this interpretation. In their eyes, the twenty years before 1914 had seen the first triumph of the modern state and of its appurtenances, of which central banks were among the most important. They did not believe short-term capital movements to have been always benign, and considered central bank action to restrain them as completely justified in view of macroeconomic stability and of the long-term orderly functioning of financial and real markets. Ralph Hawtrey, who wrote in 1914 an extremely lucid ‘situation paper’ for the British treasury on the matter of the adequacy of British gold reserves, took the view that the raids on British gold were led by foreign central banks which when they laid their hands on the metal were loath to part with it again. Central banks thus disrupted, with their actions, the carefully woven fabric of international finance which saw the Bank of England and the City of London as the centers of a nervous system they deftly controlled. The British secret had always been to be able to run a gold standard without having gold reserves, by just being able to muster them when necessary. It was essentially British capital that floated in and out of Britain at the call of the Bank of England. Thus the British could have their cake and eat it, too. It was Hartley Withers who gave the clearest, and plainest, description of how this magic worked. British capital, however, was private and not particularly animated by patriotism. It moved following the profit motive. The Bank of England also was a private institution, which worked to maximize dividends for its shareholders. It was thus a delicate balance that kept the international financial system based on London’s supremacy working smoothly. This balance was endangered and in the

MYTHS AND REALITIES OF THE GOLD STANDARD 103

end destroyed by the sheer imitation of British financial institutions carried out by other countries. As they started institutions, like banks and central banks, which being in their infancy were naturally weak, they protected them by etatiste behavior, exactly as Britain had done in the seventeenth century to get rid of Dutch financial supremacy. This is the behavior which Hawtrey, Viner, and Bloomfield clearly detected as the origin of disequilibrating short-term capital movements. There was, however, another large source of instability in the international financial system before 1914, and it arose from the purely profit-maximizing behavior of private agents. We refer to the practice of international arbitrage, which, although being as old as trade, and in fact being the very origin of all trade, rose to much greater importance around the turn of the century, because of great technical innovations which increased information and communication enormously. The amount of financial resources involved in international arbitrage was greatly increased by the huge increase in trade itself that occurred after 1870 because of the coming of age of steam navigation and steel shipbuilding. Each act of commerce gave rise to a bill to finance it, and bill accepting and discounting became a huge financial activity. There was nothing new in this, but the size and speed of transactions increased enormously. No market could consider itself as naturally isolated either from the trade or the financial point of view. Local bankers, if pressed for funds, could send local bills to be discounted in London, or avail themselves of finance bills drawn on imaginary trade transactions, to receive finance from London. Central banks were thus founded to attempt to put some discipline in local markets, which tended to follow movements initiated very far from them. They had therefore a protectionist goal, to repress and segment, to keep control of national financial resources and use foreign financial resources without allowing them to rule the roost. Again, this was nothing new. Financial crisis was sparked in Florence in the fourteenth century (as Carlo Cipolla has explained admirably) by a scarcity of gold initiated in the Far East and reverberating as far as Italy through the Italian merchants. But by the last quarter of the nineteenth century, technical innovations caused phenomena which previously took months and years to work themselves out to occur in days and to mobilize enormous resources the world over in the name of profit maximization. This new flexibility of international financial prices hit against the fix-price systems prevailing in national markets, even financial markets. Short-term capital movements were the vehicle which carried the new flexibility and shook the national fix-price systems and the policies based on them. Another phenomenon which had always existed also took new wing in the same period, and was the cause of very large short-term capital movements. It was sovereign debt, which rose to reach a huge and ever-increasing size, beginning in the 1880s. There is no need here to go over the issue of whether sovereign loans were motivated by pure profit maximization on the part of the bankers which made them. What matters here is that many countries tried by

104 SHORT-TERM CAPITAL MOVEMENTS

borrowing in the international market to maintain convertibility or near convertibility which they had established to try to get rid of the wild oscillations their currencies had experienced in the 1870s, as a result of the many concomitant new factors that had intervened in that decade to shake the international economy (German unification being probably the most important among them). Sovereign loans, and in general international loans, gave rise to a recurrent flow of interest payments, which could be guessed by the markets because they occurred at regular and known intervals. Debtor countries had therefore to face the phenomenon of arbitrage on a massive scale, when the pressure on the exchange which was felt when they had to purchase foreign currencies to make interest payments was made harsher by arbitrageurs purchasing those currencies to make them even scarcer. The problem was made more acute by the widespread use of forward transactions. Again as old as the hills, forward transactions were made much easier by technical advances. National monetary authorities of debtor countries thus found themselves trying to maintain convertibility while international arbitrageurs made huge bets against their currencies, bets which they could make without much risk, as the direction of demand and supply in the foreign exchange market could be easily guessed because of the regular intervals at which interest payments on sovereign loans occurred. Sovereign loans were discrete occurrences, they did not take place in a continuum, and so were interest payments. Trying to outsmart the arbitrageurs thus became the principal activity in which the central banks and treasuries of debtor countries engaged. Things were not made easier for them because of the fact that most of the arbitrageurs were bankers and financiers from their own countries; they were the ones who knew best about the dates and amounts of their governments’ financing needs. The international financial market, however, was essential to the game the arbitrageurs played against their own monetary authorities, as it allowed them to find the resources to place their bets. The massive recourse to the international financial market for sovereign debt thus gave debtor countries a continuing and pressing need to measure themselves against short-term capital flows which did not aim at thwarting their efforts to maintain convertibility, but had this as their practical result. This need induced them to try and perfect techniques of foreign exchange intervention which were seldom known and even more seldom practiced in earlier days. Here we have a real break with the past. The financial arena where speculators and authorities now play their games encompasses the whole world, and central banks and national treasuries see their task as consisting of preventing the market from compelling them to abandon currency convertibility. Things are made even more interesting by the fact, which is more frequent than historians have cared to underline, that the national currency is not tied to one particular exchange rate by law. In the case of both Austria and Italy, which, as is known, have in the pre-1914 years a parallel monetary history, there was no formal convertibility of either the crown or the lira, fixed by law. Thus the authorities in both countries, which were large international debtors, had great

MYTHS AND REALITIES OF THE GOLD STANDARD 105

discretionary powers for devising and executing exchange rate strategies. The gold premium at which the crown and the lira were exchanged could and did change, and in the case of Italy in the twenty years before 1914 there were very wide oscillations below and above par. The most important decision the monetary authorities of debtor countries had to take was whether to intervene in the bond market or in the exchange market, if they wanted to defend a particular exchange rate level. They had also to decide whether they wanted to intervene in the national bond market or in the bond market of the financial center where their sovereign (and other national) bonds were mostly quoted. This, for sovereign bonds, was by and large Paris, which specialized in the financing of European governments. To choose the intervention technique, the authorities had to decide whether the speculative attack had started in the bond market or in the foreign exchange market. It could be either, as a perfectly balanced bond market could be disturbed by the Lira gold premium turning to a discount because, say, of trade reasons, and opening up arbitrage opportunities for bond speculators. The most frequent occurrence, however, was that the bond market induced arbitrage operations which, to be carried out, required sales of the national currency, which endangered the exchange rate. Once the authorities had realized that it was most frequently the bond market that caused exchange rate oscillations, they thought of how best to intervene in it while at the same time saving their scarce gold and foreign currency reserves. First of all, intervention in foreign bond markets was a reason for the authorities to keep regular balances with banks in the same centers, or foreign treasury bills which they could sell to get the foreign exchange to intervene in the bond market. Central banks had to request authorization from their governments to keep those balances instead of gold and count them as reserves for their note issue. Sometimes this was granted as a new rule, sometimes treasuries preferred to keep the banks on a short leash, and give them the foreign exchange necessary for the intervention as it was needed, thus participating in the day-to-day management of the exchange rate. Intervention was, however, costly, even if the need to keep reserves under the form of bank accounts in foreign centers or foreign treasury bills had the nice feature that it allowed central banks and treasuries to earn interest on their reserves, which was better than holding gold, which was barren. The monetary authorities of debtor countries thus started to resort to non-quantitative methods of influencing the foreign bond markets where their paper was quoted. Friendly actions by banks which had their base in the financial centers in question were sought and often obtained by direct recourse to their political authorities. If the debtor countries were home to commercial banks large enough to operate in the foreign exchange market, the monetary authorities liberally availed themselves of their power over them to enlist their services. This was, however, done at a later stage, when the foreign exchange market, rather than the bond market, had become the premium mobile. Monetary authorities also discovered a cheaper way to stem oscillations in the foreign bond

106 SHORT-TERM CAPITAL MOVEMENTS

markets which could spark off runs on their currencies. This had to do with something we are today extremely well aware of, the power of the media. The media at the time were exclusively newspapers, which had in those very years become a mass means of information for the European bourgeoisies, who were also the ultimate lenders in the sovereign bond market. Monetary authorities discovered that if they could manage to have appropriate articles printed in the financial and political sections of the most important newspapers, they could prevent runs on their bonds from occurring, or stem them when they had started. This was the time when the difference between inside and outside information began to be discovered, a discovery which was to be much used in the years between the wars by economic theorists such as Keynes. From this, to trying to purchase journalists, was a very short step which was climbed by most national authorities, generously helped by the journalists themselves, who were quick to understand how the public good of information could be used for private gain. The twenty years before 1914 were an increasingly complex time, when a plurality of very large players had occupied the world stage, surrounded by even more numerous secondary players, who also wanted to grow in importance and have their say. All this gave rise to a growing difficulty to correctly interpret facts and trends, coupled with the possibility of making a lot of money very fast if one could. It also was, as we have already noted, the time that saw the definitive triumph of the modern centralized state, which started extending its influence in all directions, invading fields where private players had grown used to being completely free of state interference. The study of the archives of debtor countries has yielded a wealth of information on these non-canonical ways of influencing international short-term capital movements. Soviet authorities, soon after the October Revolution, published papers proving that the famous financial journalist A.Raffalovich had been for many years in the pay of the Russian government, with the task of propping up Russian imperial bonds to the French public, and denigrating the national debt of other countries, like Italy, which could compete with them for the favor of the same public. In the archives of the Italian central bank we found evidence that Giacomo Grillo, director general of the most important Italian bank of issue until 1894, the Banca Nazionale, replied to the Russian offensive in kind, by enlisting the services of Fiorillo Fournier, a well-known financial journalist who wrote up Italian bonds in several French newspapers. Previously, Grillo had also treated with great liberality, buying up many copies of his book at a very considerable price, the German economist Otmar Haupt, who published an annual survey of world monetary affairs. Careful research in the archives of other central banks and treasuries would no doubt yield similar finds. By working for several years on the archival material of the Bank of Italy I have had the privilege of being able to follow the formation of the central bank of a semi-peripheral European country, and its many struggles to keep ahead of an international financial market which, in the twenty years before 1914, subjected the Italian currency to very wide oscillations. After suspending

MYTHS AND REALITIES OF THE GOLD STANDARD 107

convertibility in 1893 Italy never went formally back to it, preferring, like Austria, to shadow the gold standard without being officially tied to it. As in the case of Austria, whose foreign exchange operations were analyzed by Paul Einzig several decades ago, the Italian monetary authorities managed an informal gold-exchange standard, and had to face continuous and serious problems arising from the existence of a very large stock of Italian government debt in the hands of foreigners who kept it not as rentiers but as professional portfolio managers do today, in order to get quick capital gains. The prices of Italian bonds oscillated wildly on the foreign markets where they were quoted, and offered Italian financiers, who were many and extremely well funded, the opportunity of very lucrative arbitrage operation. In the absence of any exchange controls, the appearance of a price differential on Italian bonds between Paris and the Italian market induced these Italian professional arbitrageurs to buy them in Paris to sell them again in Italy, thus depressing the lira exchange rate, as they bought the foreign exchange necessary to operate and sold Italian currency. This is a story we are today extremely familiar with, as it has been plaguing many monetary authorities in recent years, since the U.S. Federal Reserve started its policy of radical interest rate reduction to save the endangered U.S. banks from insolvency and revive the American economy, while at the same time the authorities of reunited Germany have been struggling with the difficult task of keeping tax rates low and stemming the inflationary consequences of large public expenditure destined to rebuild structures and infrastructure in the eastern Länder. Like the twenty-year period before 1914, recent years also have witnessed a huge rise of sovereign debt, first of developing, then of developed countries, and the widespread purchase of this debt by the managers of massdiffused investment and pension funds. Exchange rates have thus become directly influenced by the opinions markets form of the sovereign bonds of individual countries. At the same time, the demise of the Soviet Union and the end of cold-war foreign policy has rejuvenated opposed foreign policies on the part of former allied countries. The end of the cold war has thus engendered the re-politicization of foreign economic relations among former Western allies. The consequences for the bond and foreign exchange markets are beginning to be felt. A close look at the world of the two decades before 1914 is therefore useful both from the practical and historical viewpoints. As in those two decades, the consequences of fast technological advance are being felt in the international financial market, as well as the high importance of information in a world where financial transactions are conducted mainly by specialists who have no fixed principles on which to base their operations. What Keynes so aptly described as the ‘beauty contest dynamics’ prevailing in the financial markets first started to operate in the two decades before 1914. It was then that arbitrage between the financial instruments of many countries became a mass phenomenon, giving rise to extremely volatile short-term capital movements. Monetary authorities, academic economists and politicians were quick to realize that short-term capital

108 SHORT-TERM CAPITAL MOVEMENTS

movements had grown to a size and speed that made them inherently destabilizing, if only because the flexibility and speed of international financial markets came into friction with the fix-price system that ruled most domestic markets. It was only the Cunliffe Committee that gave credence, in 1918, to the speed and flexibility of adjustment of crucial real markets, like the labor market in pre-war years. Before the war, the greatest economists, like Marshall and Pareto, had frequently and in detail remarked on the stickiness of money wages, both downward and upward. The conflict that arose between the behavior of financial markets and that of most domestic markets in the two decades leading to World War I became much more noticeable, and much more widely noticed, after 1900, when political crises between the great powers became more and more frequent, the shadow of a great European war became longer and longer, and international financial crises broke out with increasing frequency. The crisis of 1907 in particular opened the eyes of many to the new world which massive short-term capital mobility coexisting with slow-moving national political and economic variables had created. The 1907 crisis convinced many experienced observers that new policies and institutions were needed to face the problem we have just mentioned. This explains the blossoming of proposals to organize compulsory recycling of destabilizing short-term funds, like the one put forward by Luigi Luzzatti. Reading it, one is struck by the fact that is might have been produced by J.M.Keynes or H.D.White at Bretton Woods. The landscape it paints is the same, and the remedy it proposes is very similar. Compulsory recycling of short-term balances was exactly what Keynes and White proposed at Bretton Woods. The enmity of the main beneficiaries of capital flight before, during and after World War II, American large banks, prevented the Keynes and White proposals from being enforced. But foreign exchange controls were, in 1944, the order of the day in every country, and short-term capital flows could be stemmed by that imperfect method. It is interesting to notice how fast the thinking of experienced observers evolved towards radical cooperative solutions in the years before 1914, when the problem of a too volatile international financial market became serious. One remarkable proposal, made by the French journalist Edmond Thery in 1893, would have involved the introduction of foreign exchange controls and a state monopoly of foreign exchange in Italy, to solve the huge crisis of confidence in the Italian currency that exploded in the early 1890s, a result of the prolonged and eventually unsuccessful attempt by the Italian authorities to maintain convertibility while inflating the economy, borrowing their way to external balance. Italian academic observers were at the time inclined to lay the responsibility for the crisis at the Italian authorities’ door, accusing them of excessive spending and of cavalier policies towards their banks. (A memorable article was written by the young but already highly respected Rodolfo Benini describing in great detail the borrowing policies of the Italian authorities since

MYTHS AND REALITIES OF THE GOLD STANDARD 109

the return to convertibility and deprecating their attempt to stick to their foreign currency option in the face of a profligate budgetary policy. He advocated the end of that policy and the pursuit of basic external balance by a devaluation, which would bolster exports and restrain imports.) This was certainly true, but we must not forget that it was the international financial market which conceded more and more money to spendthrift governments like the Italian one. We must also consider the fact that the Italian financial crisis exploded together with those of several other peripheral countries, in particular Argentina. It was thus the same international financial market which had so freely lent to doubtful sovereign borrowers such as Italy and which, suddenly coming to its senses, pulled the plug on them and precipitated the crisis. A foreign observer like Edmond Thery could thus be inclined to recognize in destabilizing speculation the chief culprit of the Italian crisis. His diagnosis coincided with that put forward by Giacomo Grillo, Director General of the Banca Nazionale, the embryonic Italian central bank which was to be destroyed by the crisis and born again in 1894 as the Bank of Italy, under a new management and especially under a new and stricter central banking law. Grillo had an axe to grind, but the same could not be said of Thery. He was an independent observer and yet came to the same conclusions as Grillo about the role of destabilizing international speculation in the Italian crisis. It is now time to bring this short survey to a close, drawing a few conclusions It seems clear that, in the two decades preceding World War II, extremely rapid and deep change was occurring in the international financial market. The expansion of trade gave rise to a huge amount of bills of exchange, and uncertainty over the exchange rates of semi-peripheral but still important countries like Russia, Austria-Hungary, and Italy made it mandatory for economic agents involved in foreign exchange transactions to entertain views about the future of exchange rates, and to adopt defensive policies to safeguard their trading profits. At the same time, pure short-term speculation on exchange rates and sovereign bonds grew to very large size. Monetary authorities, and central banks in particular, found themselves devising all manner of intervention systems to prevent these short-term capital flows from dictating exchange rates of their national currencies which did not reflect either fundamental trade flows or the long-term capital position of their countries. The view that it was short-term capital flows that drove the balance of payments of most semi-peripheral countries gained ground, especially among central bankers, treasury bureaucrats, and politicians, until it became the leading theory of exchange rate determination in those circles, with only academic economists left to hold faith in the HumeRicardo theory of trade balance. Hundreds of years of familiarity with a double or even triple standard had made financiers thoroughly conversant with arbitrage and with one-side speculation. When bimetallism gave way, in the developed world, to monometallism, financiers started to use their wits to arbitrage between paper money and gold currencies and to practice bond and foreign currency arbitrage.

110 SHORT-TERM CAPITAL MOVEMENTS

The archives of the Bank of Italy are full of evidence of the continuous battle the monetary authorities fought to prevent exchange rates from being dictated by a volatile foreign exchange market not much given to taking its lead from fundamentals. While this great battle was being fought between monetary authorities and international speculators, most of whom were nationals of the countries whose currencies they speculated on, political authorities were intent in using foreign balances as an instrument of machtpolitik. Central banks thus found themselves buffeted by short-term capital flows which they tried to resist, while at the same time having to adjust to the dictates of more and more threatening politicians, who could not see why foreign balances should not be used in the exciting game of power politics. With Oscar Wilde, we may conclude that pre-1914 monetary history is just a little too exciting. The First World War might not, after all, have broken out by mistake. REFERENCES Angell, Norman (1910), The Great Illusion, London, Heinemann. Bloomfield, A. (1963), Short-term Capital Movements Under the Pre-1914 Gold Standard, Princeton Studies in International Finance, Princeton. de Cecco, Marcello (1983), The International Gold Standard, Money and Empire, (second edition) London, Frances Pinter. de Cecco, Marcello (1991), L’Italia ed il sistema finanziàrio internazionale, 1861–1914, Bari-Roma, Laterza. Del Vecchio, Gustavo (1914), La politica delle Divise, in de Cecco (1991), doc. 212. Eichengreen, Barry (1992), Golden Fetters, Oxford, O.U.P. Eichengreen, Barry and Richard Portes (1988), Settling Defaults in the Era of Bond Financing, Discussion paper no. 272, London, C.E.P.R. Hirschman, Albert O. (1958), The Strategy of Economic Development, New Haven, Yale University Press. Luzzatti, Luigi (1883), Dell’attinenza dei biglietti di banca col bimetallismo, in Nuova Antologia. Luzzatti, Luigi (1908), Une conference internationale pour la paix monétaire, Note pour l’Institut de France, Imprimerie Chaix, Paris, reproduced in de Cecco (1991), as doc. 20. Raffalovich, A. (1931), L’abominable venalité de la presse. D’après les documents des archives russes (1897–1917), Librairie du travail, Paris. Salazar, António d’Oliveira (1916), O ágio do ouro. Sua natureza e suas causas, Coimbra, Imprensa da Universidade. Viner, Jacob (1932), Gold and Monetary Stabilisation, in International Economics, London, Allen and Unwin, 1951. Viner, Jacob (1937), Studies in the Theory of International Trade, New York, Harper. Withers, Hartley (1909), The Meaning of Money, London, Smith and Elder.

5 THE GEOGRAPHY OF THE GOLD STANDARD* Barry Eichengreen and Marc Flandreau

INTRODUCTION The gold standard is a subject that has garnered much attention, but its geography remains strangely uncharted. Monetary systems in which gold coin and assets convertible into gold provided the basis for the domestic circulation are commonly portrayed as the normal state of affairs prior to 1913. England, in this view, was on the gold standard for two centuries, ever since Sir Isaac Newton, in his role as Master of the Mint, set a high silver price for the gold guinea in 1717, driving full-bodied silver coins out of circulation. Portugal, which relied heavily on the British market for exports and on British industry for imported manufactures, adopted gold in 1854. Germany’s accession to the gold standard dated from her Bismarckian unification. In conventional accounts, virtually the entire world was on some form of gold standard during the final part of the nineteenth century. These ‘conventional accounts’ serve us, admittedly, in the capacity of straw men. Scholars like Bloomfield (1959) have been careful to note that not all gold standards were alike, distinguishing gold coin standards from gold bullion standards, coin and bullion standards from gold-exchange standards, and full gold standards from limping gold standards. But by focusing on the experience of Western Europe and North America, most of their studies run the risk of distorting the geography of the monetary system and hence of misrepresenting its evolution.1 A better rounded portrait reveals a situation in which gold was the basis for the domestic circulation in only a limited number of countries. In many places, especially outside Europe, gold convertibility was first established only in the final years preceding World War I. Fathoming the operation of the international monetary system thus requires comprehending how it worked prior to the completion of these transitions. Tracing its evolution requires understanding the timing and nature of these monetary regime transformations. Some studies have acknowledged the limited domain of the nineteenth century gold standard, emphasizing the distinction between gold and silver convertibility. Gold, it is said, was the basis for the circulation in

112 MYTHS AND REALITIES OF THE GOLD STANDARD

advanced countries at the core of the international system, while lower-income countries at the periphery utilized the less valuable metal, silver. We reject this view. In the 1860s, the start of the period we consider, whether countries operated gold, silver or bimetallic standards depended not so much on whether they were rich or poor, as on the foreign country or region with which their transactions were linked and on monetary standard of that trading partner or partners.2 The countries of the ‘British area’ (Britain, Portugal, Canada, Australia and Egypt) were essentially alone on the gold standard,the monetary regime on which, to exaggerate the point, Newton had incidentally placed them. The ‘French area’ (France, Belgium, Italy and Switzerland) was bimetallic. The ‘German area’ (the German states, Scandinavia, the Netherlands and Austria) was on silver, along with much of Asia. But for most of the nineteenth century the important distinction was not between gold and silver but between metallic convertibility and inconvertibility. Countries whose currencies were convertible, whether into gold, silver or both metals, enjoyed relatively stable exchange rates, while countries with inconvertible currencies might see their exchange rates fluctuate widely. There was only a weak correlation between the level of economic development and the choice of convertible monetary standard, but a strong correlation between the level of development and convertibility or inconvertibility.3 In our view, it is the distinction between convertibility and inconvertibility at least as much as that between gold and silver on which analysis should focus. Our goal in this paper is to suggest how such an analysis might be undertaken.4 First we present snap-shots of international monetary arrangements in 1868 and 1908. The information is not new, but assembling it in one place allows us to flesh out the portrait of international monetary arrangements. There is then a reconstruction of the movements that connected the snap-shots. It maps the routes that led countries to gold convertibility and offers hypotheses to explain their transitions. Finally we draw out the implications for research on the history of the international monetary system. MONETARY ARRANGEMENTS IN 1868 Table 5.1 summarizes the state of the international monetary system two-thirds of the way through the nineteenth century.5 The term ‘standard’ as used in Table 5.1 refers to the asset that was legal tender in unlimited amounts and could be freely coined. Countries that we classify as being on the gold standard may have had some silver in circulation but limited the value of the transactions for which it could be used. Countries classified as on silver may have had some gold in circulation but the value of gold coins in terms of the national unit of account could change with market conditions. Countries labelled bimetallic conferred legal tender status on both metals.6 The term ‘convertibility’ refers to legal provisions affecting paper money. A convertible-currency country was one in which issuing banks were obligated to redeem their notes for specie. This distinction had important implications for

GEOGRAPHY OF THE GOLD STANDARD 113

Table 5.1 Monetary systems of the world, 1868

Note Based on Seyd (1868).

Figure 5.1 The geography of the gold standard, 1868

114 MYTHS AND REALITIES OF THE GOLD STANDARD

GEOGRAPHY OF THE GOLD STANDARD 115

the exchange rate. If the currency was inconvertible, additional issues could produce a premium on specie, as in e.g. Austria, Russia, Italy, and the Ottoman Empire. A conspicuous feature of Table 5.1 (which is depicted graphically in Figure 5.1) is that as of 1868 only two European countries, Britain and Portugal, were on the gold standard in the sense that their currencies were unconditionally and exclusively convertible into gold.7 Gold convertibility was also maintained in the German state of Bremen and in Canada, Chile, Egypt and Australia. Other countries officially adhered to monetary standards based partially or wholly on silver.8 Also evident in Table 5.1 is the fact that the tripartite distinction between gold, silver and bimetallism incompletely captures the reality of nineteenth century monetary arrangements. Many countries whose currencies were officially convertible allowed their obligation to convert domestic currency into specie to lapse and permitted inconvertible paper to circulate alongside silver and gold. The distribution of countries among these categories was related to the timing and pace of their economic development, as suggested by Figure 5.2. Early participants in the industrial revolution (Britain, France, Belgium, Switzerland, and the German states, for example) were more likely to possess a convertible currency, while those on the periphery of the industrializing world (Spain and Italy, for example) were more likely to utilize inconvertible paper.9 The pattern suggests that bullion, not yet ‘a barbarous relic’, was rather ‘an expensive luxury’ that mainly high-income countries could afford.10 But within Europe at least, as also shown in Figure 5.2, there was no analogous correlation between the level of income per capita and the metal into which the currency was convertible. The presence or absence of a convertible currency was correlated not just with per capita incomes but with the state of the public finances.11 Many countries that operated de facto paper standards experienced persistent fiscal problems and utilized seigniorage as their revenue source of last resort. This tendency was most obvious in periods of military conflict and insurrection. In Austria-Hungary, the National Bank was forced to suspend convertibility in 1848 following a revolution in Hungary and the military expense of its suppression.12 Until 1858 the exchanges were at a discount relative to the bimetallic parity, when an attempt to restore convertibility at the statutory par collapsed after seven months due to the outbreak of war with Italy and the havoc it wreaked with the budget. A second attempt to reform the public finances and restore convertibility in 1865– 66 was again undermined by war, this time with Germany and Italy.13 Greece was forced to suspend bimetallic convertibility in 1868 in response to the Crete Revolution and the consequent increase in military spending. Bimetallic convertibility was restored in 1871 but suspended again in 1877 due to the outbreak of war with Turkey.14 In the case of Italy, the bimetallic Piedmontese system adopted following unification gave way to inconvertibility following the outbreak of war with Austria in 1866, since France opposed Italy’s campaign and

116 MYTHS AND REALITIES OF THE GOLD STANDARD

Figure 5.2 Income, convertibility and monetary standard in 1868

prevented borrowing in Paris.15 The convertibility of the U.S. dollar was suspended with the outbreak of America’s civil war; nearly two decades were required for its official restoration. Gold and silver were driven from circulation in Russia and the paper ruble depreciated in the wake of the Crimean War.16 Japan’s attempt to adopt gold convertibility in 1871 was frustrated by the

GEOGRAPHY OF THE GOLD STANDARD 117

Figure 5.3 Sterling exchange rates, 1860–1890

expense of suppressing internal rebellions.17 Spain’s efforts to stabilize its currency were disrupted by the outbreak of the Spanish-American War. Figure 5.3 shows that exchange rates between gold, silver and bimetallic currencies were remarkably stable vis-à-vis one another until the 1870s. In contrast, the exchange rates of countries with inconvertible currencies could fluctuate widely against countries maintaining convertibility. The distinction circa 1868 with implications for exchange-rate behavior, in other words, was between convertibility and inconvertibility, not between gold and silver or gold and bimetallism.18 The mechanism responsible for the pre-1873 stability of the gold-silver exchange rate was the passive buffer-stock role of the bimetallic bloc, which adjusted its gold and silver circulation to accommodate global bullion market conditions. Following the discovery of new supplies of one of these metals (say, gold), the abundant metal flowed toward the bimetallic bloc, supplanting the scarcer metal (say, silver), which was exported to the rest of the world, thereby stabilizing relative supplies of gold and silver. Through the operation of this buffer-stock mechanism, stable exchange rates between gold and silver standard countries were successfully maintained.19 In the 1870s, this mechanism began to break down. With the failure to negotiate arrangements based on international bimetallism and the departure of France and the rest of the Latin Monetary union from the bimetallic standard, the relative price of silver and gold began to vary. Exchange rates between currencies convertible into the two metals began to fluctuate, and the choice between gold and silver convertibility attained new importance. The consequences are the subject of the next section. MONETARY ARRANGEMENTS IN 1908 Table 5.2 (and Figure 5.4) tabulate monetary arrangements circa 1908. Virtually all of Europe had joined the United Kingdom and Portugal on the gold standard.

118 MYTHS AND REALITIES OF THE GOLD STANDARD

Table 5.2 Monetary systems of the world, 1908

GEOGRAPHY OF THE GOLD STANDARD 119

The only exceptions were Greece, which was formally on gold but where an extended period of inconvertibility ended in 1910, and Spain, which remained on paper, albeit with a relatively stable exchange rate (at least compared to its own previous record). Much of the non-European world had also gone over to gold; exceptions included China, still on silver, portions of Latin America on silver and bimetallic standards (Honduras and El Salvador remained on silver), and bimetallic Persia. Why were so many more countries on gold in 1908 than four decades earlier? As already noted, until the 1870s gold, silver and bimetallism all provided exchange rate stability, rendering the choice of standard a matter of relative indifference. This is not to deny that the maintenance of different monetary standards in different countries imposed transactions costs on individuals doing business across national borders. Merchants in silver standard countries importing from the gold standard world had to exchange their silver for gold in order to pay for their purchases, for example. Given the United Kingdom’s growing importance in world trade, the gold standard hence became an increasingly attractive option for many countries. The more countries shifted to gold, the more appealing the option became for the remainder. Once the bimetallic countries, whose buffer-stock function had stabilized exchange rates between silver- and gold-based currencies, defected to the gold standard, currencies convertible into silver experienced wide fluctuations. Exchange risk grew, discouraging international transactions. In 1886, for example, the Hong Kong Bank committed uncovered silver funds to finance tea exports from Hankow. When the Hong Kong/London exchange rate moved by 8. 8 per cent, it suffered very substantial losses. Such experiences led the Hong Kong Bank to hedge its exchange risk by holding gold deposits in London. Similarly, Mexican officials blamed silver-induced exchange rate variability for discouraging foreign investment. Indian commentators argued likewise.20 With the spread of industrialization came a shift in the balance of power between groups favoring monetary standards with inflationary and deflationary biases. With the advent of modern economic growth, the silver-mining interests and agricultural debtors who formed natural constituencies for the more inflationary silver standard lost clout to new industrial and financial elites. Such trends cannot explain the shift toward gold in parts of the world where economic growth remained slow, but they operated powerfully in Western Europe and in the United States. As de Cecco (1986) puts it, industrialization ‘went pari-passu with monetary stabilization, leading to the establishment of gold-based monetary systems.’21 Throughout the period, convertibility had important credibility effects. It signalled a government’s commitment to sound budgets and balanced external payments with sustainable volumes of foreign borrowing. Gold convertibility was particularly effective in this regard. The gold premium was a ready indicator of whether the debtor was keeping its part of the bargain. Foreign banks were particularly inclined to lend to governments that had staked their reputations on

Figure 5.4 The geography of the gold standard, 1908

120 MYTHS AND REALITIES OF THE GOLD STANDARD

GEOGRAPHY OF THE GOLD STANDARD 121

the maintenance of gold convertibility.22 A convertible currency thus promised to enhance a country’s access to international capital markets. Given that the principal lenders—the United Kingdom, France, Germany, the Netherlands, Belgium and Switzerland—were all early members of the gold standard club, gold convertibility came to be seen as the logical arrangement for aspiring borrowers. None of this is to deny that there was a strong element of serendipity in the spread of the gold standard. Industrialization, expanding trade, and capital flows all may have encouraged the ‘standardization of the standards,’ but impetus was also provided by the desire to restore the exchange rate stability that had prevailed before 1873. Stability could have been obtained on a number of bases. That the United Kingdom, the leading trading and lending nation, was already on gold encouraged countries to coordinate on one of those equilibria. To this extent, convergence on the gold standard (instead of the silver standard) was in a part a matter of historical happenstance. TRANSITIONAL DYNAMICS While all of the factors enumerated above encouraged countries to move from silver, bimetallism and inconvertible paper to gold, different countries followed different routes from the de facto silver and inconvertible-paper standards of the 1860s to the international gold standard of the 1900s. Is it possible to generalize about their paths? Moving from silver or bimetallism to gold required transforming the metallic basis of the circulation and augmenting the gold reserves of the authority responsible for the convertibility of paper money and token coins. While it was not necessary to replace the entire silver stock with gold, the establishment and maintenance of gold convertibility required that gold reserves exceed some minimum share of domestic monetary liabilities.23 Meeting this condition entailed exchanging silver for gold where holdings of the former were ample and accumulating additional reserves in gold where they were not. Countries seeking to complete this task followed one of three routes. Some moved from silver or bimetallism to gold at an early date and without suffering an interlude of exchange rate instability. This was true of Germany, which limited silver coinage in 1871 and implemented gold convertibility within two years. A similar path was followed by the Netherlands, which did not establish gold convertibility immediately but still managed to stabilize its exchange rate vis-à-vis the gold standard world (see Figure 5.5). Other examples include Belgium, France, Switzerland, and the Scandinavian states. A second group of countries (Austria, Russia, and some Latin American nations) also moved to the gold standard, but after a delay that involved an extended period of inconvertibility. Many of these countries’ monetary standards were formally bimetallic but effectively inconvertible. As the price of silver fell, their silver parities declined to the point where paper was no longer depreciated

122 MYTHS AND REALITIES OF THE GOLD STANDARD

Figure 5.5 Exchange rate between London and Amsterdam, 1865–90

in terms of silver. (Figures 5.7 and 5.8 illustrate these dynamics for Austria and Russia.) This forced them to choose between restoring the silver standard and changing their monetary constitutions. Austria suspended silver coinage and allowed its exchange rate to float in the neighborhood of a new devalued gold parity: it finally stabilized after 1899. Russia, when faced with this situation, stabilized more quickly, completing the task before the end of 1893. In some cases such as Spain (Figure 5.9), the depreciation of paper currency accelerated later in the century at a time when the price of silver was so low that paper currency, although wildly depreciated, remained steadier than silver. Still another group of countries followed a third route. Members of this group operated silver standards. They tended to be poorer economies with formal or informal imperial ties to the industrial core. They saw their silver-based currencies depreciate against those of the gold standard world and often experienced bouts of inconvertibility. At some point, usually after 1890, they limited the coinage of silver and moved to the gold standard. (Figure 5.6 illustrates the experience of India, one member of this group.) But the form of their gold standards differed; they adopted gold-exchange standards rather than traditional gold standards. We now consider factors that can help to account for countries choice of path.24 National income How countries completed the transition to gold depended in part on their levels of income and economic development. High-income countries initiated the transition early and completed it quickly, typically without experiencing

GEOGRAPHY OF THE GOLD STANDARD 123

Figure 5.6 Exchange rate between London and Calcutta, 1870–1914

Figure 5.7 Exchange rate between London and Vienna, 1870–1914

inconvertibility. Low-income countries for whom the acquisition of gold reserves was more difficult navigated a more circuitous route. In the first category were early industrializers (Germany, France and other members of the Latin Monetary union) that moved from de facto silver standards to gold convertibility in the 1870s. On the eve of the transition their reserves were composed of a melange of gold and silver. By trading silver for gold on world markets, they sought to amass the specie necessary for the operation of a gold standard. Their possession of relatively ample reserves allowed them to move directly to gold. In the second category were countries that industrialized later, that navigated the transition more slowly, and that followed an indirect route along which their silver-based monetary systems first gave way to inconvertible paper and only

124 MYTHS AND REALITIES OF THE GOLD STANDARD

Figure 5.8 Exchange rate between London and St. Petersburg, 1870–1914

Figure 5.9 Exchange rate between London and Madrid, 1870–1914

later were replaced by gold. Russia, Japan and Austria all moved from bimetallic standards based primarily on silver first to inconvertible paper and then to gold. They suspended silver coinage, devalued their currencies, and gradually augmented their gold reserves before establishing gold convertibility. These countries lacked the financial and economic resources of their richer counterparts and required more time to accumulate the reserves needed for the operation of a gold standard. While there was some scope for augmenting their gold reserves by borrowing abroad, as in the case of Greece in the 1880s, their debt-bearing capacity was limited, restricting the amount of external debt they could accumulate to finance the introduction of the gold standard.25

GEOGRAPHY OF THE GOLD STANDARD 125

National incomes per capita were also associated with the type of gold standard that countries ultimately adopted. A pure gold standard was a costly way of stabilizing the exchange rate, because it entailed the acquisition and maintenance of large amounts of bullion. A gold exchange standard, in contrast, allowed the central bank or government to hold its reserves in the form of interest-bearing assets that were convertible into gold, such as foreign bills deposited in a financial center like London, Paris, New York or Berlin. This was a cheap way to augment limited reserves, albeit at some loss of control of those assets. Poorer countries, for whom the budget constraint bound, were most likely to adopt the gold exchange standard.26 The connections between low per capita incomes (as an index of economic development) and the adoption of a gold-exchange standard suggest a monetary version of Gerschenkron’s ‘backwardness’ hypothesis. In many early industrializers, it was only necessary for the government to swap silver for gold on the world market and to revise the statute governing convertibility. Such countries already possessed smoothly-functioning financial markets and wellestablished central banks with which to manage the transition to gold. Late industrializers, by contrast, required more extensive government intervention to complete the transition. For those with inconvertible currencies and whose inherited reserves were inadequate to support the operation of a gold standard, the authorities had to launch an extended campaign in order to acquire them. This process often required the cooperation of a reserve-center country and complex intergovernmental negotiations. Concentration of the metallic base at the central bank The transition could be completed more quickly if the metallic basis of the circulation was concentrated at the central bank. When the nation’s metallic circulation was concentrated there, it was possible for that one entity to undertake the transaction that transformed the metallic basis of the national reserve. Thus, the Scandinavian countries, responding to the announcement by their major trading partner, Germany, that the latter intended to move to the gold standard, could beat their larger partner to the punch because their silver holdings were concentrated at their central banks, which could exchange them for gold at will. (See Figures 5.10–5.12.) In countries where silver circulated internally, in contrast, it was necessary to induce individuals to exchange their silver for gold. This the central bank or government could accomplish by suspending its offer to purchase silver at a fixed price and allowing the internal price of silver to decline relative to world levels. Individuals would respond by exporting silver and importing gold, on which the authorities might bestow legaltender status or stand ready to coin in unlimited amounts. But this process would take time, especially in countries where financial information flowed slowly.27 The Indian case illustrates the resulting dynamics. India was on the silver standard, the silver rupee having been established as the standard coin for British

126 MYTHS AND REALITIES OF THE GOLD STANDARD

Figure 5.10 Reserves of the Bank of Denmark, 1852–85 Source: Soetbeer (1889)

India in 1835.28 The post-1870 depreciation of Indian silver coin and notes against the moneys of the gold standard led to the above-mentioned complaints about exchange-rate instability and created pressure for monetary reform. Free coinage of silver was suspended in 1893, and the Indian government started issuing paper rupees. Although one could obtain rupees in exchange for gold, the reverse operation was not yet possible.29 When the government used Council Bills (technically claims on India) to finance its expenses, thereby monetizing its short-term debt, the exchange rate depreciated. Moreover, the transformation of the circulation required time. When Council Bills were issued, augmenting domestic liquidity, gold inflows slowed. Only after 1896, following fiscal retrenchment, did large amounts of gold begin to flow into the government’s coffers. In 1899 the British sovereign was made legal tender in India, and the government established a reserve in London. Even then, however, there was no statutory requirement that the government intervene to support the rupee. Following long discussions and substantial improvement in the reserves of the Indian Government, the system was completed in 1901–02 by the introduction of a policy of paying gold for rupees. The credibility of the new regime was fostered by the creation of the Gold Reserve Fund, on which gold claims could be drawn. Institutional innovation thus was used to launch the gold standard without withdrawing a penny from the silver circulation.30 Size Small countries were more likely than large ones to move from silver or bimetallism directly to gold. Completing this transition, as explained above,

GEOGRAPHY OF THE GOLD STANDARD 127

Figure 5.11 Reserves of the Swedish State Bank, 1850–86 Source: Soetbeer (1889)

Figure 5.12 Reserves of the Bank of Norway, 1850–86 Source: Soetbeer (1889)

required exchanging silver for gold on world markets. Small countries could do so without worrying about the implications of their transition for the relative price of the two metals. Large countries, in contrast, had to worry that their silver sales would destabilize existing bimetallic or silver standards and depress the value of the asset they were attempting to liquidate. Thus, the Scandinavian countries, whose silver stocks were small, could exchange them for gold without disturbing global markets. Larger countries like France and Germany had a more complicated task. In the early 1870s they held

128 MYTHS AND REALITIES OF THE GOLD STANDARD

Figure 5.13 Reserves of the Bank of Belgium, 1870–86 Source: Soetbeer (1889)

the equivalent of 3 billion French francs worth of silver, or six to ten times the annual flow supply. Dumping these reserves on world markets would have driven down the price of silver, frustrating efforts to accumulate a gold stock adequate to support monometallic convertibility. This encouraged larger countries to undertake the transition more gradually and to rely on cooperative monetary arrangements which might support the price of the metal that was undergoing liquidation. Cooperative monetary arrangements Belgium and Switzerland had the advantage, as members of the Latin Monetary union, that France was committed to purchasing their silver coins at a fixed price.31 This permitted them to unload silver without depressing its market price (Figure 5.13).32 Hence, the Bank of Belgium could move directly in 1874 from bimetallism to paying out only gold for its notes.33 As late as the eve of World War I, the Bank of France held half of all five-franc Belgian silver coins.34 Similarly, the 1874 Amendment to the Latin Union Treaty obligated the Bank of France to accept Switzerland’s silver coins at a fixed price. This made it easier for Switzerland to transform the metallic basis of its circulation and follow its neighbors onto gold.35 The Dutch case serves to illustrate the more complicated task facing countries which did not enjoy the support of a benevolent neighbor. The Netherlands moved from silver to gold in the 1870s.36 In 1872, the government decided to limit silver coinage. In May of the next year, when Germany’s intention of adopting gold convertibility was clear, the suspension of silver coinage was made permanent.37 On June 6, 1875, the government authorized the free coinage of gold florins, officially placing Holland on the gold standard.38

GEOGRAPHY OF THE GOLD STANDARD 129

Figure 5.14 Reserves of the Netherlands Bank, 1850–86 Source: Soetbeer (1889)

Figure 5.14 shows how the Netherlands altered the composition of its gold and silver reserves.39 This process did not result in the immediate and total dominance of gold, since to liquidate the country’s silver stocks threatened to depress the metals price on world markets. Though silver was no longer coined, it remained legal tender, and much silver circulated internally, in contrast to the situation in Scandinavia. For a considerable period, gold and silver continued to circulate side by side. This caused considerable difficulties for the Netherlands Bank, which constantly ran the risk of being drained of gold and flooded with silver.40 Like Belgium and Switzerland, Germany hoped to trade its silver for gold via the French bimetallic system.41 No sooner did it begin doing so, however, than France limited silver coinage. The German government was forced to reduce its silver sales: in 1879, when these sales were suspended, Germany had disposed of only a third of its initial stock. In effect, the two countries’ diplomatic sparring delayed Germany’s efforts to complete the transition while pushing France toward a de facto gold standard.42 Imperial ties If they possessed a colonial tie or were willing to cede control over their reserves to a foreign financial center, countries seeking to join the gold standard might accumulate not gold bullion but interest-bearing foreign exchange. Doing so could speed the accumulation of international reserves. Imperial ties could be formal, as with India and the Straits Settlements, which held sterling balances in London and the Dutch East Indies, which held exchange reserves in Amsterdam.

130 MYTHS AND REALITIES OF THE GOLD STANDARD

They could be informal, as with the dollar balances of the Philippines and Nicaragua, which adopted gold-exchange standards following military intervention by the United States.43 Thus, the Philippines went onto the gold standard following the American occupation of 1898. A special investigation under the authority of the U.S. War Department, prompted by the depreciation of the peso, led to an Act of 1903 establishing the gold exchange standard. Much of the country’s reserve was held as gold deposits in New York. Similarly, Korea went onto the gold standard following an agreement of August 1904 which confirmed Japanese protection of the country. It agreed to accept a Japanese financial adviser, to withdraw its depreciated coins, and to adopt the Japanese system of gold and token coinage.44 The Brazilian case is worth considering in some detail, since although Brazil had been independent since 1824 its dependence on London for capital imports gave its monetary affairs an imperial flavor. Initially, the country had some success in operating a gold standard. An official parity was established in 1833 and adjusted in 1844, and the paper circulation was maintained close to par for extended periods.45 The abolition of slavery in 1888, establishment of the Republic in 1889, and civil war in the south inaugurated a period of monetary instability.46 Not only did these events place financial demands on the government, but Rui Barbosa, the first republican finance minister, attempted to secure the support of leading bankers, investors and merchants by creating three regional banks of issue, under the financial elite’s control, with the power to issue notes backed by treasury bonds equal to three times the total currency in circulation. The milreis depreciated as a consequence and plummeted dramatically when a new civil war broke out in 1893. Financial scandals then caused the finance minister to be replaced, and his successors, free of the demands of a civil war, were better positioned to pursue balanced budgets, free trade and ultimately, gold convertibility. The regional banks were reorganized, strengthening state control. While the planter class that supported these governments benefited from the stimulus to exports provided by a depreciated exchange rate, it suffered from the impediment it represented to obtaining capital imports. London demanded the restoration of gold convertibility as a precondition for lending, regarding it as a signal of the country’s commitment to responsible finance.47 Moreover, the Rothschilds warned the Brazilians that a failure to put their financial house in order might threaten their very sovereignty, in the extreme provoking a foreign invasion.48 This constraint proved crucial. The Paulista governments that ruled from 1894 attached priority to withdrawing from circulation the inconvertible paper currency issued by their predecessors. But acquiring the reserves needed to reestablish convertibility was costly for an economy subject to violent export fluctuations. Coffee prices weakened in the mid-1890s; budget deficits soared and the milreis reached its nadir in 1898. That year the treasury finally reacquired the monopoly of the note issue. For a funding loan from the Rothschilds in London, the government pledged the revenues of the Rio customs

GEOGRAPHY OF THE GOLD STANDARD 131

house. It agreed to withdraw from circulation a sufficient quantity of inconvertible notes to offset any additional debt it issued and gave foreign banks authority to oversee the treasury’s operations. Provision was made for accumulating a gold reserve in London. Helped by these measures and by strengthening coffee and rubber prices, the milreis recovered. Reform was solidified in 1906 by exchange rate stabilization and the establishment of the Caixa de Conversao, essentially a currency board authorized to issue convertible paper notes backed by foreign exchange reserves (mainly pounds sterling). Despite the retirement of part of the inconvertible circulation as a consequence of the agreement with the foreign bankers, and despite the Caixa’s issue of convertible notes, only 40 per cent of the note issue was convertible as late as 1912. While the exchange rate had been stabilized, full return to the gold standard had not yet been effected. The catalytic role of war War could accelerate the transition of both the victors and the vanquished. For the former it could make available reparations in assets convertible into gold and provide the reserves needed to establish convertibility. For the latter it could provide discipline and constraints and thereby lead to the accretion of the requisite gold. The German case is a classic example of a battlefield triumph supplying the impetus and resources for a transition to gold.49 The indemnity provided the collateral that the government required to draw the proceeds of bills of exchange on London and Paris in order to acquire gold in both centers. Germany obtained about 1 billion marks in this manner.50 The Japanese case also illustrates the role of battlefield victory. Not only was Japan a low-income country by European standards, but a turbulent political environment disrupted the state finances. Hence, Japan required nearly three decades to complete the transition from silver to gold, a journey that was marked by experiments with bimetallism and an extended period of inconvertibility. The coinage in circulation in 1868, the year of the Meiji Restoration, was based on a system that dated to 1600. As in much of the rest of Asia, the standard was silver based.51 Successive debasements and the general disorganization of the monetary system led the imperial government to promulgate a new coinage law in 1871. It authorized the issuance of both gold and silver coins (though the silver trade dollar was given legal tender status only in 1878).52 However, the wars consolidating the Meiji Restoration, the need to suppress a rebellion in the southern provinces, and the abolition of feudalism all entailed financial demands on the government that were met through issues of inconvertible paper currency. Once these disturbances passed, financial reform could proceed. Convertibility was restored in 1886. By this time, Japan was a de facto silver standard country. Although the finance minister strove to accumulate a reserve sufficient to support convertibility, the task was expensive and progress slow. As in Germany

132 MYTHS AND REALITIES OF THE GOLD STANDARD

two decades before, a critical event in the country’s transition to the gold standard was a battlefield victory, in this case over China. The Shimonoseki Peace Treaty negotiated in 1895 specified that Japan was to receive her indemnity in silver taels. Japanese financial officials, concerned about the depreciation of silver, asked that the treaty be modified to provide for an indemnity to be paid in pounds sterling and raised by China through the negotiation of gold loans in Europe. These funds, deposited with the Bank of England, provided the resources to back Japan’s currency in gold and prompted the Diet’s passage of a gold standard law in 1897. War could also influence the monetary standard of the vanquished. France’s defeat in the Franco-Prussian War, for example, encouraged its government to adopt an early version of the ‘franc fort’ strategy: officials feared that a depreciating exchange rate would make it difficult to float on international capital markets the bonds (Rente Thiers) needed to finance the indemnity. The importance of the Rente Thiers thus reinforced the weight they attached to stabilizing the French exchanges against Britain and Germany, a goal which could be accomplished by pegging to gold. Greece provides yet another example. The country was defeated in the GrecoTurkish war that erupted in the spring of 1897. Forced to pay an indemnity in funds convertible into gold, and having been in default on her foreign debt since 1893, Greece opened discussions with her creditors in the hope of consolidating and funding the debt. Negotiations led to the establishment of an International Committee for Greek Debt Management. The Committee set as a condition for the provision of new funds the passage of a law prohibiting money creation as a means of financing budget deficits. The same article required the government to retire at least 2 million drachmas in banknotes each year until its domestic debt was reduced to 40 million drachmas. The peace treaty assigned a portion of tax revenues to servicing the external debt. In return, a 150 million French franc gold loan was provided to fund the war indemnity. With a binding constraint on the note circulation, increases in the demand for currency could be met only by importing gold. A decade of gold inflows provided the resources needed to establish gold convertibility in 1910. Historical contingencies The economist’s inclination is to search for systematic factors that can explain the timing of the gold standard’s adoption. Fully accounting for this timing, however, requires introducing a role for historical contingency. In Argentina, for instance, favorable harvests and a surge in British lending facilitated the transition to gold in 1900. Unusually good harvests and large wheat exports in 1898 first strengthened the exchange rate. Following the passage of laws in November 1899 providing for the creation of a conversion fund and mandating that paper pesos be converted into gold at a rate of 44 gold centavos, another good harvest resulted in a surge of exports and a gold inflow. These events raised

GEOGRAPHY OF THE GOLD STANDARD 133

the gold backing to 59 per cent of the note circulation.53 Argentina’s subsequent success in maintaining gold convertibility was at least partially attributable to capital inflows from Britain, made available by favorable conditions on the London market.54 The transition could be eased if discoveries of precious metals or changes in their prices fortuitously increased the value of reserves. Mexico, for example, was a significant producer of both gold and silver, but most of the precious metal it produced had long since circulated outside the country. It was able to move quickly from silver to gold between 1905 and 1908. The transition was eased by a rise in the world price of silver, the metal in whose liquidation the government was engaged, due to silver purchases by Italy and increasing global gold production.55 With the world market price of silver temporarily above its legal tender value in Mexico, large amounts of Mexican silver were exported to the United States and other countries in return for gold. Between 1905 and 1908 Mexico was able to transform its internal circulation from one based almost entirely on silver to one based primarily on gold, without depressing the price of the metal it was selling. It held a significant portion of its reserve as interestbearing assets in the United States.56 IMPLICATIONS FOR RESEARCH IN INTERNATIONAL MONETARY HISTORY In this paper we have sought to chart the geography of the gold standard. Our account has highlighted the late date of the move to gold and the variety of transition strategies. Whether a country with a currency convertible into specie operated a gold, silver or bimetallic standard at mid-century depended not so much on whether it was rich or poor as on the monetary standard of the foreign country or countries to which its transactions were linked. When it came to the distinction between specie convertibility and inconvertibility, however, domestic economic conditions came into play. In particular, there was a strong correlation between economic development, as proxied by the level of per capita incomes, and possession of a convertible currency. Most countries went onto the gold standard between the 1870s and the first decade of the twentieth century. We have enumerated the factors propelling this transition and analyzed variations in its timing. Factors shaping the transition include the afore-mentioned level of economic development, the magnitude of reserves relative to world specie markets, whether reserves were concentrated at the central bank, and the presence or absence of imperial ties. International politics, evident especially but not exclusively in times of war, had important consequences for the transition. Our analysis suggests, in contrast to most earlier core-periphery distinctions, that the prewar system had several cores and several peripheries. The classical gold standard did not revolve simply around the City of London; it evolved out of the British, French and German zones of economic influence and

134 MYTHS AND REALITIES OF THE GOLD STANDARD

consequently had several centers—London, Paris and Berlin—corresponding to several peripheries. While the capital imports that allowed countries at the periphery to acquire the reserves needed to establish convertibility did not flow from any one financial center, they nonetheless tended to follow geographic channels that had been established earlier in the century.57 Better understanding the operation of these monetary-cumfinancial blocs should be a priority for research. Finally, it is possible to argue on the basis of the evidence presented here that the gold standard was less a dramatic departure from the status quo ante than an attempt to restore the monetary stability that had prevailed before 1870. It was a way to solve problems, including exchange-rate instability and disruptions to international capital markets, caused by the collapse of the bimetallic system. It was a response to the failure of countries to agree on steps to sustain international bimetallism. To better understand how countries cooperated in the operation of the international gold standard thus requires understanding how they failed to cooperate in the maintenance of its bimetallic predecessor. NOTES * Much of the work on this paper was completed while Marc Flandreau was Visiting Professor at Stanford University. We thank Andrea Cu for research assistance. We are grateful to Jorge Braga de Macedo, Jaime Reis, Angela Redish and other conference participants for comments on an earlier draft. Remaining errors are the authors’ responsibility. 1 Late nineteenth-century monetary experience outside the gold standards European core is not entirely terra incognita. There exist studies of particular countries in Eastern Europe, Asia and Latin America. At the same time, a number of national cases remain virtually unremarked upon. And there exist only a handful of attempts to link the experience of countries at the core of the international monetary system with other parts of the world. The few pioneering studies that succeed in doing so are limited in scope. Thus, Ford (1962) focuses on the interaction of two countries, Britain and Argentina. De Cecco (1974) concentrates on the interaction of the gold standard world with large trading partners and bullion suppliers like Russia and India. See Hawtrey (1934), Foreman-Peck (1983) and Kindleberger (1984) for further efforts along these lines. 2 Here ‘trade’ is shorthand for the entire range of commercial and financial transactions that we consider below. 3 Indeed, within each of the ‘areas’ identified above, the convertibilityinconvertibility distinction was reproduced. In the French area, for example, Italy was unable to maintain full convertibility for much of the period. The same was true of Austria in the German area. In the British area, Portuguese and Brazilian convertibility was fragile. In Asia, China had been forced to suspend convertibility earlier in the nineteenth century. 4 In referring to this system as ‘the gold standard’ in our title, we follow common parlance, not to mention Ford and de Cecco, both of whose titles refer to the gold

GEOGRAPHY OF THE GOLD STANDARD 135

5

6

7 8

9

10 11

12 13

14 15 16 17

standard despite the fact that they devote attention to silver or inconvertible paper. Given the ‘currency’ of the gold-standard label, alternatives may do more to confuse than clarify. It is also worth emphasizing the distinction between a monetary standard and a monetary system. The latter connotes the existence of a number of institutions (a mint, a central bank, a government concerned to guarantee the enforcement of private contracts) that did not always exist in the nineteenth century. We choose 1868 because it coincides with the publication of Seyd’s Bullion and Foreign Exchanges, a rich source of information on monetary arrangements and financial transactions. Although Switzerland, for example, was on a dual standard, it did not grant free coinage to private parties. It did, however, give legal tender status to French coins, which were freely coined in France, implying that Switzerland was de facto fully bimetallic. Portugal raised its bimetallic ratio in 1847 and adopted gold convertibility in 1854. Martin (1977, pp.654–5). Even in Britain, the Bank of England had taken silver for the reserve of the Issue Department, until the mid-century rise in the metal’s relative price made it prohibitively expensive. The practice was abolished in 1848. Cottrell (1992, p. 223). That a late industrializer like Portugal had adopted gold convertibility while an earlier industrializer like Austria remained on inconvertible paper indicates that other factors, to which we turn momentarily, also influenced the choice. The two quotes are from Keynes (1925) and Young (1925). The two explanatory variables—the level of economic development and the condition of the public finances—were correlated with one another, with more advanced countries often having more highly developed and reliable budgetary systems. The connection is less than tight: some middle-income countries, in Scandinavia for example, had solvent governments and maintained specie convertibility, while Austria-Hungary, Russia and the Ottoman Empire, whose incomes per capita were not dramatically different from those of Scandinavia at this early date, had more serious public-finance problems, which can be characterized as symptoms of the ‘decaying empire’ syndrome. Although these correlations are loose, one might argue that the same political obstacles that delayed industrialization in decaying empires also gave rise to public finance problems that led to currency inconvertibility. Conant (1915, pp.225–6). The government resorted to paper issues backed by the revenues of the salt mines of Gmund, Hallein and Aussee. See Willis (1899a) and Conant (1915). The gulden continued to fluctuate until 1892. In 1879, the world market price of silver had fallen to the point where the paper gulden was again worth as much as its statutory silver content. The Austro-Hungarian authorities then suspended the free coinage of silver on private account to prevent silver inflation. See Yeager (1969). See Lazaretou (1992). For details on the Italian case, see Fratianni and Spinelli (1984). Redemption was first discouraged and then refused starting in 1857. Willis (1899b, p.86). Uchida (1900, p.19).

136 MYTHS AND REALITIES OF THE GOLD STANDARD

18 Collins (1986) reports coefficients of variation for a number of sterling exchange rates. These show that the stability of exchange rates between countries with convertible currencies was almost as impressive before the gold standard as under it. Note that while Amsterdam was on silver and Paris was bimetallic at the beginning of our period, by 1880 both countries had moved to gold. See below. 19 See Flandreau (1994a). 20 Conant (1915, pp.486–7); Ambedkar (1923, ch. 3). 21 De Cecco (1986, p.373). He has in mind specifically the second wave of industrializers such as France, Italy, Austria-Hungary, Russia and Japan. 22 See Fishlow (1989). As Fishlow’s comment in this volume reminds us, convertibility was neither necessary nor sufficient to insure access to foreign funds. But other variables like the decision to hold foreign exchange reserves in a foreign financial center could have a reinforcing effect (Lindert, 1969). 23 Many countries which went onto the gold standard in the final decades of the nineteenth century continued to hold significant quantities of silver. The Reichsbank’s silver holdings remained larger than its gold holdings until the mid-1880s, those of the Bank of France until 1890, those of the Austro-Hungarian Bank until 1894, those of the Netherlands Bank until 1904. Bloomfield (1959, p. 16). 24 The role of each of the factors we describe below should be thought of as holding other determinants of the transition constant. This is analogous to a multiple regression framework in which the dependent variable is the length of time required to complete the transition and a number of independent variables, starting with the level of national income, are considered in turn. 25 Lazaretou (1992, p.296). 26 Keynes (1913) classifies Russia, Austria-Hungary and Japan as possessing goldexchange standards. Bloomfield (1959) mentions in addition Canada, South Africa, Australia, New Zealand, the Netherlands and most of Scandinavia. 27 Even in financially sophisticated economies like the Netherlands, this process could be time-consuming. In addition, in countries with convertible currencies in which paper money (convertible into specie at par) circulated widely, it was necessary to obtain only gold sufficient to back some 40 or 50 per cent of the circulation. Where the circulation took the form of gold and silver coin, in contrast, it was necessary to obtain the gold needed to inject the additional gold coin. Since, by the final third of the nineteenth century, few countries relied heavily on the internal circulation of gold, differences across countries in this propensity affected only modestly the difficulty of completing the transition. Gold coin formed an important part of the internal circulation only in Britain, France, Germany, the United States, post-1897 Russia, Australia, South Africa and Egypt. 28 This act also authorized the coinage of gold mohurs. The gold mohur and the silver rupee being of identical weight and fineness, this effectively established a ratio between gold and silver of 15 to 1. When Californian and Australian gold discoveries led to a flood of silver from Europe to Asia, gold’s legal tender status was revoked and government treasuries stopped accepting it. (In fact, gold had ceased being legal tender in transactions between private persons in 1835; in 1853 the government stopped accepting it.) This put India on the silver standard, where it remained until 1893. Details on India may be found in Ambedkar (1923) and Spalding (1924).

GEOGRAPHY OF THE GOLD STANDARD 137

29 Mertens (1944 p.296) calls this a ‘one way gold standard.’ In target zone vernacular, this type of arrangement is known as a ‘reflecting barrier.’ 30 Another illustration of backwardness forcing institutional creativity was the accumulation by the Bank of Spain of a portfolio of foreign bills (principally on Paris) in order to implement an active exchange rate policy (while other countries merely pegged to gold). Thus, although Spain did not rejoin the gold standard, it was one of the first countries to adopt modern methods of exchange rate management. See Root (1898) and Nogaro (1910). 31 In fact, this guarantee had only been obtained in January 1874, through a special agreement, not in 1865 when the Latin Monetary Union was formed. See Flandreau (1993). 32 Belgium had weathered a series of stormy changes in its monetary legislation which contrasted with the dull stability of its exchanges. It operated a silver standard until March 1847, when it turned to bimetallism. In 1850, in response to the decline in the price of gold, it reimplemented the silver standard. Gold coins were demonetized in 1854, but their legal tender status was restored in 1861. See Perlman (1992). 33 See Luzzatti (1883) and Kauch (1950). 34 Conant (1915, p.283). 35 On the Swiss case, see Weber (1992). 36 It had switched from bimetallism to silver monometallism at mid-century, when the decline in the relative price of gold threatened inflation due to free gold coinage. 37 Because of peculiar features of the Dutch system, the exchange rate displayed a tendency to strengthen. Silver coins remained legal tender. Although they could not be struck, they could still be melted. Excess demands for Dutch currency, due to balance of payments surpluses, for example, were met by an appreciation of the florin relative to its intrinsic value. In contrast, excess supplies were met by melting down coins. 38 The new gold florins were slightly devalued relative to the previous gold-silver parity. 39 It also reveals that this expanded gold reserve was needed to meet the economic fluctuations of the 1880s. 40 See Luzzatti (1883). This strategy of suspending silver coinage and authorizing the coinage of gold was an attempt to transform the monetary standard, in anticipation of the technique followed by India some two decades later. The suspension of free silver coinage made currency relatively scarce, creating pressure for gold coinage and thereby working to endogenously alter the metallic basis of the circulation. A number of other countries, including Spain in the 1870s, pursued this same quantity-theory-inspired strategy. 41 Flandreau (1994) shows that France in fact possessed sufficient gold to have accommodated Germany’s monetary reform without destabilizing the gold-silver exchange rate. 42 Similar difficulties arose a century later, when the U.S. government attempted to dispose of some if its gold reserves following President Nixon’s closure of the gold window. See Salant and Henderson (1978). 43 A more extreme case was that of Puerto Rico, which had maintained a variety of different circulations under Spanish rule, although most coin was based on Spanish silver; its circulation was integrated with the U.S. monetary standard following

138 MYTHS AND REALITIES OF THE GOLD STANDARD

44 45 46 47 48 49 50

51

52 53 54

55 56

American acquisition of the colony due to the Spanish-American War. See Kemmerer (1916). Conant (1915, pp.567–8). There were also extended periods of depreciation and exchange-rate fluctuation, as Conant (1915, p.500) notes. The discussion that follows draws on Topik (1987). See Fishlow (1989). Topik (1987, p.36). In fact, it has been argued that Franco-German rivalry played a central role in the emergence of the gold standard. See Flandreau (1994b). The government kept an account in London, believed to have a balance as large as $20 million, which it used to purchase gold at propitious moments: Conant (1915, p.198). It is important to note that the war indemnity was not meant to facilitate the move to gold, but rather to weaken French political influence in Europe. Bismarck had imposed indemnities before on defeated rivals (as with Denmark in 1864) without using the proceeds to underwrite a transition to gold. See Mertens (1944) and Kindleberger (1984). The Tokugawa governments had minted both gold and silver coin (as well as some copper); at the time Japan was opened to commerce with the Western Powers (1854–59), the ratio of their prices was eight to one, compared to 16 to one in the outside world. Naturally, gold flowed out, silver in. Masayoshi (1899, pp. 179–80). Details in this paragraph are drawn from Masayoshi (1899), who was finance minister from 1881. See, for example, Ford (1962) and Fishlow (1989). Conant (1915, pp.510–11). One might argue that the Argentine Government’s policies of financial reform were responsible for it being on the receiving end of these flows. But reform, though necessary, was not sufficient; in other decades, London was largely unwilling to lend independent of policies in the debtor countries. After 1900, in contrast, large quantities of British capital flowed to Argentina and the other regions of regional settlement. The experience drives home the point that the movement onto the gold standard by Argentina, Mexico and other latecomers was facilitated by the state of the international capital market, whose general condition was largely independent of their own national policies. Kemmerer (1916, pp.538, 547). Keynes (1913, p.35). The role of gold production in the transition is also evident in Russia. Upon resumption in the 1890s, the gold reserves of the State Treasury and the State Bank amounted to approximately 50 per cent of the note issue. (Other domestic liabilities were backed by convertible foreign exchange reserves.) Nearly half was accumulated in the first half of the 1890s. Two-thirds of this total accumulated simply as a result of domestic production. This is not to minimize the importance of the policies that caused that gold to stay at home: these included steps to rein in the budget deficit by increasing revenues from the state railways and forests and to encourage agricultural modernization as a way of stimulating cereal exports. Fiscal and economic reform also encouraged lending by French and other foreign investors, which raised the Russian debt from 11 billion to 16 billion French francs between 1887 and 1900. This led Conant (1915, p.265) to conclude that ‘Substantially, the [Russian] gold standard was financed by means of borrowed

GEOGRAPHY OF THE GOLD STANDARD 139

capital.’ Without these ancillary measures, domestic gold production would not have sufficed. For further details on the Russian transition, see Willis (1899b). 57 For a congruent view, see Fishlow (1985).

REFERENCES Ambedkar, B.R. (1923), The Problem of the Rupee, London: P.S.King. Bloomfield, Arthur I. (1959), Monetary Policy Under the International Gold Standard, New York: Federal Reserve Bank of New York. Collins, M. (1986), ‘Sterling Exchange Rates 1847–1880,’ Journal of European Economic History 15, pp. 511–533. Conant, Charles A. (1915), A History of Modern Banks of Issue, New York: G.P. Putnam’s Sons. Cottrell, P.L. (1992), ‘Silver, Gold and the International Monetary Order 1851–1896,’ in S.N. Broadberry and N.F.R.Crafts (eds), Britain in the World Economy, 1870–1939, Cambridge: Cambridge University Press, pp.221–243. Crafts, N.F.R. (1984), ‘Patterns of Development in 19th Century Europe,’ Oxford Economic Papers 36, pp.438–458. de Cecco, Marcello (1974), The International Gold Standard: Money and Empire, 2nd edn., 1984, London: Francis Pinter. de Cecco, Marcello (1986), ‘The Choice of a Monetary Standard: National Dilemma and Supranational Solutions, 1890–1914,’ in Wolfram Fischer et al. (eds), The Emergence of a World Economy, 1500–1914, Wiesbaden: Franz Steiner, pp.371–382. Fishlow, Albert (1985), ‘Lessons from the Past: Capital Markets During the 19th Century and the Interwar Period,’ International Organization 39, pp.383–439. Fishlow, Albert (1989), ‘Conditionality and Willingness to Pay: Some Parallels from the 1890s,’ in Barry Eichengreen and Peter Lindert (eds), The International Debt Crisis in Historical Perspective, Cambridge: MIT Press, pp.86–105. Flandreau, Marc (1993), ‘On the Inflationary Bias of Common Currencies: The Latin Union Puzzle,’ European Economic Review 37, pp.501–506. Flandreau, Marc (1994), ‘An Essay on the Emergence of the Gold Standard,’ unpublished manuscript, Stanford University. Flandreau, Marc (1995), Cette Faim Sacrée d’Or: la France, le Bimétallisme et la Stabilité du Système Monétaire International, Paris: L’Harmattan. Ford, A.G. (1962), The Gold Standard: Britain and Argentina, Oxford: Clarendon Press. Foreman-Peck, James (1983), A History of the World Economy: International Economic Relations Since 1850, Brighton: Wheatsheaf. Fratianni, Michele and Franco Spinelli (1983), ‘Italy in the Gold Standard Period 1861– 1914,’ in Michael D.Bordo and Anna J.Schwartz (eds), A Retrospective on the Classical Gold Standard 1821–1931, Chicago: University of Chicago Press, pp.405– 454. Friedman, Milton (1991), ‘The Crime of 1873,’ Journal of Political Economy 98, pp. 1159–1194. Hawtrey, Ralph (1934), Currency and Credit, London: Longman. Kauch, P. (1950), La Banque Nationale de Belgique, Brussels: Sobeli. Kemmerer, Edwin W. (1916), Modern Currency Reforms, London: Macmillan. Keynes, John Maynard (1913), Indian Currency and Finance, London: Macmillan.

140 MYTHS AND REALITIES OF THE GOLD STANDARD

Keynes, John Maynard (1925), The Economic Consequences of Mr. Churchill, London: Macmillan. Kindleberger, Charles P. (1984), A Financial History of Western Europe, London: Allen & Unwin. Lazaretou, S. (1992), ‘Monetary and Fiscal Policies in Greece, 1833–1914,’ Journal of European Economic History 22, pp.285–312. Lindert, Peter H. (1969), ‘Key Currencies and Gold, 1900–1913,’ Princeton Studies in International Finance no. 24, International Finance Section, Department of Economics, Princeton University. Luzzatti, L. (1883) ‘Delle Attinenze dei Biglietti di Banca col Bimetallismo,’ Nuova Antolgia di Scienze Lettere ed Arti, Seconda Serie, Rome. Martin, David A. (1977), ‘The Impact of Mid-Nineteenth Century Gold Depreciation Upon Western Monetary Standards,’ Journal of European Economic History 6, pp. 641–658. Masayoshi, Matsukata (1899), Report on the Adoption of the Gold Standard in Japan, Tokyo: Government Press. Mertens, J. (1944), La Naissance et le Développement de l’Etalon-Or, 1896–1922, Paris: PUE. Nogaro, Bertrand (1910), ‘Le Problème du Change Espagnol,’ Revue economique internationale 4, pp.60–75. Nugent, J.B. (1973), ‘Exchange-Rate Movements and Economic Development in the Late Nineteenth Century,’ Journal of Political Economy 81, pp. 1110–1135. Perlman, Mark (1992), ‘In Search of Monetary Union,’ Journal of European Economic History 21, pp.313–332. Redish, Angela (1990), ‘The Evolution of the Gold Standard in England,’ Journal of Economic History 50, pp.789–806. Root, L.Carroll (1898), ‘Currency System of Spain,’ Sound Currency 5, pp.242–256. Salant, Stephen and Dale Henderson (1978), ‘Market Anticipations of Government Policies and the Price of Gold,’ Journal of Political Economy 86, pp.627–648. Seyd, Ernest (1868), Bullion and Foreign Exchanges, London: Wilson. Soetbeer A. (1894), Matériaux pour faciliter l’intelligence et l’examen des rapports économiques des métaux précieux et de la question monétaire, Paris and Nancy: Berger-Levrault. Spalding, William F. (1924), Eastern Exchange, Currency and Finance, 4th edition, London: Pitman & Sons. Topik, Steven (1987), The Political Economy of the Brazilian State, 1889–1930, Austin: University of Texas Press. Uchida, S. (1900), ‘The Gold Standard in Japan,’ Sound Currency 7, pp.18–30. Weber, Ernst Juerg (1992), ‘Free Banking in Switzerland After the Liberal Revolutions of the 19th Century,’in Kevin Dowd (ed.), The Experience of Free Banking, Routledge: London. Willis, H.Parker (1899a), ‘The Austrian Monetary Reform,’ Sound Currency 6, pp.1 14– 128. Willis, H.Parker (1899b), ‘Monetary Reform in Russia,’ Sound Currency 6, pp.82–110. Yeager, Leland B. (1969), ‘Fluctuating Exchange Rates in the Nineteenth Century: The Experiences of Austria and Russia,’ in Robert A.Mundell and Alexander K. Swoboda (eds), Monetary Problems of the International Economy, Chicago: University of Chicago Press, pp.61–89.

GEOGRAPHY OF THE GOLD STANDARD 141

Young, John Parke (1925), Central American Currency and Finance, Princeton: Princeton University Press.

142

COMMENT Angela Redish

In this time of uncertainty over monetary regimes, it is natural to see what can be learned from historical experience. Correspondingly, over the last few decades the classical gold standard has been used as a case-study by those attempting to measure the impact of alternative monetary regimes for price stability, output cycles and economic growth.1 The papers presented in Chapters 3 and 5 look for rather different lessons—what are the causes of changes in monetary regimes and how are transitions between monetary regimes effected? While it is possible that the choice of regime may have implications for (for example) price and output stability, it does not necessarily follow that it was the search for such behavior that led to the regime change. Indeed, the two papers discussed here emphasize the role of historical contingency and political symbolism in the ascendancy of the gold standard. Let me begin with Alan Milwards’ discussion of the origins of the gold standard. The central thrust of his paper is that Germany’s adoption of the gold standard in the 1870s had a domino effect and led to the international spread of the gold standard. Germany’s decision, in turn, he attributes to the desire by a politically powerful liberal bourgeoisie, not for deflation as is sometimes supposed, but for economic development and growth.2 This group believed that the gold standard would ‘guarantee…a liberal middle class constitutional order’. The overall story here is internally consistent and yet I think there are gaps in the argument that need to be addressed to make the case compelling. I would like to see the link between German adoption of the gold standard and subsequent choices by other European powers expanded upon. Was the fall in the price of silver resulting from Germany’s silver sales the primary channel of influence, or was it the desire for fixed exchange rate with major trading partners? This latter effect, which Milward emphasizes for example in the case of Scandinavia, needs elaboration. For example, in Chapter 6, on Portugal’s adoption of the gold standard, Jaime Reis argues that the Portuguese in 1854 were rather nervous about adopting the same monetary standard as their major trading partner, for fear that the Portuguese economy would be less insulated from Britain. Similarly, there is now a debate over whether Canada should fix its exchange rate with the United States, a debate which to date has been won by the ‘no’ side.3

144 COMMENT

Let me turn to the rationale for the German decision. Milward presents relatively little primary evidence to document the claim that the decision reflected the desire of businessmen to link with sterling. Furthermore, the relative significance of the desire (a) to grow like free-trade Britain, (b) to have a political order like constitutionally liberal Britain, (c) to have a fixed exchange rate to promote trade with Britain and (d) to avoid having the indemnity paid in ‘depreciated silver’ are unclear. The first three factors require further elucidation in terms of timing. Britain adopted the gold standard in 1816 (or 1821), while Germany waited until the 1870s. Milward explains the delay in terms of the dramatic expansion of international trade, and especially Britain’s rising share of that trade in the 1860s, as well as the move to free trade in that era. However, the inflow of gold from California and Australia in the 1850s and 1860s must have played a major role in the transition. By the mid-1860s gold coin dominated the circulation of France and Germany, significantly reducing the cost of the adoption of the gold standard.4 Indeed a de facto gold money stock was a precursor of the adoption of the gold standard in both the United Kingdom and Portugal—the early adherents of the gold standard. The role of the war indemnity following the Franco-Prussian war remains something of a puzzle. It is now well understood that the amount of actual gold coin that passed between France and Germany was insignificant; however, the indemnity could have been important in one of two ways. First, it may have represented a significant transfer of wealth to Germany, who thus may have been able to ‘afford’ an expensive change in monetary regime. In that case, we need to have a better quantitative understanding of the cost of the regime change. A slightly different view would be that while the actual wealth transfer was small, liquidity constraints and limits on taxing power meant that it represented a significant increase in the funds available for the transition. While Milward’s paper presents a close-up view of the decisions in Germany in the 1870s, Eichengreen and Flandreau use a panoramic approach. They present two snap-shots of worldwide monetary regime choices, one in 1868 and one forty years later. These pictures highlight the change over time—the convergence, they argue, from a variety of monetary regimes to the gold standard. They then categorize the factors which made this transition more or less difficult. Chapter 4 emphasizes the distinction between inconvertible and convertible monetary standards and suggests that the transition from inconvertibility was the more significant transition. This is an important distinction, one that the authors could have followed a little further, by providing more discussion of the nature of inconvertible monetary regimes and by integrating the distinction between convertible and inconvertible regimes more tightly into the body of their analysis. More specifically, it would be of interest to know whether or not their classification of convertible and inconvertible regimes is consistent with Bordo and Kydland’s (1992) characterization of commodity money standards as

MYTHS AND REALITIES OF THE GOLD STANDARD 145

contingent rules. Were inconvertible regimes just contingent suspensions of convertibility?5 In Table 5.1 the authors assign a particular monetary commodity to each inconvertible regime, where the appropriate commodity is defined by the statement that ‘in a normal state of affairs’ debts would be paid in a money convertible into that commodity. This suggests that at some level of commitment inconvertibility was a temporary phenomenon. If inconvertible regimes were just contingent suspensions of convertibility, then examining snap-shots of a point in time seems inappropriate, as the monetary regime is a dynamic process. For example, in 1868 the United States had an inconvertible currency, which markedly reduces the high correlation the authors find between levels of economic development and convertibility status, but I would interpret it as a prototypical contingent suspension of convertibility. If inconvertible regimes were not just temporary phenomena, then were countries with inconvertible moneys using the money as a tax source, or were the governments of such countries ahead of their time in appreciating the resource cost of commodity money?6 In either case, why would a government that has established an inconvertible currency chose to adopt a gold standard? Flandreau and Eichengreen argue that the ease of transition to the gold standard depended on (a) the level of economic development, (b) the concentration of reserves, (c) economic size, (d) imperial ties and (e) historical contingencies. The actual variable being determined here is not entirely clear. It is described as ‘the speed of transition but the discussion might apply equally to the cost of transition or the starting date of transition. It would be interesting to see more discussion of how the speed of transition, starting date of transition, and even the decision to change regimes, were interwoven. Turning to the ‘explanatory variables,’ the authors emphasize that the level of economic development affects not the choice of monetary commodity but the choice of convertible versus inconvertible money. This is in contrast with late nineteenth century discussions which frequently argued in favor of gold as the choice of developed countries. If a third snap-shot were taken—say, in 1888— would there still be no correlation between monetary commodity and economic development?7 Still, if inconvertibility reflected the use of seignorage for an administratively easy but economically inefficient tax, economic theory would readily predict the correlation between economic development status and convertibility status. The authors argue that the level of economic development helped the transition because the transition was costly and richer countries could afford it. It is not clear to me that this would affect the speed of the transition more than the starting date of the transition. Additionally, there may have been more than a wealth effect: more developed countries may also have found more use for gold coins in circulation, and have started with a larger stock of the same.8 Here, particularly, it would have been useful if the cases of transition between silver and gold and between paper and gold were treated separately.

146 COMMENT

The role of the ‘concentration of the metallic reserve in the central bank’ is very interesting and deserves amplification, perhaps through a theoretical discussion preceding the examination of particular examples. For example, the ‘concentration of the metallic reserve in the central bank’ could (necessarily?) imply the circulation of convertible paper money but convertible paper money did not imply a central bank. Which was the critical factor? While India is proposed as the counter-example it is complicated by the decision to monetize the debt. A theoretical analysis could have presented a clearer example of the contrast between the cost to the government of (a) a central bank selling its silver to buy gold and (b) a central bank agreeing to buy silver at (the inflated) par for a limited time. In the discussion of imperial ties the authors present an extended discussion of Brazil’s difficult transition to the gold standard, which certainly confirms the significance of these ties. Yet counterposing these ties were a series of domestic issues, some idiosyncratic, some probably systemic, which more than offset these ties. Brazil’s experience would seem to depend far more on historical contingency than on its imperial ties.9 Which brings us to ‘historical contingency,’ a category akin to dummy variables and error terms. The authors explicitly emphasize the role of contingencies in Argentina, Japan, Germany, France, Greece and Mexico. Economic historians tend naturally to be determinists and to look for consistent patterns, so the large role that Eichengreen and Flandreau find for contingency distressed me somewhat. My distress was alleviated when I remembered that it is the transition to the gold standard, not the decision to adopt the gold standard, which they so convincingly attribute to contingency. These minor caveats aside, let me turn to the bigger picture of what we can learn from all this. In a nutshell, Eichengreen and Flandreau have shown that acquiring a convertible gold-based currency was expensive, and that despite this almost everyone did it in the late nineteenth century. But the careful elaboration of the difficulties of the transition to the gold standard makes it even harder to understand the rationale for the transition. Perhaps part of the problem lies in the multiple dimensions of the gold standard. Its characteristics included provision of a system of fixed (nominal) exchange rates, convertible currencies, a British-style monetary order, and a monetary system that obviated the problems of bimetallism and Gresham’s Law. In theory, monetary authorities seeking one of these objectives might have ended up with a gold standard, and therefore each of the other characteristics as sideeffects. Was one objective paramount? The idea that convertibility was the raison d’être is appealing, particularly today, when the costs of the lack of a nominal anchor for the economy are so evident. Convertibility implied that the monetary system could not be manipulated expediently by the government to acquire tax revenue, and it imposed fiscal discipline on its adherents. Yet what were the costs of inconvertibility? A frequent answer is that there would be less access to capital

MYTHS AND REALITIES OF THE GOLD STANDARD 147

markets (see, for example, Chapter 2 by Bordo and Schwartz) but countries on inconvertible standards typically issued gold bonds to foreigners; would issuing an inconvertible currency make a government more likely to default on those bonds? Or lower the probability, since the government could, ceteris paribus, draw on more domestic resources? Did paying the costs of convertibility generate a return in terms of a lower interest rate gold (or hard currency) bonds?10 Alternatively convertibility in the periphery may have had the same rationale as in the ‘core’—that is, domestic economic concerns. Convertibility would convince domestic agents that there would be no (or minimal) seignorage tax, which might promote domestic monetization and growth. Perhaps the goal was less specifically convertibility but, rather, fixed exchange rates. Certainly there were many countries for which this seems the relevant criterion, particularly amongst those changing from a silver standard to a gold standard.11 What needs to be spelled out then are the benefits of a fixed exchange rate—as distinct from the provision of a nominal anchor. That is, today the choice of a fixed exchange rate is frequently described as a way of tying the hands of the monetary authority and imposing fiscal discipline. But for economies on commodity standards (whether bimetallic or silver), the nominal anchor was already in place, and a fixed exchange rate was just that. This is the world typically inhabited not by international finance economists but by monetary theorists who seek to find the origins of money. What is particularly interesting is that these theories (for example Jones (1976), Kiyotaki and Wright (1989, 1994)) focus on a world of goods trade, yet historical experience suggests that it is the transactions in the capital market that motivated the transition to the fixed exchange regime. Perhaps the transition to the gold standard was primarily an attempt to eliminate the problems of bimetallism, undervalued coins and Gresham’s Law. Elsewhere (Redish, 1990), I have argued that this was the case in Britain in the early nineteenth century, and such ideas were certainly common in France in the 1860s. Furthermore, Reis argues in Chapter 6 that Portugal in 1854 adopted the gold standard not because of admiration of the British political system, but because the system worked—it provided a useful and efficient medium of exchange. If convertibility was the objective, why gold? If monometallism was the objective, why gold? If fixed exchange rates were the objective, why gold? There are many possible answers: gold was abundant; gold was the choice of monetary standard of the leading economy of the day; gold was the choice of monetary standard of a politically liberal regime; gold was more valuable. A standard tool whereby economic historians gauge the significance of alternative hypotheses is to use a counterfactual methodology. If silver had been the abundant metal in the 1870s, would we now be discussing the classical silver standard? If England had been a small economy with a liberal political regime, would Germany still have opted for gold? If England had been a rapidly growing large economy that didn’t trade with Germany, would the Germans still have wanted gold? And so on!

148 COMMENT

In fact silver was the abundant metal for much of the early nineteenth century, and yet there had been a relatively easy transition to gold in the middle of the century. Perhaps it was harmony created by the fact that in the 1860s the leading national economy used as a monetary standard the same metal as comprised the circulation of many economies of Europe that led to international action to create a common currency area and to prevent a return to a discordant world. NOTES 1 Bordo (1993); Bayoumi and Eichengreen (1994). 2 See, for example, Gallarotti (1993). This argument is frequently ascribed to American political forces promoting the gold standard, but recent work by Jeffrey Frieden (1993) argues that the silver forces were advocates of depreciation rather than inflation. 3 Of course the similarity should not be overdrawn. It is important to the case that these were commodity moneys not fiat moneys. 4 See Flandreau (1994) and Oppers (1994) for the case of France. 5 Elsewhere in the paper, the authors suggest that inconvertibility was sometimes a stage that poorer countries would go through during the transition between silver to gold. Can this transition be an example of a ‘contingency?’ 6 ‘Appreciating’ here should perhaps be read as ‘having the institutional framework to eliminate.’ 7 For reasons of symmetry, as well as content, the authors might want to replicate Figure 5.1 for 1908. 8 I am not sure what to make of the authors’ comment that ‘few countries’ had a significant circulation of gold coin, when the few countries included the United Kingdom, France, Germany, and the United States—clearly the G-4 of the day. 9 This is not a contradiction with the authors who are only saying that imperial ties had a positive significant effect, not, they explain, a large share of the cross-country variance. 10 Perhaps through the more subtle ‘lemons effect,’ as suggested by Bordo and Schwartz. 11 However, the fall in the price of silver during the 1880s makes it hard to discriminate between the desire for a fixed exchange rate with the ‘gold bloc’ and the desire to avoid a depreciating exchange rate.

REFERENCES Bayoumi, T. and B.Eichengreen (1994) ‘Economic Performance Under Alternative Exchange Rate Regimes: Some Historical Evidence.’ In Peter B.Kenen, Francesco Papadia and Fabrizio Saccomanni (eds) The International Monetary System. Cambridge: Cambridge University Press, pp.257–97. Bordo, Michael D. (1993) ‘The Gold Standard, Bretton Woods, and Other Monetary Regimes: A Historical Appraisal.’ Federal Reserve Bank of St. Louis Review, Vol. 75, No. 2, March/April: 123–91.

MYTHS AND REALITIES OF THE GOLD STANDARD 149

Bordo, Michael D. and F.Kydland, (1992) ‘The Gold Standard as a Rule.’ Federal Reserve Bank of Cleveland Working Paper No. 9205, March. Flandreau, M. (1994) ‘As Good as Gold? Bimetallism in Equilibrium, 1848–73.’ University of California, Berkeley, mimeo. Frieden, J. (1993) ‘The Dynamics of International Monetary Systems: International and Domestic Factors in the Rise, Reign, and Demise of the Classical Gold Standard.’ In J.Snyder and R.Jervis (eds) Coping with Complexity in the International System. Colorado: Westview Press. Gallarotti, G. (1993) ‘The Scramble for Gold: Monetary Regime Transformation in the 1870s.’ In M.Bordo and F.Capie (eds) Monetary Regimes in Transition, Cambridge: Cambridge University Press. Jones, R. (1976) ‘The Origins and Development of Media of Exchange.’ Journal of Political Economy, August, Vol. 84:757–75. Kiyotaki, N. and R.Wright, (1989) ‘On Money as a Medium of Exchange.’ Journal of Political Economy, August, Vol. 97:927–54. (1994) ‘A Search Theoretic Approach to Monetary Economics.’ American Economic Review, Vol. 83 March.: 63–77. Oppers, S. (1994) ‘Arbitrage in Bimetallic Money Supplies: Evidence from the Exchange Rate.’ University of Michigan, March, mimeo. Redish, A. (1990) ‘The Evolution of the Gold Standard in England.’ Journal of Economic History, Vol. 50, December: 789–806.

150

COMMENT Albert Fishlow

The two papers by Michel Bordo and Anna Schwartz (Chapter 2) and by Marcello de Cecco (Chapter 4) are important efforts to place the gold standard, particularly in the classical period between 1870 and 1914, in an appropriate historical perspective. Bordo and Schwartz go beyond such a goal, substantial enough on its own, and offer observations on the interwar and Bretton Woods periods in addition. Moreover, these two papers stand in apparent significant contrast. De Cecco, as he has done before, regards the gold standard more as myth than practice. He sees increasing conflict, increasingly over the last two decades prior to World War I, between the rapidly evolving international financial markets and domestic objectives. In particular he emphasizes the potentially adverse consequences of short-term capital movements upon the economic fate of semi-peripheral countries. Not surprisingly, he draws extensively upon Italian and Austrian experience. Bordo and Schwartz emphasize three central points when explaining why the gold standard did work: first, the specie standard was admittedly a contingent rule, but during the interval before World War I ‘the basic convertibility rule was followed by a larger group of countries’. One of the reasons this was so was that the period was one of relatively stable growth and political stability; and the different experiences of the core and peripheral countries, even before 1914, can be attributed to different stages of economic development. But as I hope to show in my brief remarks, both of these views—de Cecco, and Bordo and Schwartz—are correct. What is essential to appreciate is the different motivations of different countries at different times in adhering to or deviating from the standard. After all, even Portugal, one of the first adherents to the gold standard in 1854, suspended convertibility for forty years from 1891, before rejoining for only 82 days in 1931. That story is told extensively in Part III of this volume. National interest hardly disappeared in the gold standard interval; indeed, rising nationalism before 1914 is widely cited as one of the factors contributing to the war. Its effects helped to determine the operation of the system, both then and in later periods as well.

152 MYTHS AND REALITIES OF THE GOLD STANDARD

Specifically, de Cecco focuses rather too much on short-term capital flows and their irregularity as factors explaining why the gold standard became progressively inoperative. To be sure, as Bloomfield noted and studied, shortterm flows tended to increase at the end of the gold standard period. And the central banks of the period reacted slowly, if at all, to them; they played the rules of the game badly. But the magnitude of such commitments and their destinations, relative to long-term investment, can be exaggerated. When de Cecco writes, moreover, that ‘short-term capital movements were the vehicle which carried the new flexibility and shook the national fix-price systems and the policies based upon them,’ he exaggerates a bit. After all, the movement toward monopoly and trusts was current in exactly this same period, rather than previously. The basic point is simply that long-term flows for investment greatly dominated the flows to peripheral countries. It was the anticipated profitability of railroad, mining, public utility, and other investment that motivated the significant flows outward from England, France and Germany, and even from the United States to Latin America. De Cecco speaks about the ‘central banks and treasuries of debtor nations’ with such passion and vigor, but they were the agencies of Italy and perhaps one or two other countries very near the center of the system in Europe. He thereby excludes the very large flows going out to the United States, Latin America, Australia, Canada and India at the end of the century. Moreover, he is really focusing upon the Paris markets, when he speaks of capital flows; and he is speaking particularly about Italy when he analyzes the critical choice of intervention in either the exchange or the bond market. For peripheral countries as a whole, the role of speculation and arbitrage was much less important and the flows of truly long-term commitment more so. The market of importance was London, and the developmental investments were substantial. In 1914, as I cite elsewhere, more than 50 per cent of France’s is long-term foreign investment went to Europe, and more than 40 per cent of Germany’s; for the United Kingdom, whose total commitment was a third greater than the sum of the two, only 5 per cent went to Europe; the rest flowed to the countries mentioned above that did not even have central banks. An important additional point to appreciate, even with long-term flows, is their inherently cyclical variation. This pattern, conveniently put aside until suddenly recalled by the debt crisis of the 1980s, creates an inevitable imbalance between inflows and outflows. The former initially are emitted in massive amount to finance new projects, while the latter are delayed, growing continuously and cumulatively. At some point, net inflows become smaller and smaller, and countries are soon in the midst of an impossible balance of payments situation. Payments cease, and rescheduling follows. This was exactly the pre-1914 pattern; and the 1890s as a whole give ample evidence of its importance. What made matters simultaneously worse and better was the lagged pattern of primary production in these peripheral countries. Investment was not entirely

COMMENT 153

wasted, despite the incidence of failure and the need for debt relief. Railways were constructed and did function. But that increase in agrarian output only came later and saw the beginning of recovery during which exports of the agricultural and mineral sectors grew. After a while those larger sales initiated a further new round of investment, which was inherently long-term cyclical and in advance of production because of the indivisibilities involved. The size of the initial commitment was especially large when the source of production was distant from the market and much of the area neighboring railway investment could not be exploited. Where de Cecco rings true is in his insistence upon the growing problems plaguing the European borrowing countries that were part of the system, but peripheral to it. For them, the objective was long-term development. But the reality was a need for central bank action to offset potential negative effects of financial crises involving short-term flows. However, one should not generalize his discussion of this difficulty to the operation of the capital market as a whole. Bordo and Schwartz focus upon the long term. Their treatment is extensive and rich. I wish here to single out only three topics for brief discussion: the relationship between convertibility and new flows of investment during the period before 1914; the role of peripheral countries to the system; and their focus upon disturbances as a means of separating out policy actions and environmental changes. In analysing the first of these, the authors choose annual flows to Argentina and the United States as a basis for assessing the relationship between adherence to gold standard rules and investment. They conclude, recognizing the need for further research, that it appears that ‘restoration of convertibility during suspension and threats to convertibility during adherence seem to be associated with increases and declines in capital calls on new issues of securities in London.’ I would somewhat disagree. In the instance of Argentina, the largest flows during the first of two cycles considered, in the late 1880s, came despite suspension of the rules. Nor is this surprising. After all, neither Brazil nor Mexico, also large recipients of investment, were convertible at this time. Moreover, the important thing to appreciate is that large long-term capital flows did not lend themselves to gold standard conditions of re-establishing equilibrium. They inevitably went beyond limits, creating domestic expansions and price increases, and were followed by banking crises. In the Argentine case, the gold premium on the paper peso had already reached 94 per cent in 1889, despite record inflows in that year. Moreover, as commented above, the investment was followed only subsequently by output increase. That was also characteristic of the United States in the 1830s, 1850s, and 1870s. More interesting is the case of Argentina’s resumption in 1899 and the subsequent increase of large flows of foreign resources. During that later period some 47 per cent of gross fixed investment was estimated to be foreign-financed. The Argentina Republic, in 1914, as noted by H.E.Peters, the author of The

154 MYTHS AND REALITIES OF THE GOLD STANDARD

Foreign Debt of the Argentine Republic, ‘enjoyed a credit rating exceeded by none except the old creditor nations of Europe’ (p.49). But was it due to gold convertibility? Brazil increased its flow considerably in the first years of the twentieth century, as did Mexico and Chile, without yet committing to a fixed exchange rate; indeed, when it did, in 1906, it was to prevent the continued appreciation of domestic currency. There is obviously no fully satisfactory test. Certainly, Argentina opted to return to the gold standard in expectation of a favorable reception abroad; but whether it was necessary, versus the attraction of the accompanying rise in quantity and prices of exports, is much more difficult to assess. Given the earlier experience, and that of other Latin American recipients, I am cautious of giving too much credit to gold standard adherence. For the United States, I am similarly skeptical. The reason is the same. Cyclical events and decisions regarding currency backing interact in ways that are equally difficult to disentangle. According to the authors, capital calls ‘rebound significantly only after the 1900 Gold Standard Act,’ but this was precisely when they had started to increase worldwide. In the second instance, I point to the limitations of the treatment of peripheral countries in the post-World War II period. Granted that many of the recipients of capital flow before 1914 had colonial status, one understands the small number of developing countries included in the 21 country sample in that era. But it becomes more difficult to justify a constant sample as a basis of generalization in the Bretton Woods period, when independence becomes the rule. One only has inflationary Argentina, Brazil and Chile representing a much broader group of countries that extended to Asia and Africa, whose monetary experiences were rather different. Surely, this diminishes the validity of conclusions that can be drawn from extending the period of analysis beyond 1914. Finally, I wish to comment briefly upon the effort to use supply and demand disturbances to determine the relative importance of environmental impacts versus policy actions. Two observations are in order. They will not be a surprise to the authors, whose caution is noteworthy; but they may be a useful reminder to the reader not to exaggerate the findings unduly. First, the basic assumption required to perform the vector autoregression technique, that demand shocks are temporary and supply shocks permanent, may not really hold. In the period before 1914, for example, the decision by Argentina to return to the gold standard may certainly be anticipated to have yielded a near permanent effect before the Great Depression; and some supply shocks affecting agricultural, rather than industrial, output would be temporary. Clearly in the more modern period, as industrial production rises and policy changes are frequently more temporary than permanent, the assumption holds somewhat better. Note, as well, as commented earlier, that several of these countries did not have fully functioning central banks before 1914, thereby raising additional questions about the meaningfulness of the exercise in that time period. How would demand shocks follow from policy interventions, when these were relatively few?

COMMENT 155

Second, in analyzing the different sources of shocks to a group so large and diverse, and over a period in excess of a century, treatment is necessarily limited. What could be helpful for this methodology to yield its potentially positive return is more focus and somewhat less breadth. One wants to be able to identify the major shocks and their sources—like the oil shocks in 1973 and 1979, or the beginnings of major business cycles, or adoption of a different monetary standard —rather than to have them historically nameless. The method, in other words, ought to reinforce historical analysis, rather than confound it. In sum, this paper, like de Cecco’s, has its strengths in setting out how the gold standard actually worked over the longer term in a number of different settings. But, like his as well, there are occasional weaknesses that derive from generalizing too far on the basis of limited information. Still, both sets of authors much more merit our thanks than our criticism.

156

Part III PORTUGUESE CURRENCY EXPERIENCE

158

6 FIRST TO JOIN THE GOLD STANDARD, 1854 Jaime Reis*

INTRODUCTION Portugal’s adoption of the gold standard in 1854 is noteworthy for several reasons. Portugal was the first country in Europe, after the United Kingdom, to do so. It went against the then current trend which favored either demonetizing gold, as in the case of Holland, Spain or Naples, or a ‘wait and see’ posture while formally continuing with bimetallism, as in the case of France.1 It was a decision taken two decades before the major continental powers began to switch to gold monometallism, during the 1870s, and some thirty years before the onset of what has come to be termed ‘the classical gold standard era’. Although we still do not have a satisfactory theory to account for the wave of countries which adhered to the gold standard in the 1870s, recent comparative studies have attempted to provide a general framework of explanation.2 In this perspective, Portugal is interesting too, precisely because none of the factors advanced so far to account for the emergence of the gold standard as a multimember international system satisfactorily fits the events which led to the 1854 decision. In the first place, Portugal was hardly one of those richer and more industrialized nations for whom gold was preferable because of its greater convenience for large transactions. On the contrary, it was one of the poorest countries of Europe, with an economy that was predominantly agricultural and much less monetized than most. The argument that this was a choice aimed at favoring a rising and increasingly assertive ‘urban-industrial class’ to the detriment of a relatively less influential agricultural sector also fits the situation poorly. In Portugal, the fact that the scales of power were clearly tipped in the other direction makes this view implausible. On the other hand, if the aim of the reform was to benefit a debt-prone landed interest, it was gold, not silver, that was the ‘weak’ standard in the early 1850s and therefore likely to achieve this. For the socially and politically powerful agro-export interests, who might have stood to gain from the better terms of trade that a ‘weak’ standard made possible, the gold option made little difference since it merely placed Portugal on the same

160 PORTUGUESE CURRENCY EXPERIENCE

monetary system as its main trading partner, Britain. Lastly, and to make matters still less clear, it should be noted that commercial farmers, who probably made small payments to a greater extent than any other class of entrepreneurs, for this reason alone should have preferred silver since it reduced their transaction costs. In the second place, there is little evidence to show that the adoption of the gold standard by Portugal had anything to do with the dominance of an ideology of gold. In the early 1850s, rather than being drawn to gold as a symbol of sound practice and a badge of honour and decency, the strongest doubts were voiced concerning gold’s soundness as a monetary standard. The likelihood of further immense amounts of it descending upon the civilized world, as a result of mining developments in North America and Asia, did nothing to allay such suspicions. Similarly, while it has been noted that the rise of free trade ideals after the middle of the century was often associated with pro-gold standard sentiment, in the Portuguese case this could hardly have been the case. Although lip service was paid occasionally to free trade ideals, protectionism for both agriculture and industry was a firmly entrenched policy which was unlikely to be altered in a country whose fiscal machinery rested squarely on tariff revenues. Finally, with only one other country on the gold standard at this time, albeit one that was very much envied for its economic success, this was not enough to bring about the ‘chain effect’ whereby country after country adhered to the gold standard during the 1870s and 1880s simply because they could not ‘afford’ to be left out of this arrangement. In any case, in the Latin world, bimetallic France also commanded enormous respect, both politically and economically, and in the 1850s was a major source of capital and technological transfers to Southern Europe.3 This paper attempts to explain the circumstances which led to the adoption by Portugal of the gold standard. It starts with an outline of the country’s monetary conditions during the late 1840s and early 1850s. This is followed by an analysis of the monetary, political and broad economic factors which prompted monetary reform in 1854. A fourth section examines the details of the solution adopted and tries to explain why the silver and bimetallic options were rejected. A final section concludes. THE MONETARY REGIME BEFORE THE GOLD STANDARD Like many other countries during the first half of the nineteenth century, Portugal was bimetallic in theory, while in practice it oscillated between a predominance of gold and a predominance of silver. This depended on the current relation between the legal and the market values of the gold-silver ratio and on the arbitrage flows which a gap between them could induce. Thus, during the first two decades of the century, it appears to have been predominantly on a silver standard, until ‘given the need to bring back gold coin, which at present does not circulate’ a law of 1822 altered this ratio, from 13.5:1 to 16:1, and put

FIRST TO JOIN THE GOLD STANDARD 161

an end to the legal export of gold coin. From the mid-1830s, the de facto silver standard seems to have returned, as a result of the measures taken to encourage the circulation of this precious metal. In 1835 the official ratio was reduced to 15. 5:1 and the mint started buying silver at a premium, to convert it into coin.4 The practical effect of this can be seen in the fact that from 1836 to 1846, the Bank of Lisbon, the only significant institution of its kind in Portugal at this time, held almost only silver in its vaults, gold representing typically no more than 5 per cent of its holding of coin.5 In common with most of the then civilized world, a second feature of the monetary system was the circulation of a variety of foreign coin alongside that which was minted in Portugal.6 This can be traced initially to the presence of foreign troops, during the Napoleonic period and on several other occasions subsequently, which led to the spread of both Mexican and Spanish pesos and of English gold. Another major cause of this prevalence of foreign commodity money was the failure of the authorities to supply national coin in sufficient quantity and of adequate quality, apparently a consequence of the penury of the state’s coffers and of the technical inadequacies of the mint.7 Apart from the question of hurt national pride, the confusion which this situation brought about and the higher transactions costs it entailed encouraged more than one government to try and purge the system of at least some of these exotic elements. The fact that all legislative efforts in this direction appear to have met with failure is a good measure of how dependent the country’s stock of money had become on outside supplies by the 1830s. The governments of the new liberal regime established in Portugal after the civil war of 1832–1834 repeatedly tried to demonetize English sovereigns and Spanish, Brazilian and Mexican pesos and doubloons but had to recognize that the country could not do without at least some of them and accordingly declared them to be legal tender. The diversity of type, weight and fineness of legal coin and the changes in the legal gold-silver ratio were not the only monetary problems the authorities had to contend with, however. Another was the existence of a substantial mass of inconvertible paper money issued by the state, as a ‘temporary expedient,’ between 1797 and 1799, and of which a considerable part was still in circulation in the 1830s. What made this ‘papel-moeda’ such a scourge was the fact that by law it must enter into all payments at face value for at least one-third of the value of the transaction, yet at the same time, in the market, it discounted heavily and was subject to sharp fluctuations. For ordinary people, this meant a degree of uncertainty in their day-to-day transactions which they would have been glad to avoid and which was further compounded by a not insignificant risk of counterfeiting, due to the poor quality of the notes and to the widespread condition of illiteracy of the population. To complete the picture, reference must be made to the Bank of Lisbon, a joint-stock bank founded in 1821 with the privilege to issue convertible notes, something which it did successfully and on a reasonable scale. Apparently, the misfortunes of the holders of state unbacked notes were not sufficient to

162 PORTUGUESE CURRENCY EXPERIENCE

discredit other paper monetary instruments. Thanks to a solid backing and a generally cautious attitude on the part of the managers, those of the Bank circulated at par until 1846 and, except for a brief suspension in 1827, were readily converted into either gold or silver. By international standards, the arrangements we have just described do not appear to have stood out for their shortcomings. Matters in the 1840s, however, took a turn for the worse. In 1846, Portugal entered a period of intense social, political and economic turbulence which, aside from its other consequences, considerably magnified the distortions and malfunctioning of the monetary system. The starting point was the Maria da Fonte popular uprising, which swept the country in the spring of 1846 and was soon followed by civil war and foreign intervention, lasting in all for more than one year. Coin tended to disappear from urban markets owing to hoarding and to the need for the central authorities to transfer large amounts of cash to the theatre of military operations, in the north to pay for troops and their supplies. The situation was made worse by two extraneous circumstances which in the meantime caused an external drain of commodity money. One was a bad harvest in 1846, which necessitated imports of grain. The other was the cyclical downturn in the British economy starting sometime in 1845 and which reputedly caused a flight of capital to London.8 To fill this sudden monetary void, a spate of legislation ensued—nine decrees were enacted, between June 1846 and July 1847—the aim of which was to attract foreign coin into the country as quickly as possible. This was achieved by conferring upon it the status of legal tender and, at the same time, by officially overvaluing it relative to its local market price in order to make it worthwhile to import such coin. The French five franc silver coin, for example, which had been valued in the Lisbon market at 832 reis, was now monetized at 860. The Spanish gold doubloon was declared to be worth 14,600 reis, when previously it would have fetched only 14,060 for its weight in precious metal. The official value of the sovereign meanwhile was raised from 4,140 to 4,500 reis.9 The first of these laws gave legal tender to the sovereign, the French franc, the Mexican doubloon and the Spanish and Mexican silver pesos, and its successors merely widened its scope by admitting into circulation further coinage from other countries, namely Colombia, Brazil, Venezuela, Ecuador and the United States. In the second place, by devaluing the real slightly more in terms of the gold sovereign than of the remaining coinage, the new monetary rules ensured that the inflow of foreign money would originate mainly in England, a perennial feature of the decades to come. In the third place, they set an official gold-silver ratio of 16.5:1 at a time when the international market ratio was 15.9:1, something which, although probably not intended, was bound to drive out the silver coinage sooner or later and cause problems with small change.10 The monetary arrangements which emerged in the period 1846–47 were anything but clear, consistent or practical. The plethora of legal foreign moneys meant that the average individual now had not only to learn to distinguish between thirty-six different coins but also to check them for their metal content,

FIRST TO JOIN THE GOLD STANDARD 163

since there were at least three different proofs—the English (916), the French (900) and the Spanish (835)—not to mention a great many different legal weights. All of this obviously entailed higher transactions costs, as well as a higher risk of counterfeit. ‘Everyone knows that we have many factories of false coin in this country…and that they endeavour to imitate sovereigns and other real coins, either with other metals or with other alloys, or by altering their weight or proof.’11 In addition, there was the strong possibility of encountering substandard foreign money. One reason lay in the fact that some Latin American mints were given to issuing coin at less than its official weight. Another was the vagueness of the Portuguese law itself regarding the permitted margin of tolerance with respect to both proof and weight. Owing to the hurry in which it was enacted, and to a superficial knowledge of many of the coins it monetized, taken as a whole the new legislation utterly lacked internal coherence. Instead of a single standard of valuation, it encompassed several. The gold contained in English sovereigns, for example, officially was worth 603 reis per gram, while that in the old Portuguese peça was valued at 528 reis and that in the Spanish doubloon at 579 reis. In the case of silver, that in the Mexican peso was officially rated at 36.5 reis per gram but in the case of the Portuguese cruzado novo it only reached 33.0 reis per gram. Consequently, instead of a single official gold-silver ratio, there were several, depending on which pair of coins was being considered, a situation which was described as ‘the most incredible anarchy known to the present times, a consequence of those grandiose measures, which opened our doors to every kind of gold and silver coin from the old world and the new and made our circulation a veritable Babel.’12 The difficulties caused by the legalization of so many different foreign coins were not the only ones to plague the Portuguese monetary system of the late 1840s. The suspension of payments by the Bank of Lisbon, in May 1846, and its reconstruction into a new bank, the Bank of Portugal, in the following November, caused almost as many problems and provoked an even louder public outcry. The trouble here was the considerable mass of inconvertible Bank of Lisbon notes which were left in circulation after these events as part of the Bank of Portugal’s liabilities and which, despite legal injunctions to the contrary, met with a large and volatile discount in the market. Their total amount of initially 5 million reis nominal may have represented something like a fifth or a sixth of the national stock of money and during the first two years of the new situation, they suffered an average discount of 30 per cent, with maximum and minimum values of 50 per cent and 2.5 per cent, respectively.13 Since civil servants were paid entirely in these notes at their par value, it is easy to understand why this was one of the most important issues of economic policy that governments had to face during these years. In theory, there were two ways to solve the problems caused by these notes. One was for the state to pay the enormous debt which it had owed the Bank of Lisbon previously and which had now become one of the Bank of Portugal’s

164 PORTUGUESE CURRENCY EXPERIENCE

main assets. This would have provided the liquidity the Bank needed in order to restore the notes to convertibility, but was totally out of the question, given the utter penury of the treasury’s coffers. The other was to oblige the Bank of Portugal to close down altogether for not being able to meet its obligations, thereby rendering the Bank of Lisbon’s notes worthless. This, however, was no more acceptable as it would have entailed the ruin of not a few of the bearers of these notes, a sharp contraction of the money supply and the hostility of the Bank of Portugal ‘s shareholders, many of whom were influential in the highest financial and political circles. Instead, the notes were decreed to be the joint responsibility of both the state and the Bank and every effort was made to have them circulate at par. This was done by declaring them legal tender for the discharge of all private debts up to a certain proportion of the value of the transaction, as well as acceptable in full in the payment of taxes. Underlying this was the authorities’ belief that the market value of the Bank of Lisbon’s notes could be forced up simply by legislative fiat, thus automatically and cheaply solving the problem of inconvertible paper money. Despite the stiff penalties prescribed, market resistance was strong and not only did the discount on these notes not behave as predicted, but a great many contracts were made, illegally, in coin alone. Altogether six decrees or laws were enacted between November 1846 and July 1848 with a view to rendering this policy effective, mainly by trying different proportions for the notes entering into ordinary payments. With each one of them came the provision that all transactions previously undertaken were to be governed by the legislation in force at that time. The only result, however, was that economic agents had to contend with a mind-boggling array of monetary regulations and consequently faced ever-higher transactions costs on account of difficulties both of information-gathering and of contract enforcement. The proliferation of court litigation in connection with problems of payments in notes during these years clearly attests to this. THE TRANSITION TO THE GOLD STANDARD The monetary situation we have just described is the starting point for understanding what was to happen in 1854. It accounts for why monetary problems were such an intensely discussed issue in the years prior to this reform and why they generated such a strong clamor for the rationalization and simplification of the monetary system, as well as for the eradication of the elements of uncertainty in it.14 By itself, however, it is not sufficient to explain why such a radical reform was decided upon so early by European standards. After all, other countries faced similar monetary problems at this time and endured them for much longer without attempting major reforms.15 The Portuguese polity and economy too were able to live with an inadequate and costly monetary system for years without any moves being made to alter it significantly, except for the worse, as in 1846.

FIRST TO JOIN THE GOLD STANDARD 165

Nor was the option for monometallism based on gold such an obvious and indisputable one in the context of the early 1850s. On the contrary, at a time when a fierce debate over monetary standards was raging throughout European political, scientific and business circles, it is hardly surprising that its echoes should have reached Portugal and that its inconclusiveness should have been reproduced here too. Indeed, public discussion of the monetary question was pervaded with references to foreign authors, journals, and examples. Chevalier and others were cited constantly, the latest issues of The Economist and of the Journal des Débats were invoked, and the details of the monetary policies of various countries were adduced in defense of one point of view or another. A strong pro-silver party brandished the threats of inflation and loss of state revenue which were associated with gold, and held up Holland and Belgium as shining examples. Bimetallists lauded the wisdom of France and the United States for not committing themselves one way or the other in the face of so much uncertainty in world monetary affairs. They also claimed that if gold were to depreciate sufficiently, equilibrium would be reachieved by the stimulus this would give to silver production. The defenders of gold denied that there was too much of it in the world, despite the Californian and Siberian strikes, and claimed that its abundance in an expanding world economy was actually an asset and not a disadvantage. While referring to the United Kingdom’s lack of concern over the expanding supply of gold, they pointed out that, unlike silver, the gold standard was not subject to sudden destabilizing drains to Asia and, unlike bimetallism, did not require continual adjustments of the official gold-silver ratio.16 At the beginning of the 1850s, two new circumstances had a decisive effect on the evolution of the monetary problem and helped propel the country towards a much needed reform. The first was the solution of the inconvertible currency problem. From a high point of 54 per cent in June 1848, the discount on Bank of Lisbon notes gradually and steadily diminished and, by the end of 1850, had reached a virtually negligible amount (less than 5 per cent), at which it more or less stabilized. The immediate cause was a sustained policy of programmed monthly withdrawals and destruction of this paper carried out jointly by the government and the Bank, and which brought its supply down to roughly the level of the demand for it. Though by now 2,000 contos of notes still in circulation continued to be inconvertible—as they did until their final withdrawal in 1856—what was significant was that they had become a minor and trouble-free element in the Portuguese monetary system, which from 1851 could be considered therefore to have gone back again to convertibility under formal bimetallism.17 This had three important implications. The first was that policy-makers’ and politicians’ energies ceased to be absorbed by this issue and could be diverted to something else, such as the reform of commodity money. The plethora of pamphlets on how to eliminate the note discount dried up all of a sudden, no more laws to regulate the use of these notes came on to the statute books and Parliament ceased to show concern over the matter. The second was that it

166 PORTUGUESE CURRENCY EXPERIENCE

became easier for the authorities to clean the slate as regards contracts entered into under one of the many rules enacted at different times for the circulation of these notes. Given that they were now virtually at par and no longer so much in use, a new law could be devised which would reduce all monetary obligations to gold or any other metallic standard without this entailing any loss for the bearers of notes. Given that many of these had been civil servants, politically this was an important point. Finally, in spite of enormous revenue shortfalls, all through the late 1840s successive governments resisted the temptation to issue their own unbacked notes. Schemes of the kind were attempted, but always met with such resistance on the part of the Bank of Portugal, the commercial interests and many politicians that governments found it preferable to accumulate debts or simply not to pay, as happened with the interest on the foreign debt.18 Had they not done so, this would probably have released on to the market a large quantity of inconvertible paper which would have made it subsequently far harder to restore any commodity standard. The second circumstance was the well-known worldwide shift in relative goldsilver values, which started around 1850. Owing to a combination of increased gold production with the demonetization of gold in certain European countries, the ratio of the London market prices of these two precious metals began to fall from levels of around 16:1 in the mid-1840s, to 15.8:1 in 1849, to 15.7:1 in 1850. Apart from a great deal of uncertainty and speculation in monetary markets around the world, the result was a tendency for silver to disappear from circulation in bimetallic countries, either due to speculative hoarding or to its export to countries where the market and legal values were closer to each other.19 Given that the Portuguese legal gold-silver ratio of 16.5:1 was one of the highest internationally, these effects were bound to make themselves felt early and strongly.20 At the end of October 1850, gold coin, which had been at par in the market until then, began to depreciate relative to silver (by at 1 per cent), experienced another fall at the end of December (another percentage point) and reached a discount of 4 per cent in mid-February 1851 (see Figure 6.1). Naturally, a panic set in. The holders of gold sovereigns, in particular, rushed to the currency dealers to buy silver, many tradesmen refused to accept gold and the prime minister himself reported that ‘every day I am receiving complaints from the Heads of the Customs Houses that the export of silver and the import of gold are proceeding on a rising scale.’21 A commission of enquiry was appointed by the Queen to study the causes of these events and ‘to propose the changes in the present monetary system which the circumstances render advisable’22. Alengthyparliamentary debate into these causes and the possible remedies ensued. The government was toppled by a military coup shortly afterwards and never had the chance to hear the commissions views. Before this, however, it managed to pass a monetary law which was an important first step in the direction of an overhaul of the system, even if it only went some of the way towards this

FIRST TO JOIN THE GOLD STANDARD 167

Figure 6.1 Gold-silver exchange rate in Lisbon, 1850–54 Source: Diario do Governo

objective. One important contribution of this legislation was to put an end to the plethora of foreign gold in circulation by restricting legal tender to Portuguese and British coins alone.23 Ostensibly the aim was to rationalize a highly complicated and wasteful system. But it also served to withdraw from circulation a portion of ‘weak’ coin, i.e. gold, in emulation of what had lately been done in Belgium and Spain. In this sense it may be regarded as a timid attempt to shift the market gold-silver ratio closer to the official level; or even as a cautious move to pave the way for a silver standard.24 In 1851, both the government and the opposition realized that the heart of the problem lay in the discrepancy between the two ratios. The former lacked, however, the strength or the resolve to reduce the legal ratio in order to bring back the silver coin which had fallen into short supply. To this extent, by maintaining official coin values intact, the new law simply confirmed the slight over-valuation of the sovereign relative to all other coins and therefore its supremacy within the Portuguese system, which continued as a result to drift towards an informal gold monometallism based on sterling. As a palliative, and rather than meddle with the ratios, it included instead a duty on the export of all silver, whether monetized or not. With this it was hoped to deter the drain of this precious metal, since the surcharge represented some 12.5 per cent by value. Whether this would ensure the restoration of a payments system that would satisfy the needs of both large and small transactions by providing sufficient gold and silver coin was questioned from the start, however, and needless to say, a considerable part of this debate focused on the likely efficacy of such a measure

168 PORTUGUESE CURRENCY EXPERIENCE

in the context of a weakly organized state and of a habitually high level of smuggling.25 Contrary to what had been hoped, the market gold-silver ratio failed to return to par and the shortage of silver change continued to be an affliction. It might be thought that, under the circumstances, the scene was now set for the profound reform of the monetary system that everyone recognized as being necessary, indeed essential. Yet it took another three years before the final gold standard legislation was approved and implemented. Several reasons help us to understand why it took until 1854 for this to happen. One of them pertains to the intellectual climate which surrounded the problem and the genuine difficulty for many who pondered over it to know the correct solution. Throughout the parliamentary debate on the monetary question in 1851, speaker after speaker had to recognize that the situation was not yet sufficiently clear to them and that anything more than a temporary measure would be imprudent. No one was too sure about what was really causing the change in the international gold-silver ratio. Even less did they know how it would evolve in the future. Suppose, for example, that California, the Urals and Siberia were not to keep their promise as massive producers of gold?26 To this uncertainty was added that which came from the recent and frustrating experience with the multiplicity of laws which had been passed to try and control the market discount of the Bank of Lisbon’s notes. These had shown public opinion how complex a task it was to cure monetary ills with legislation. Nor was economic science felt to be all that much help. Typically, Fontes Pereira de Melo, who was to be the minister responsible in 1854 for the bill which instituted the gold standard, had to admit in 1851, after studying Chevalier and others, that ‘I do not feel sufficiently elucidated to say which is the best course of action.’27 On the other hand, José Lourenço da Luz, a banker, felt that ‘in this matter, if I were to follow the principles which guide me in my profession, I would say that up to a point theory is often contradicted by practice.’28 By 1853, when the public discussion of the monetary regime was taken up again, and by 1854, when the matter was finally resolved, these doubts seemed far weaker. Two or three years of persistently lower gold prices relative to silver had convinced most politicians and commentators that the gap between the market and mint ratios was unlikely to disappear soon. The continual vagaries of the precious metals trade had also persuaded most that monometallism was best, as it did away with any need to continually readjust the mint gold-silver ratio, a requirement that was universally considered to be quite impractical. There was also a growing conviction that the root of the trouble lay not in the demonetization of gold, which some countries had undertaken during the 1840s, but rather in the steady rise in the worlds gold output, a situation which was unlikely to change in the short run.29 Interestingly, neither did anyone any longer seriously argue, in 1854, that too little was known or understood about the monetary situation. Disagreements now were over how many milligrams of gold there really were in a standard sovereign, not over how much coin the national mint

FIRST TO JOIN THE GOLD STANDARD 169

had produced in the last few years. In fact, in 1853, output statistics for the last century were published for the first time, a sign of the new public interest in such data. A stable and strong government was another indispensable ingredient which had been absent earlier but which the country’s political evolution in the meantime had made available. As he was to claim later, finance minister Avila had intended, in 1851, a sweeping change of the monetary system which included the complete withdrawal of English sovereigns and their replacement with Portuguese coin, not a small matter. But he had lacked the necessary parliamentary support and even the 30 January law had encountered the ‘greatest difficulties,’ with the opposition paralyzing a weak and divided government with rumors that it was unable to put even this modest improvement into effect.30 A few weeks later, the government had been toppled and for some time its successor had been compelled to concern itself more with its own survival than with monetary reform. The Regeneration movement of 1851, headed by the Duke of Saldanha, not only changed the constitutional system and pacified the army but, after years of internal strife, ushered in a new mood of national reconciliation which gave rise, during the rest of the 1850s, to reasonably stable governments based on large parliamentary majorities. While these Regeneration governments enjoyed the requisite political conditions for reforming the economy’s institutional framework, they also held to the ideals of modernization, rationalization and simplification which were then increasingly in vogue in Europe. They introduced the metric system, modernized taxation and streamlined the Public Debt Office. New laws for mining and subsoil rights were passed, an up to date postal service was created and an ambitious program of public works, which included the building of the first two trunk railway lines, was launched. Under these conditions, to have reformed the currency as well, no matter how drastically, appears not only as a perfectly logical step but also as quite a practicable one. What really tipped the scales, however, was the behavior of the currency market. As can be seen from Figure 6.1, following the panic of early 1851, the exchange rate between gold and silver coin in the Lisbon market stabilized at 96– 97 per cent of par for the rest of the year and throughout 1852. At the same time, official exports of silver sank to almost nothing and a certain amount of this metal was even brought to the mint for coining. Yet after roughly two years of this, in May 1853, the value of the sovereign in relation to silver began to slide again and by the end of the year had hit the unheard of level of 90 per cent of par. A new panic set in, with reports coming in from the provinces already in May, when the discount of the sovereign was still at 94 per cent, to the effect that ‘there is nothing there but sovereigns and a little copper; commercial transactions are impossible to carry out except for the most insignificant ones and these only with great difficulty.’31 According to one school of thought, the cause of this was simply the government’s decision, at the end of 1852, to lift the export duty on silver. The

170 PORTUGUESE CURRENCY EXPERIENCE

consequence was a sudden price gap between London and Lisbon and an outflow of silver to London. Its resulting scarcity in Lisbon induced the change in the local market ratio. This theory is not convincing, however, when one considers that the alleged causal mechanism took five months to make itself felt, much too long a period for two such closely integrated markets. Probably all that happened as regards the change in tariffs was that contraband was converted into official trade and silver exports reappeared in the published statistics, while Lisbon prices were unaffected until the London prices changed. A better explanation seems to be the existence, throughout the period and without any regard for restrictions on trade, because of the ease of smuggling, of efficient arbitrage between the London and Lisbon currency markets, with the decline in the gold-silver ratio which occurred in London in 1853 dragging down the Lisbon ratio to a roughly equal extent. For this to be true, however, coins would have had to be traded in Lisbon at prices which were consistent with those of the London market after taking into account the costs of shipment, in this particular case, of silver to London and of gold to Lisbon. At least until late 1853, this appears to have been the case, as the following data show. As Angela Redish has recently pointed out, ‘bimetallic arbitrage occurs when, under a bimetallic standard, coins of one metal are sufficiently over-valued to make it profitable to melt down coins in the other metal, export the metal, and exchange it for the overvalued metal which is imported and sold to the mint.’32 In the early 1850s, shipping, insurance and commission charges between Lisbon and London for silver added up to 1.4 per cent of the value of the shipment. For gold, they were 0.7 per cent.33 Let us consider, for example, in the late 1840s, before the first monetary crisis, the case of a Lisbon speculator who traded a unit of gold for 16.5 units of silver. After deducting expenses for sending the latter to London and trading it there for sovereigns at the market rate of 15.9:1, he would be left with 1.02 units of gold. Back in Lisbon and after further deducting the cost of the return journey, he would have 1.015 gold units and therefore a profit of around 1.5 per cent on the initial investment.34 Whilst merely an illustration, this example serves to show two things. One is that although small, this was the sort of gain that would prompt the continual movement of precious metals from market to market, since it was realized over a short period and therefore annualized into quite a high figure. The second is that clearly prices in Lisbon were unlikely to diverge significantly from those in London, as otherwise one of these two precious metals would tend to disappear more or less at once from the former market. Moreover, since the Lisbon legal ratio was higher than in most other countries and since London was the dominant market, unless there were some non-market factor coming into play, the Lisbon ratio would always be the higher of the two. The drop in the relative price of gold coin in Lisbon in 1851 is equally well explained with recourse to the arbitrage mechanism. The London gold-silver ratio for this year was 15.46:1 and if we add to this the aggregate margin of 2.2 per cent for transactions costs, as shown above, this would imply a Lisbon ratio of

FIRST TO JOIN THE GOLD STANDARD 171

15.8:1, which is very close to the 96 per cent of par which in fact gold fetched there during the same period. In the course of 1852, the London ratio rose slightly, to 15.6:1, and in Lisbon, during most of this year and until April 1853, gold coin exchanged for silver at around 97 per cent of par, that is again the London price plus approximately 2.2 per cent for costs. Starting in the summer of that year, when the second and final monetary panic manifested itself, bimetallic arbitrage ceases to provide an adequate explanation for the movements in the Lisbon currency market. The London average ratio for 1853, which was 15.33:1 corresponded closely enough to the 94–95 per cent of par at which gold coin was being bought and sold in Lisbon until September. Thereafter, the Lisbon gold market continued to fall, to as low as 90 per cent in November, December and January 1854. Even after a recovery in February–June, to 93 per cent, its behavior still contrasted with the stability which meanwhile had characterized the London market and its level remained much too far removed from the latters price level. At one point, the differential between the two was almost 10 percentage points. To account for this systematic deviation without discarding the arbitrage which all contemporaries believed to have existed, we must suppose that politically induced expectations on the part of the Portuguese speculators now began to influence the currency market too. What seems likely is that the latter had now come to suppose that the government was really going to have to alter the monetary system radically, a belief which was no doubt bolstered by a petition to the Queen from the highly influential Bank of Portugal, in November 1853, asking her respectfully to provide a solution through the government for the ‘terrible’ scarcity of means of payment.35 It was also becoming clear that the authorities would opt for the gold standard, and in so doing would enact something like there was already in Britain. This being the case, the Lisbon market ratio might be expected to move to the proximity of the presumed future Portuguese mint ratio. If this was going to be like the British one of 14.1:1, in Lisbon this would mean something in the region of 14.44:1. Significantly, the market ratio in Lisbon from November 1853 and during early 1854 was around 14.85:1 and rose to only 15.34:1 during that spring.36 If it is true that such expectations existed and were strong, then the renewal of the earlier public discussion and the appearance of several proposals for the reform of the monetary system at this time are hardly surprising. As happened before, two basic solutions were again envisaged. One of them was to keep the situation basically unchanged and, in order to insulate the market from the rest of the world, to slap a stiff export duty on silver whilst hoping that Portugal’s borders would not prove too porous.37 The other was to revalue silver by an amount sufficient to ‘beat the market,’ while making it a subsidiary coin with limited legal tender, in other words, adopting the English system. According to the Count of Taipa, since dealers were selling silver to London at 8,160 reis per mark of silver, this ought to have been enough to raise its mint value to 8,500 reis. If the sovereign were left unchanged at 4,500 reis, this would imply an official gold-

172 PORTUGUESE CURRENCY EXPERIENCE

silver ratio of 15.24:1 and would provide an ample margin of safety relative to the London market ratio of 15.33:1, if account were also taken of shipping and other costs.38 During the second half of 1853, the government stalled and did little except accept legislation that would raise the export duty on silver again. Public opinion, however, continued to press for fundamental reform. The rising tone of the editorials and articles of the Jornal do Comercio, the mouthpiece of Lisbon’s commercial interests, provides a good indication of this. Besides the accounts of daily disturbances and even riots caused by the problems of small change, by November 1853, it was claiming that the monetary crisis was getting steadily worse and that the public clamor for something to be done was getting stronger. In December, it warned that ‘our monetary system is profoundly vicious’ and that the government had to take steps to put an end to ‘the disorganization of our monetary circulation.’ The only solution was a monometallic system which would obviate the need to change the mint gold-silver ratio periodically while constantly finding it out of harmony with the market. It was also the only way to halt the external drain of silver which upset the economy and even society so.39 Meanwhile, reports were coming in from the countryside that commercial farmers were facing similar problems. In the Douro valley, port wine producers were finding it difficult to pay laborers in gold and were having to pay a premium of 160 reis for each sovereign they changed into silver for this purpose.40 In April 1854, the government finally introduced a bill in the Chamber of Deputies to reform the monetary system once and for all. The seriousness of the situation lay in the fact that ‘our circulation of coin is reduced to sovereigns and half sovereigns, silver is missing for the majority of transactions, which are those of smallest value; and when it is exchanged for silver and even copper, gold suffers an exaggerated discount, particularly in the provinces, where the market in coins is more limited than in the capital and other major cities. This cannot continue, as it burdens transactions, holds up the transmission of property, hampers trade, even in its simplest operations, and puts public order at risk.’41 The debate, which centered largely around technical issues, was a lengthy one— it lasted for fifteen days and filled 155 pages of the parliamentary register. On July 29, the government’s proposal became law and Portugal became the first continental nation to adopt the gold standard. THE NEW MONETARY SYSTEM The Portuguese monetary law of 1854 created a system which conformed in every way to the classical model of the gold standard found in the textbooks. It established gold as the single monetary standard, of which only a limited number of types of coin were to be allowed into circulation, their content of precious metal rigorously defined in strict proportionality to their respective face values. The only foreign ones enjoying legal tender were the English sovereign and half sovereign. Gold in any form could be taken to the mint to be exchanged at no

FIRST TO JOIN THE GOLD STANDARD 173

cost for legal coin. Silver, on the other hand, became a subsidiary means of payment, its legal tender limited to 5,000 reis per transaction. To ensure that there was a sufficient amount of silver change and that it did not ‘disappear’ again, the gold-silver ratio was reduced to 14.1:1. This was so much lower than the current market value that it was extremely unlikely that any silver would be hoarded or exported. Since this also implied a substantial seignorage, a specific parliamentary authorization was needed to mint further silver, in order to put temptation out of the government’s reach. Both gold and silver could be freely transacted in any form and there were no limits to their export or import. Finally, all previous legislation was revoked, including the rules which had been legislated in recent years concerning special monetary clauses in contracts. As it turned out, the system worked well on the whole and survived unscathed for almost forty years. Nevertheless, at the time of its inception some of its crucial aspects were by no means an obvious choice. For one thing, while it was evident to most that since the early 1850s bimetallism was no longer suitable for Portugal, opting for gold as the standard was much less obvious. There was still plenty of pro-silver sentiment both at home and abroad, and the examples of the various countries that were heading this way were a constant reminder that there were other possible paths to follow. Possibly more controversial still was the fact that this was not simply a gold standard. It was a sterling-gold standard in which the principal coins would be the sovereign and the half sovereign, and this had several negative implications. The first of these was the offense to feelings of national pride for, as Aragão was later to comment bitterly, ‘one rarely sees the arms or the effigy of the kings of Portugal in the gold money which circulates in our market.’42 The second was the rather more pragmatic fear that the Portuguese economy would be submitted to an even greater degree of dependence on the British one and that every time the latter experienced a downturn, sovereigns would rapidly leave Portuguese shores and cause an even sharper crisis there. A third difficulty arose from the fact that the English mint was known to have far more exacting standards of weight and fineness and to remove promptly all substandard coin that came its way. This could mean there would be a tendency for the sovereigns which were lighter and worn down to migrate to Portugal, whose monetary stock, in time, would come to consist mostly of ‘weak’ coin. Finally, there was the contradiction between, on the one hand, the general trend in Portugal towards decimalization, both in the monetary system since 1835, and in weights and measures since 1852, and, on the other, the adoption as the basic coin of one which could not but fit awkwardly into this, given its facial value of 4,500 reis.43 Why was it that, in spite of all this, Portugal not only went on the gold standard but also modeled itself in this respect so closely on the United Kingdom? The reasons are several and mainly of a practical nature, as might be expected of a nation whose rulers believed ‘that science can solve problems of this nature in general terms but governments are obliged to consider the particular circumstances of the countries they administer.’44

174 PORTUGUESE CURRENCY EXPERIENCE

One essential advantage of the reform was that it ensured an important degree of continuity in Portugal’s monetary affairs, rather than breaking with the past as would have happened had a silver standard been adopted. Gold monometallism was in line with a tradition which, despite official bimetallism, went far back in Portuguese monetary history. All through the eighteenth century, gold had been widely used, as a result of the legendary mining strikes in Brazil which had flooded the country with gold and still left enough to export to England, to cover a considerable external trade deficit. During the nineteenth century there were interludes of either silver or gold predominance, but the monetary legislation of 1846–47, by flooding the country with English coin, firmly shifted it into an enduring gold era. To cap this, in the early 1850s large amounts of gold, again from Brazil, began to arrive as a result of a movement of capital repatriation by some of the richest elements in the Portuguese emigrant community in that country.45 Under the circumstances, the question which had to be asked in 1854 was ‘can we chose between gold and silver? Tell me, what measures would you take in order to replace the gold we have for the silver we have not, without loss either to the State or to private interests? What dangers would not such a substitution entail?’46 And the same might have been said of demonetizing the sovereigns which had come to be in every household and in every hand. A second type of advantage of the new system was its relative inexpensiveness, compared with what a silver standard would have cost. To institute the latter would have required the import of a large amount of silver for coining—perhaps something like 10 per cent of GNP and more than the value of one years exports—and at least part of this from countries with which Portugal had little trade, at a time when a good deal of international commodity transactions were paid for directly in either silver or gold. Moreover, if this were done without an alteration of the exchange rate, Portugal would have actually had to buy silver on the international market at a price that was higher than the mint would then charge for the coins it issued. On the other hand, replacing one currency system with another newly minted in Portugal would have taken time and entailed a costly immobilization of resources, a serious problem for an impoverished treasury which could seldom make even the smallest of ends meet. When the subject was brought up in discussion, it was not forgotten that back in 1851, an equally penurious government had had the greatest difficulty in replacing the non-British coin in circulation and which only amounted to one twentieth of the stock of sovereigns which was thought to exist in Portugal in 1854.47 Several years after the passage of the monetary law, in 1857, Portugal’s finance minister was still cautiously trying to find out how much it would cost the treasury to replace sovereigns with coin minted in Portugal from imported gold bars, obviously a sign that the enthusiasm for any replacement operation was not great. The nature of Portugal’s international economic relations provides another set of reasons in favor of the solution which was followed in 1854. One of the most important of these had to do with the close economic ties between Portugal and

FIRST TO JOIN THE GOLD STANDARD 175

Britain, the only gold standard country at the time. In the 1840s, more than half of Portugal’s trade was with this country. Brazil, another gold-using country, came a distant second, with 20 per cent, which meant that only one fourth of Portugal’s foreign trade was with bimetallic or silver standard countries.48 A similar situation existed with regard to the foreign debt. This was already sizeable and a constant headache for governments which had to acquire the sterling needed to make service payments. Again, this was the currency in which it was denominated, even those parts of it which were raised in Paris during the 1830s. As the Finance Committee of the Chamber of Deputies put it, ‘the Government has always use for [the sovereign] in the large transactions and remittances it makes every year to London, to pay our dividends.’49 In contrast, there seem to be no grounds for supposing that one of the motives for adopting the gold standard was the desire for easier access to the international capital market.50 This is not to say that this may not have been helpful at a later date. By this stage, many countries had joined the gold standard, international capital markets had become more sophisticated and discriminating and more importance attached to monetary standards as a signal to the market. Nor does it imply that at mid-century, foreign capital was not desired. On the contrary, all through the 1840s and 1850s, Portugal entertained keen hopes of attracting external financial resources, particularly to invest in transportation and to consolidate its parlous financial situation. But there is no evidence that the question of the monetary regime was even deemed by contemporaries to be relevant to its lack of success in this field prior to 1856, or to its success subsequently. By all available accounts, foreign loans to the state were conditioned by two factors alone at this time. One was the country’s national credibility, in other words, its political stability, the government’s ability to raise revenue, and its resolve to keep to its commitments, especially to its foreign creditors. In 1852, the Portuguese ambassador in London reported being told by the London financiers he had approached for loans: ‘In order that our funds be employed in Portugal it is necessary first that you assure us that contracts will be religiously kept and that public tranquillity will not be disturbed.’ And when he pressured Barings and Goldsmith to the same end, their first question was ‘What will be the result of your forthcoming elections?’51 The other factor was Portugal’s treatment of foreign bondholders. A dismal record with regard to the servicing of the debt and a forced conversion in 1852 were both the subject of the most disagreeable comment in London financial circles. In fact, as a result of the pressure of the Committee of Foreign Bondholders, Portugal was excluded from the London Stock Exchange and it was not until 1856, when the government secured an agreement with the bondholders, that this ban was lifted and the first new loan could be raised. Symptomatically, when finance minister Fontes Pereira de Melo announced in Parliament the successful conclusion of these negotiations, he made absolutely no mention of the gold standard as a contributing factor.52

176 PORTUGUESE CURRENCY EXPERIENCE

The difficulty in attracting foreign finance for railway construction was also largely connected with Portugal’s general lack of creditworthiness, as described above. The wariness of lenders in this field stemmed from their negative assessment of the potential offered by the Portuguese economy, which was poor and backward. When some capital was finally lured for this purpose in the late 1850s and early 1860s, it was largely thanks to hefty government subsidies, the prospect of linking up with Spanish networks and the hope that railway ventures would open up fresh possibilities of business.53 Once again, the influence of the monetary system does not seem to have existed, something which is further brought out by the fact that Spain, with a bimetallic and inconvertible system, was able to attract far more capital during the 1850s. CONCLUSION Much has been written about how the gold standard worked, as well as about its demise. The political and economic processes which led different countries to joining this first and very successful international monetary system are less known. This paper has attempted to provide an account of one such process in a country which is not important in terms of its weight internationally but which is worthy of attention because of its precociousness in this respect. Portugal was already practically on the gold standard when it adopted this regime officially in 1854. That it found itself in this position was very much a matter of historical accident. As a result of a quick and not very well considered set of decisions which had to be taken in a moment of great unrest and monetary penury, it became a haven for English sovereigns in the late 1840s. Thereafter, the vagaries of the world gold and silver markets reinforced this situation and drove the country to abandon bimetallism and find another system. In the final analysis, despite the abundance of scientific and intellectual trappings, it was as a pragmatic measure that the new monetary regime gained approval. This was not the choice of an ideal model of monetary regime but rather of one which, for Portugal, would be both feasible and cheap. The fact was that the sovereign, on which the new system was based, was widely accepted, was well manufactured and cost the Portuguese state nothing to mint. More significant, when English monetary arrangements were held up as an example for Portugal to follow, it was not because Britain was rich and powerful and should therefore be imitated. Rather, it was because the British system had worked successfully since 1821, providing a convenient monetary standard for large transactions—gold—while ensuring that silver token coin was available in the right quantity as change for small transactions.54 Lastly, rather than wonder why Portugal joined the gold standard so early, perhaps the most important point to draw from this paper is why other de jure bimetallic countries, when faced with the same international context, did not do the same as Portugal. Apart from the degree of economic dependence on Britain, the explanation may lie in the differences between these countries with respect to

FIRST TO JOIN THE GOLD STANDARD 177

their mint gold-silver ratios. Portugal’s was so much higher that it suffered a strong and premature external drain of silver and had no choice but to reform its monetary system early. Other countries could hold out longer because they may have suffered a relatively weaker silver drain and then, when they adopted the gold standard, the reasons for doing so were already quite otherwise. NOTES * Instituto de Ciencias Sociais. The author is grateful to the Bank of Portugal for the financial support which made this research possible. 1 See P.L.Cottrell (1992), ‘Silver, Gold and The International Monetary Order 1851– 1898’, in S.N.Broadberry and N.F.R.Crafts (eds), Britain in the World Economy, 1870–1939, Cambridge, Cambridge University Press, p.223. 2 Giulio M.Gallarotti (1993), ‘The Scramble for Gold: Monetary Regimes Transformation in the 1870s’, in Michael D.Bordo and Forrest Capie (eds), Monetary Regimes in Transition, Cambridge, Cambridge University Press; and Jeffry A.Frieden (1993), ‘The Dynamics of International Monetary Systems: International and Domestic Factors in the Rise, Reign, and Demise of the Classical Gold Standard’, in Jack Snyder and Robert Jervis (eds) Coping with Complexity in the International System, Boulder, Westview Press. 3 There is ample evidence for this in Rondo Cameron (1961), France and the Economic Development of Europe, 1800–1914, Princeton, Princeton University Press. 4 The best account of Portuguese monetary history for this period is still Augusto Carlos Teixeira de Aragão, Descripção Geral e Histórica das Moedas Cunhadas em Nome dos Reis, Regentes e Governadores de Portugal, Lisbon, Imprensa Nacional, 3 vols. 5 See Bank of Portugal Archives: Banco de Lisboa, Livro da Caixa de Reserva and Livro Mestre. The Bank of Lisbon had two separate cash vaults: the Caixa de Expediente and the Caixa de Reserva. We have a detailed breakdown only for the latter but it is unlikely that the former should have been appreciably different. Nearly all the gold coin was Portuguese, very few sovereigns being held at this time, in contrast to what was to happen later. 6 In the 1830s, Belgium, for example, another small, bimetallic country granted legal tender to Austrian, Dutch and French coin and added sovereigns to this list in 1848. See P.Kauch (1950), La Banque Nationale de Belgique. I. 1850–1918, Brussels, Banque Nationale de Belgique, pp.13–14. 7 According to Aragão, Descripção, p.179, in 1835, the productive capacity of the mint was 7.5 million reis a day. If it were to have worked at full capacity 250 days a year it would still have produced less than one tenth of the country’s monetary supply. A recent account of technical aspects of the mint’s history is given by Carlos Bastien (1991), ‘Para a Historia da Casa da Moeda de Lisboa—Aspectos Técnicos e Organizativos da Produção da Moeda Metálica’, Estudos de Economia, XII, pp.43–78. 8 David Justino (1981), ‘Conjuntura Económica e Maria da Fonte. Algumas Notas’, Bracara Augusta, XXXV, pp.5–12; Wallace E.Huffman and James R.Lothian

178 PORTUGUESE CURRENCY EXPERIENCE

9 10

11 12 13

14

15

16

17

18

19

20

(1984), ‘The Gold Standard and the Transmission of Business Cycles, 1833– 1932’, in Michael D.Bordo and Anna J.Schwarz (eds) A Retrospective on the Classical Gold Standard, 1821–1931. Chicago, University of Chicago Press, p.468. The Portuguese monetary unit was the real (plural, reis). There was also a unit of account, the conto, which was equal to one million reis. In the succeeding period Portuguese exports of oranges and wine were often traded directly for sovereigns while imports were purchased with silver coin, which therefore gradually disappeared from circulation. See Rebello da Silva speech, in Diario da Camara dos Deputados, January 21, 1851, p.128. Speech by Julio Pimentel, in Diario da Camara dos Deputados, May 4, 1854, p.75. Ibid., p.74. Rita Sousa (1991) in ‘Money Supply in Portugal 1834–1891’, Estudos de Economia vol. XII, p.28, has estimated M0 in 1847 at 36 million reis. Daily quotations for the discount on Bank of Lisbon notes are given in Livro do Agio Notas do Banco de Lisboa, Bank of Portugal Archives. As soon as Parliament reopened after the civil war, in 1848, bills to clarify the confusion which reigned in the country’s means of payment were tabled and over the next two years a considerable amount of debating time was devoted to the question of how to put an end to the market discount of the Bank of Lisbon’s notes. See the first two proposals presented in February 1848 to the Portuguese House of Peers, in Diario do Governo, 1848, pp.249 and 429. Foreign coin circulated in the United States until 1860 and in the early 1800s was said to represent 19/20ths of the stock of minted silver, according to David A.Martin (1977), ‘The Changing Role of Foreign Money in the United States’, Journal of Economic History, XXXVII, p.1017. See Marquês Camilo Pallavicino Grimaldi (1855), A Legislação Monetária em Portugal, Lisbon, Typ. do Progresso; João Mouzinho de Albuquerque (1862), Memória sobre a Moeda Portuguesa e sua Origem, seus Usos e Abusos: Oferecida às Classes Menos Versadas na Sciencia do Credito, (Lisbon, Typ. Elvense; Carlos Morato Roma (1861), A Questão da Moeda, Lisboa, Typ. da Academia. See also various speeches during the parliamentary debate of January and February 1851, in Diario da Camara dos Deputados. The Bank of Portugal, which had also been given the right of issue in 1846, had a small amount of convertible notes in circulation too. These could be exchanged either for gold or silver and were backed by high reserves. See law proposals 11-E and 11-F dated 13 March 1848 by the Minister of Finance to print government notes, in appendix to Diario da Camara dos Deputados, March 1848. ‘Gold Silver Ratio in London’ Parliamentary Papers, 1881, LXXV, pp.228–9. This series differs from that in David A.Martin (1977), ‘The Impact of MidNineteenth Century Gold Depreciation upon Western Monetary Standards’, Journal of European Economic History, vol.6, p.648. Silver always remained in circulation to some extent since there was a need for smaller denominations, but it was transacted at a premium above its par value. See Stefan E.Oppers (1995), ‘Recent Developments in Bimetallic Theory’, in Jaime Reis (ed.) International Monetary Systems: Historical Aspects, London, Macmillan.

FIRST TO JOIN THE GOLD STANDARD 179

21 Speech by Antonio José d’Avila, Diario da Camara dos Deputados, January 1851, p.83. 22 Diario do Governo, 1851, p.59. 23 The demonetization of all the different North and South American and continental gold coin did not cause any loss to its holders since the Mint was ordered to exchange them at par for legal money, the losses being sustained by the authorities. 24 See the remarks by the Minister of Finance in Parliament, in Diario da Camara dos Deputados, January 1851, p.78. A beneficial spin-off of this, however, was that it eliminated much of the substandard gold coin then in circulation 25 According to a deputy who was also a director of the Bank of Portugal, the profit on the export of silver was 2 per cent in 8 days and the trade could therefore not be stopped by laws alone. See speech by Xavier da Silva, Diario da Camara dos Deputados, February 1852, p.136. 26 Speech by Rebello da Silva, Diario da Camara dos Deputados, February 1851, p. 130. 27 Speech, ibid., January 1851, p.117. 28 Speech, ibid., January 1851, p. 116. 29 And they were right. World gold output in the 1860s was only slightly lower than the then historically high level of the 1850s. See Barry J.Eichengreen and Ian W.McLean (1994), The Supply of Gold under the Pre-1914 Gold Standard’, Economic History Review, XLVII, 2, pp. 288–309. 30 Speech, Diario da Camara dos Deputados, May 1853, p.146. 31 Speech by Avila, in Diario da Camara dos Deputados, May 1853, p.144. 32 A.Reddish (1992) ‘The Evolution of the Classical Gold Standard: The Case of France’, University of British Columbia, Department of Economics Discussion Paper No.92–22, p.10. 33 These charges are for transactions which actually took place and are taken from the account books of the Bank of Portugal, which dealt both on its own behalf and that of the government. For a shipment of silver worth £6, 173–13–8, packing and carriage to the port were £6–15–0, freight was £31–1-0, insurance was £11–12–11 and commissions were £38–18–4, making a total of £88–6–15. See Archive of the Bank of Portugal, Diario Auxiliar, 1856, p.192. For a shipment of £20,000 in sovereigns, packing was £5–15–0, freight was £100–2–0 and insurance was £37– 12–6, making a total of £143–9–6. For no apparent reason, there were no commission charges in this instance. Ibid., p.61. 34 I have not considered here the additional cost represented by the difference in Lisbon between the mint price and the mint equivalent for gold, because the sovereigns which our speculator brought back entered directly into circulation, being legal tender, and did not have to be taken to the mint for conversion into Portuguese currency. With the silver in London, I assume it was being sold on the market for melting into bar. 35 Archive of Bank of Portugal, Registo de Papeis Oficiais, Livro N.5, p.192. 36 It is important to consider that Fontes Pereira de Melo, who was the finance minister in 1853 and who was also therefore probably going to be the architect of any monetary reform, had expressed a considerable sympathy for the English system as far back as the 1851 monetary debate, when he was in opposition. The same is true of the Marquis of Lavradio, an 1851 oppositionist who in 1854 occupied the politically important post of ambassador to London.

180 PORTUGUESE CURRENCY EXPERIENCE

37 This was the proposal which Santos Monteiro tabled in the Chamber of Deputies. See Diario da Camara dos Deputados, April 1853, p.72. 38 For Taipa’s proposal, see ‘Camara dos Pares’, session of 29 July 1853, in Diario do Governo 1853, p. 1136. See also the same idea in the proposal of Lourenço Cabral in Diario da Camara dos Deputados, May 1853, pp.551 and 596. A third solution tabled by Carlos Bento was to authorize the Bank of Portugal to issue convertible notes in small denominations and to mint small gold coins, all in order to ease the difficulty with the lack of small change in silver. See Diario da Camara dos Deputados, June 1853, p.265. The fact that it did not find favor and was not even debated should come as no surprise. In 1846–47, the Bank of Portugal had issued small inconvertible notes for the same purpose and they had not left pleasant memories. Nevertheless, Carlos Bento returned to the fray with his project in 1854. See ibid., March 1854, p.389. 39 Jornal do Comercio, 15 December and 27 December 1853. 40 Ibid., 7 November 1853. The lack of any expression of opinion on the issue by the Lisbon Commercial Association can perhaps be explained by its undergoing an internal crisis at this time. On the other hand, it is worth noting that the Oporto Industrial Association seemed interested only in debating the tariff question and never once mentioned the monetary question in its magazine, the Jornal da Associação Industrial Portuense. 41 From the preamble of the government’s proposal, in Diario da Camara dos Deputados, May 1854, p.15. The concern over public order was paramount for a country which had only recently emerged from several decades of internecine strife. 42 Augusto Carlos Teixeira de Aragão, Descripção Geral, p.233. 43 An example of this awkwardness was the need felt by the Bank of Portugal to replace its 10 and 20,000 reis notes with others denominated in 9,000 and 18,000 reis in order to make it easier to pay them out in gold, in multiples of £1. 44 From the preamble to the 1854 monetary law in Diario da Camara dos Deputados, May 1854, p.13. 45 Gervase Clarence-Smith (1985), The Third Portuguese Empire, Manchester, Manchester University Press, pp.50–6l. This is confirmed by the Marquis of Lavradio’s speech in the Chamber of Peers, 29 January 1851, in Diario do Governo 1851, p.127. Capital repatriation was associated with the expulsion from Rio de Janeiro of Portuguese slave traders, as a gesture of appeasement of the United Kingdom by the Brazilian government. 46 Speech by Julio Pimentel, Diario da Camara dos Deputados, May 1854, p.74. 47 The stock of sovereigns was thought to be around £5 million. See the speeches by Avila and Alves Martins in Diario da Camara dos Deputados, May 1853, respectively pp.147 and 148. 48 For a breakdown of Portugal’s foreign trade, see Maria de Fátima Bonifácio (1991), Seis Estudos sobre o Liberalismo Português, Lisbon, Estampa. Carl Ludwig Holtfrerich points out that Hamburg and Bremen were gold standard enclaves in a German sea of silver standard states because of their strong foreign trade orientation and connections with England: (1989). ‘The Monetary Unification Process in 19th Century Germany: Relevance and Lessons for Europe Today’, in Marcello de Cecco and Alberto Giovanninni (eds), A European Central Bank?, Cambridge, Cambridge University Press, p.223.

FIRST TO JOIN THE GOLD STANDARD 181

49 Speech by José Izidoro Guedes, Diario da Camara dos Deputados, January 1851, p.120. 50 Late nineteenth century Russia is the classic example of a country where the gold standard objective was closely connected with the desire to attract foreign capital. See Olga Crisp (1976), Studies in the Russian Economy before 1914, London, Macmillan, p.97, and James Foreman-Peck (1983), A History of the World Economy. International Economic Relations since 1850, Brighton, Wheatsheaf Books, p.181. See also Frieden, op. cit. n. 2 above, p.147, who argues that ‘the incentives to go to a monometallic gold standard increased especially after international financial flows out of London became very large in the 1850s and 1860s.’ 51 D.Francisco de Almeida Portugal (1937), Memòrias do Conde do Lavradio, Lisbon, Imprensa Nacional, part II, Vol. IV, pp.47 and 113. Lavradio attributed his failure to raise £100,000 for roads to five causes: lack of guarantees that the money would really be used for roads; lack of guarantees that the revenue assigned to the loan would be kept for it; the unilateral reduction in interest payments in 1851; the irregular payment of interest on the foreign debt; and the lack of assurances that further such reductions would not take place. Ibid., p.120. 52 Diario da Camara dos Deputados, February 1856, pp.21–24. 53 Two unpublished doctoral theses on this subject are helpful: Antonio Lopes Vieira (1983), ‘The Role of Britain and France in the Finance of Portuguese Railways, 1850–1890. A Comparative Study in Speculation, Corruption and Inefficiency’, Leicester University; and Magda Pinheiro de Sousa (1986), ‘Chemins de Fer, Structure Financière de l’Etat et Dépendance Extérieure au Portugal (1850–1890),’ University of Paris-I. 54 The Portuguese law of 1854 was such a close copy of the English one that in the initial proposal, the maximum legal tender for silver was set at 10,000 reis, approximately the equivalent of the £2 laid down for England. In the course of the debate, it was recognized that some adaptation to local conditions might be wise, and the limit was reduced to 5,000 reis.

182

7 LAST TO JOIN THE GOLD STANDARD, 1931* Fernando Teixeira dos Santos

INTRODUCTION After World War I, and following the recommendations of the Brussels (1920) and Genoa (1922) Conferences, several European countries returned to the gold standard. After the inflationary experiences during the war and post-war years, it was expected that the gold standard would provide for long-run price stability, exchange rate stability and promote conditions favorable to the development of international trade and finance. Sweden was the first European country to return to a de facto gold standard, in 1922, which became de jure in 1924, at the pre-war parity. Among the belligerent countries, Germany was the first one to return to gold convertibility, in 1924, Great Britain returned to the gold standard in 1925 at the pre-war parity, and France returned in 1928. By a decree of June 1931, Portugal decided to adopt the gold standard as of July of that year, after a forty-year suspension of the convertibility of her currency. While the major Western European countries were under some kind of gold standard until World War I, Portugal had an inconvertible regime between 1891 and 1931. Portugal had suspended convertibility in 1891 in the aftermath of a financial crisis.1 The return to convertibility however, had represented an important goal for Portugal’s authorities before the war. The importance of the relations between Portugal and Great Britain, and the international prestige of the gold standard in the promotion of long-run price stability and exchange rate stability, explain the relevance of such goals to the Portuguese authorities. Following a period of exchange rate instability, in 1906–07 the Portuguese currency stabilized at a level close to its former parity (£1=4.50 escudos). Government budget surpluses during the 1912/13 and 1913/14 fiscal years (the first Afonso Costa administration) made conditions favorable for a return to the gold standard.2 However, the outbreak of war and Portuguese participation in the conflict did not allow such an outcome. The war period, and resulting financing needs, generated significant public budget deficits, partially alleviated by loans from the Bank of England.3 After the war, public budget deficits increased sharply, financed through bank note issues from the Bank of Portugal. Up to the mid-1920s, there was high inflation and

184 LAST TO JOIN THE GOLD STANDARD

severe exchange rate depreciation. Political instability did not allow the adoption of measures to promote sound fiscal policies and price and exchange rate stability. Despite improvements in 1924 and 1925, it was only possible after 1928 to promote the stabilizing policies required for Portugal’s return to the gold standard, which occurred on July 1, 1931, with the establishment of the convertibility of the escudo. On September 21, 1931, due to substantial withdrawals of gold from the Bank of England, the British authorities were forced to suspend the convertibility of sterling. This event marked the beginning of the end of the post-war gold standard experience. As a consequence, Portugal had the most short-lived gold standard experience of any country—a mere eighty-two days. Such a period was too brief to produce sufficiently visible effects to assess the performance of the new regime in terms of price stability, exchange rate behavior, and real economic growth and stability. This period is relevant, not so much because of its brief duration, but because it illustrates the commitment of Portugal’s new minister of finance, Salazar, to the conduct of sound monetary policies, helping to explain his emergence as the most influential Portuguese politician of the period. This paper is organized as follows. First there is an analysis of the exchange rate behavior and its determinants during the war and the post-war period up to 1931. There is then a description of the events preceding the return to the gold standard as well as the behavior of major variables during the 82-day period. The reasons leading the authorities to peg to sterling in September 1931, when England abandoned the gold standard are then examined. The paper concludes by stressing the economic and political significance and importance of the Portuguese return to convertibility, despite its ephemeral life. PORTUGAL’S POST-WAR INFLATIONARY EXPERIENCE: MONETIZATION OF PUBLIC DEFICITS AND EXCHANGE RATE DEPRECIATION During the years following World War I until the mid-1920s Portugal had a record of high budget deficits, strong depreciation of its exchange rate, and high inflation. The rapid increase in the price level was a result of both the monetization of government deficits and the expense of imported goods following depreciation of the escudo. Disruptions in supply conditions reflecting social agitation, and consequent shortages, together with speculation in the goods market, significantly contributed to push prices upwards.4 Figure 7.1 illustrates the evolution of the monthly escudo/sterling exchange rate from 1910 to June 1931, as well as during the gold standard period—July to September 1931—and the following months up to the end of 1932. After a gradual, but steady, depreciation during the war period, the escudo suffered a strong depreciation between 1919 and 1924. Several factors may have contributed to the acceleration of this depreciation. First, the end of the war marked the end of financial support from the United Kingdom. Secondly, the sharp

PORTUGUESE CURRENCY EXPERIENCE 185

Figure 7.1 Escudo/sterling monthly exchange rate, 1910–1932 Source: Mata (1987)

Figure 7.2 Government budget deficit as a percentage of GDP Source: Neves (1993)

increase in imports aggravated the trade deficit. Thirdly, political turmoil and social agitation led to substantial capital outflows. Finally, there was speculation against the escudo in the expectation of a significant appreciation of the currencies of the belligerent countries, especially the German mark.5 Figure 7.2 shows the evolution of the annual budget deficit from 1910 to 1932. As mentioned before, in order to finance its war expenditures, in 1916 Portugal negotiated with the Bank of England the grant of two major loans. Besides these loans, the Portuguese authorities6 drew up a new charter for the Bank of Portugal in April 1918. The earlier charter had imposed a maximum limit to note issues by the Bank, and the obligation to back the notes issued with a minimum metallic reserve (20 per cent for the notes backed by gold and 100 per cent for the notes backed by silver). Issue beyond the defined limit (120

186 LAST TO JOIN THE GOLD STANDARD

million escudos was the limit defined in August 1914, which rose to 200 million in December 1916) was only allowed when fully backed by metal reserves. Under the new charter, however, the Bank could issue money to finance government expenditures and was not subject to such a constraint.7 There was a major change in the conduct of monetary policy in the post-war period. Although the war had ended in November 1918, the government’s ability to monetize its deficits was not revised until the 1931 monetary reform. The political instability during the years following the war8 did not favor the adoption of restrictive fiscal policies, which were unpopular and politically costly, so governments adopted the easiest way to finance deficits—by issuing money. The record of substantial government deficits and its monetary financing, pushing prices upwards, imposed a high inflationary tax on the private sector.9 The results of such a policy regime are clearly illustrated in the following figures. Figure 7.3 depicts the evolution of the government’s debt to the Bank of Portugal and of the amount of notes issued, and Figure 7.4 reports the behavior of the price level. Granger causality tests10 suggest two major conclusions with respect to price level behavior. First, money issue by the Bank of Portugal to finance government deficits led to subsequent changes in the price level. Second, the movement of the price level and of the exchange rate fed on each other. The detection of causality running from the exchange rate to the price level may be explained by expectations of future inflation and depreciation, leading to speculative behavior against the escudo. The following table reports the average annual rates of change of the relevant variables during the period 1918–24: During this high inflation period, the Portuguese price level increased faster than the UK price level, as illustrated in Figure 7.4. This is reflected in the behavior of the escudo/sterling real exchange rate (Figure 7.5). Figure 7.6 shows that, during the period running from the end of World War I to the declaration of escudo convertibility in July 1931, the behavior of the nominal exchange rate closely followed the pattern of the ratio of the Portuguese to the British price level. However, analysis of the real exchange rate shows that in 1917 and 1918 the nominal depreciation was not enough to match the inflation differential between the two countries reflected in a real appreciation of the escudo. The real exchange rate reached its pre-war average level, represented by the solid horizontal line in early 1919. Between the second half of 1919 and 1922, despite some attempts by the authorities to impede a strong depreciation of the escudo,11 the rise in the nominal exchange rate more than offset the inflation differential,

PORTUGUESE CURRENCY EXPERIENCE 187

Figure 7.3 Government debt to the Bank of Portugal and note issue Source: Direcção Geral de Estatistica, Boletim Mensal and Annual Reports of the Bank of Portugal

Figure 7.4 Price level: cost of living in Portugal and in the UK Source: Portuguese cost of living index: Amorim (1946) and Boletim da Previdência Social; UK cost of living index: Capie and Webber (1985)

leading to a sharp real depreciation. In 1922, the average real exchange rate was 114 per cent higher than its 1919 average. Facing such a nominal and real depreciation of the escudo, important measures were implemented in order to reduce budget imbalances, to promote confidence, and to restrain speculation. In September 1922, Parliament voted a law increasing existing taxes and creating new ones such as a sales tax, an additional 1 per cent tax on imports of consumption goods, and taxation of personal income. These measures, followed by the budgetary discipline imposed in 1924 by Alvaro de Castro,12 significantly reducing public expenditures, together with a further increase in some indirect taxes, were able to reduce the size of the government deficit. The deficit, as a percent of GDP, decreased from 8.14 per

188 LAST TO JOIN THE GOLD STANDARD

Figure 7.5 Escudo/sterling monthly real exchange rate

Figure 7.6 Nominal exchange rate and relative prices: Portugal vs. UK (log scale)

cent in 1922 to 6.53 per cent in 1923, 3.25 per cent in 1924 and 3 per cent in 1925, as shown in Figure 7.2. In July 1922 exporters were obliged to deposit in the Bank of Portugal an amount of foreign currency corresponding to 50 per cent of the value of their exports. This percentage was raised to 75 per cent in 1924. Also in 1924, the government sold the silver coins removed from circulation in 1917 (£2 million). As a result of these measures there was a substantial increase in the gold and foreign currency reserves which enabled the Portuguese authorities to intervene in the exchange market in order to stabilize exchange rate movements. Furthermore, the government reduced its foreign exchange needs, imposing the payment in escudos of the charges related to the outstanding external debt of bonds denominated in foreign currency and held by Portuguese citizens,13 i.e. abrogating any gold clauses. At the beginning, the market interpreted these latter measures as a symptom of weakness of the escudo and reacted with a depreciation of the currency. However, following the reduction of the government deficit, intervention in the

PORTUGUESE CURRENCY EXPERIENCE 189

exchange market, and important regulations imposing a more strict discipline in exchange trading operations in the second half of 1924, the movement of the nominal and real rates began to be reversed.14 The nominal exchange rate peaked at 154.75 escudos per pound in July 1924 and started declining sharply. By the end of 1924, the escudo/sterling rate had dropped to 100.5 and continued to decline in such a way that the authorities had to intervene in order to avoid too strong and fast an appreciation. In the beginning of 1926, the exchange rate was stabilized at 94.75 escudos per pound. The real exchange rate accompanied the movement of the nominal rate, appreciating 36 per cent between 1922 and 1926. In the following periods both the nominal and the real exchange rates had a more stable behavior compared to the previous years.15 Although the policies conducted by the Álvaro de Castro and Daniel Rodrigues ministries reversed the depreciation of the escudo and induced a significant appreciation, they were not able, as expected, to attract capital from abroad. THE 1928–31 STABILIZATION OF THE ESCUDO AND THE RETURN TO CONVERTIBILITY: THE 82 DAYS OF ESCUDO CONVERTIBILITY The relative stability of the exchange rate beginning in the second half of 1924 suggests that Portugal could have declared convertibility of the escudo at that time, following the example of other countries which, following the recommendation of the Genoa Conference (1922), had returned to the gold exchange standard. However, several difficulties prevented such a decision. First of all, political instability, which led to the military coup of May 1926, and instability within the victorious forces of the coup prevented a return to the gold standard.16 Secondly, despite the reduction of the government deficit which occurred between 1922 and 1925, its level was still high and, had increased in 1926 to 4.3 per cent of GDP. Finally, as Figures 7.7 and 7.8 illustrate, Portugal had accumulated a significant public debt, representing 74 per cent of GDP in 1926. Forty per cent of this debt was floating debt (15 per cent internal and 25 per cent external). The high level of the debt, particularly of its floating component, represented a threat to the stability of the exchange rate. In 1926 and 1927, under the ministry of Sinel de Cordes, the budget deficit deteriorated and the exchange rate depreciated in early 1928 from 94.75 to 98. 75. The financial situation of the government became so delicate that a £12 million loan was requested from Baring Brothers in January 1927.17 The conditions imposed by the lender were so severe that the government had to abandon its request. Again, in late 1927 the government requested a new loan under the auspices of the League of Nations, but again the demands imposed by this international organization were unacceptable.18 These difficulties in international capital markets forced the authorities to reduce deficits through the adoption of expenditure-reducing measures and an increase of government revenues. It was, however, under the ministry of Salazar, who again became

190 LAST TO JOIN THE GOLD STANDARD

Figure 7.7 Government debt as a percentage of GDP: total and floating Source: Mata and Valério (1994)

Figure 7.8 Government domestic and external floating debt as a percentage of total debt Source: Mata and Valério (1994)

minister of finance in April 1928, that the budget was balanced and the floating debt substantially reduced. In face of the collapse of the gradualist strategy of Sinel de Cordes, Salazar had to opt for a strategy imposing higher short-run costs, which would require a stronger political will. Salazar made stabilization of the value of the currency his first priority and defined balancing of the government budget and reduction of the public debt as the means to achieve that goal.19 A fiscal reform was implemented in 1928 and 1929. This reform, adopting the distinction between ordinary and extraordinary expenditures, established the principle of a balanced ordinary budget. Extraordinary expenditures were subject to a careful classification, including only major investments and development expenditures as well as expenditures implied by exceptional defense and national security situations. Most important was the discipline imposed on the spending practices of the ministries and the

PORTUGUESE CURRENCY EXPERIENCE 191

major supervisory functions assigned to the Ministry of Finance. With respect to taxation, the measures adopted aimed to increase tax revenues mostly based on an increase of the tax base but avoiding any increase in tax rates. The 1928–29 budget followed the new principles and, after 14 years of consecutive deficits, a positive balance was recorded. Budget surpluses were obtained in the following years. The positive budgetary results were accompanied by a reduction in the size of the public debt, namely of its floating component. In 1929 the external floating debt was liquidated and the internal one was partially paid or converted to medium or long-term debt.20 The total public debt was reduced from 74 per cent of GDP in 1926 to 61 per cent in 1931. After a three-year period of balanced budgets, the exchange rate remained relatively stable at £1=108.25 escudos, and foreign reserves were made available to meet the needs required for external payments and stability of the currency. Under these favorable conditions, the Portuguese authorities, following suggestions made by the Genoa Conference of 1922, decided to return to the gold standard, in its gold-exchange version, by defining a new gold-escudo parity, based on its current stabilized market level. On June 9, 1931 a decree was published defining the new monetary regime to be effective as of July 1. The new monetary unit—the gold escudo—was defined with a gold weight of 0.0739 grams with a fineness of 900/1000, which corresponded to an escudo-sterling parity of 110.00.21 This parity defined a nominal exchange rate slightly higher than its market value. This rate was justified on the basis of the ease of making conversions between the Portuguese escudo and the currency of its major trade and financial partner. However, the purpose of setting a higher rate was undoubtedly to improve competitiveness and the trade balance in the face of an international recession.22 In the terms of this law and according to its new charter, the Bank of Portugal was required to maintain the escudo-gold parity. To do so, the Bank had to hold a reserve of gold (in either coins or bullion), foreign currencies or other shortterm foreign assets convertible in gold. This reserve would have to represent at least 30 per cent of the monetary liabilities of the Bank (notes and deposits). An important innovation with respect to the functioning of the Bank of Portugal was the creation of a Stabilization Commission, composed of the governor and two vice-governors appointed by the government and three administrators elected by the shareholders of the Bank. The function of this Commission was to ‘conduct the Bank’s policy with respect to monetary stabilization, namely through the regulation of the price and global volume of the credit in any of its forms.’23 This reform did not transform the Bank of Portugal into a full fledged central bank, since its supervisory functions over banking activity were still limited. However, it was recognized that the Bank should be committed progressively to the function of regulating the credit activity of the commercial banking system. The creation of the Commission was intended to be a step towards this objective,

192 LAST TO JOIN THE GOLD STANDARD

Figure 7.9 Escudo/sterling daily exchange rate May–October 1931 Source: O Comércio do Porto, daily issues

balancing the commercial function of the Bank of Portugal, a privately owned issuing bank, with the functions of a central bank. The inflationary effects of the monetization of public deficits were still present and so severe limitations were put on the issue of bank notes to meet the financing needs of the government. First, besides the need to back money creation with the reserve, a second constraint was established through the imposition of a limit of 2,200 million escudos to the notes issued. Only when fully backed with gold coins or bullion could the Bank of Portugal exceed this limit. Secondly, the government debt to the Bank was limited to a maximum of 1, 100 million escudos, which was about its current level after the deduction of the gains from the revaluation of the gold stock. Also, the floating debt in the form of treasury bills should be extinguished, which happened in 1935. Finally, the government would hold a 100 million escudos current account in the Bank of Portugal. Withdrawals from this account had to be paid for until the end of the fiscal year, using the receipts collected. Indeed, the imposed budgetary discipline, together with these limitations to money financing, reduced the amount of the public debt outstanding (Figures 7.7 and 7.8) and stabilized the issuing activity of the Bank (Figure 7.3). A major aspect of the new regime was the resumption of convertibility. Bank of Portugal notes could be exchanged for gold coins or foreign exchange convertible into gold on the basis of a rate within the gold points of the defined parity. The choice of conversion into either gold or foreign exchange was at the option of the Bank. This legal disposition clearly defined a gold-exchange standard regime for Portugal. For its full operation, some restrictions on exchange transactions, still in practice, would have to be removed at a date to be agreed between the government and the Bank. Finally, the obligation to exchange the outstanding notes for gold coins would start at a date also to be decided by the government and the Bank of Portugal.24

PORTUGUESE CURRENCY EXPERIENCE 193

Figure 7.10 Escudo/dollar daily exchange rate May–October 1931 Source: O Comércio do Porto, daily issues

With the exception of the adjustment of nominal exchange rates (as Figures 7.9 and 7.10 illustrate), there were no significant reactions to the establishment of the new monetary regime. Indeed, the new legislation formalized an already de facto situation achieved by the stabilization over the previous three years. The return to the gold standard was indeed expected. In an interview with the Financial Times in 1930, Salazar had expressed his intention to implement the legal stabilization of the escudo. In the beginning of May 1931, the Ministry of Finance prepared an official press statement announcing the preparation of the decrees defining the new monetary regime, which would start on July 1. At the end of May, even before the publication of the decree in the official newspaper, the press publicized the document as well as issuing several editorials stressing its importance. Figures 7.9–7.13 illustrate the behavior of the major economic variables during the eighty-two day period of convertibility. Figures 7.9 and 7.10 report the daily exchange rate of sterling and of the U.S. dollar. The rise of the escudo/ sterling rate to its new parity—1.75 per cent above its current market value for the cheque rate and 8.3 per cent higher for the rate on sight operations—is apparent. At the time, the escudo registered a slight 1.7 per cent depreciation relative to the dollar. It is worth noting that this adjustment of the exchange rates occurred in the first days of June, right after the announcement of the new legislation. After this adjustment, the sterling rate remained, as expected, at its fixed parity and the dollar rate remained stable until September 1931. Figure 7.11 illustrates the evolution of the weekly figures of the government debt to the Bank of Portugal and outstanding bank notes. Of importance is the decline of the debt from around 1,630 million escudos at the end of June to 1,059 million in July following the monetary reform. The stock of outstanding notes remained steady over the year,25 just as it had in the preceding three years. This stable behavior was also observed in the subsequent years.

194 LAST TO JOIN THE GOLD STANDARD

Figure 7.11 Government debt to the Bank of Portugal and note issue in 1931 (weekly figures) Source: Direcção Geral de Estatística, Boletim Mensal and Annual Reports of the Bank of Portugal

Figure 7.12 shows the weekly behavior of gold and gold-denominated assets reserves during 1931. In July, the gold reserves increased sharply in value as a result of their valuation to the new gold parity of the escudo. The value of gold reserves was maintained virtually unchanged until the end of October, when the Bank of Portugal acquired 3.5 tons of gold. Convertible foreign exchange and other convertible short-run assets in the Bank’s reserves increased until the second week of August, when they declined until the last week of September. This decline followed a reduction of the discount rate from 7.5 per cent to 7 per cent which occurred in August 10. In the face of the ongoing recession, this was aimed at improving the level of domestic economic activity. If we consider that stock prices reflect the state of general economic conditions and expectations about its future, it is worth observing the reaction of stock prices to the adoption of the new monetary regime. Figure 7.13 plots the evolution of five stock price indices.26 In general, stock market prices declined during the whole period. However, their decline slowed down as of mid-1931. This behavior of stock prices suggests that the resumption of the convertibility of the escudo had a positive impact, improving the state of expectations about future economic conditions. Indeed, after the stabilization efforts of the past three years, the de jure adoption of a gold-exchange standard represented a strong commitment of the authorities to sound public finance and monetary and exchange rate stability. THE BRITISH ABANDONMENT OF THE GOLD STANDARD AND THE PEG TO STERLING Following the bankruptcy of the Kredit Anstalt and the German banking crisis, massive capital outflows occurred from Great Britain, substantially reducing the

PORTUGUESE CURRENCY EXPERIENCE 195

Figure 7.12 Bank of Portugal reserves in 1931: gold and foreign assets (weekly figures) Source: Direcção Geral de Estatística, Boletim Mensal and Annual Reports of the Bank of Portugal

Figure 7.13 Stock price indices, 1929–1932 (monthly figures) Source: Direcção Geral de Estatística, Boletim Mensal

Bank of England’s gold reserves. The British authorities reacted by obtaining substantial loans from the French central bank and from the Federal Reserve Bank of New York (£130 million) raising the discount rate from 2.5 per cent to 4. 5 per cent in the space of a few weeks. Ramsay MacDonald’s national coalition government, formed in the beginning of September 1931, announced a restrictive fiscal program, raising some taxes and drastically reducing some expenditures, primarily through a reduction in the salaries of public officials. The political and economic difficulties of Britain and the naval mutiny which occurred that September severely affected the confidence of foreign investors who made massive withdrawals from the Bank of England. Unable to retain the financial support of France and of the United States, or to avoid the complete depletion of gold reserves, the Bank of England suspended the convertibility of sterling on

196 LAST TO JOIN THE GOLD STANDARD

Figure 7.14 Bank of Portugal reserves in 1931–1934: gold and foreign assets (weekly figures) Source: Direcção Geral de Estatística, Boletim Mensal and Annual Reports of the Bank of Portugal

September 21, 1931. The British abandonment of the gold standard was then followed worldwide by other countries.27 The Portuguese authorities decided to peg the escudo to the British pound.28 Only in December did the government formally declare the inconvertibility of the escudo, authorizing the Bank of Portugal to exchange bank notes for sterling at the defined exchange rate. This suspension was initially declared for a fourmonth period which was subject to several renewals until May 1933, when it was renewed without the imposition of any time limit. The importance of Portuguese trade relations with the United Kingdom (25 per cent of total foreign trade), and the fact that most Portuguese foreign assets were denominated in sterling, explains the decision to peg to the pound. Furthermore, the imposition by Brazil of a moratorium on her debt service, together with the imposition of exchange controls, severely reduced the inflow of remittances and capital incomes from that country, which had represented a major component of Portugal’s current account. Under these conditions, the depreciation of the escudo relative to third currencies, for example the dollar (see Figure 7.10), would increase the competitiveness of Portuguese exports, namely with respect to the gold bloc countries, without affecting trade with Britain. Indeed, according to the 1931 annual report of the Bank of Portugal, the trade deficit fell from 1,462 million escudos in 1930 to 920 million in 1931. The trade deficit, as a percentage of GDP, declined from 9 per cent in 1930 to around 5.5 per cent in 1931 and 1932. Although the peg to sterling was intended to avoid adverse consequences for the current account, the Portuguese authorities remained committed to the goal of a stable escudo.29 They believed that the sterling crisis was transitory and that

PORTUGUESE CURRENCY EXPERIENCE 197

the United Kingdom would be able to stabilize the pound and return soon to the gold standard. Pegging the escudo to sterling would be a way to ensure stability. However, sterling was subject to several depreciations and the authorities became aware that the adopted peg could not by itself provide the desired stability. Therefore, to avoid sharp depreciations of the escudo, it was decided to abandon sterling and peg to the dollar whenever the pound/dollar rate fell below $3.32. This happened between December 7 and December 11, 1931; from October 27 to November 11, 1932; from November 17 to December 19, 1932; on December 30, 1932; and on July 19, 1933. Later, facing large fluctuations of the dollar, the authorities decided, instead, to peg to the French franc whenever the pound declined below 84.74 francs. The escudo stayed pegged to the French franc from August 3 to November 23, 1933, then pegging to sterling until 1939. In the face of the inconvertibility of the pound, the government authorized the Bank of Portugal to keep in its reserve exchange and foreign assets denominated in sterling. However, despite this authorization, the Bank of Portugal adopted a policy of gold acquisitions in order to increase the reserve. The gold reserve increased by 6.3 million between July 1931 and December 1934, the foreign assets decreased, in the same period, by 0.7 million, and foreign exchange increased by 0.2 million.30 Figure 7.14 illustrates the weekly evolution of the reserves of the Bank of Portugal from 1931 to 1934. As a final note, it should be pointed out that the decision to peg to sterling was well accepted among the business community. At the beginning of October the press reported several meetings between Salazar and major representatives of the business community When leaving one of those meetings, bankers declared their approval of the ongoing policies of the Bank of Portugal with respect to pegging to the pound. For example, the Comércio do Porto,31 in its Financial Review, states ‘Our exchange situation continues, correctly, at the average rate of 110 escudos per pound. In general everybody benefited from the wise decision to keep the escudo stabilised at 110 per pound; the contrary would have caused a great bewilderment with disastrous consequences.’ CONCLUSION: THE ECONOMIC AND POLITICAL IMPORTANCE OF THE 82-DAY CONVERTIBILITY PERIOD After World War I Portugal had experienced, between 1918 and 1924, a record of high budget deficits, high inflation and sharp depreciation of the escudo. In the mid-1920s the stabilization of the exchange rate, the balancing of the budget, and price stability became major policy goals. Between 1924 and 1926 significant progress was achieved towards those goals. However, political instability did not allow enough progress to permit, on that occasion, the adoption of the gold standard. Only when the political regime had stabilized through imposition of a dictatorship was it possible to implement policies leading to the balance of

198 LAST TO JOIN THE GOLD STANDARD

government budgets, reduction of the public debt, especially its floating component, and stabilization of the exchange rate. The goal of resumption of convertibility was a major economic goal of the political regime, expressed by the policy priorities of the minister of finance as of April 1928—Salazar. After a long period of inflation, depreciation, and monetization of public deficits, the success of such a policy would represent an achievement of utmost importance for the still young regime. Indeed, the success of this monetary and financial policy between 1928 and the early 1930s is of crucial importance to understanding the consolidation of the new regime and the political prestige of Salazar. The adoption of a rule such as the gold standard32 was particularly important both domestically and externally. As a domestic plan, it represented an important signal of the regime with respect to its commitment to provide stable money. Having gained reputation during the stabilizing period between 1928 and 1931, the de jure adoption of the gold standard was a way of reinforcing the credibility of the policy and of the regime. At the external level, the resumption of convertibility would be a way to restore the confidence of foreign capital markets. The importance of this goal is clear if we recall that in 1927–28 Portugal was unable to negotiate two major foreign loans under acceptable conditions. The Portuguese return to gold in 1931, however, raises an important question: why did the Portuguese authorities adopt the gold exchange standard at a time when there were already clear signs of international financial instability? After the stabilization of the political regime, the completion of Salazar’s reforms with the return to the gold standard aimed, in face of the ongoing uncertainty in international markets, to attract capital from abroad. Exports and remittances from abroad had declined in 1930 and 1931 in consequence of the international recession, and such an inflow of capital would have allowed Portugal to face those payments difficulties. Also, we can speculate on the political reasons that motivated Salazar to announce the return to the gold standard, despite the international instability in financial markets. He was strongly committed to this goal and built up his political influence and prestige on that basis. Failure to complete his reforms would have represented a major political defeat.33 Finally, the Portuguese experience illustrates several points stressed by Bordo and Kydland (1992). First, the goal to adopt the gold standard in Portugal made sense because it was, at the time, an international standard. In particular, her major trade and financial partner was under such a regime. For similar reasons, the worldwide abandonment of the standard was recommendation for its suspension. Secondly, the Portuguese experience illustrates the importance of political stability (which is not synonymous with dictatorship) for the functioning of the gold standard. Finally, as those authors point out, the gold standard operated, in several countries, as a contingent rule. In fact, in the presence of events such as a major war, authorities suspended specie payments and issued paper money to finance war expenditures. The gravity of the September events for the Portuguese economy—since she was a British satellite explains the

PORTUGUESE CURRENCY EXPERIENCE 199

suspension of convertibility. However, the strong commitment of the Portuguese authorities to a rule providing for a stable money should be stressed. Unable to achieve such a goal through the gold standard, the peg to the sterling was intended to be a close substitute. The escape clauses adopted leading to the peg to the U.S. dollar or to the French franc highlight the importance assigned to this goal. Indeed, as stressed in Bordo and Santos (1993), this commitment by the Portuguese authorities to exchange rate stability was maintained during the subsequent years by the adoption of conservative monetary and fiscal policies. NOTES * The author wishes to thank Michael David Bordo and Jaime Reis for helpful comments. 1 This crisis was associated with three major shocks which occurred at the time: a British ultimatum for Portugal to abandon the African territories between Angola and Mozambique, a sharp reduction in remittances from Brazil due to political instability in that country, and the request, by the Baring Brothers Company, of the payment of one million pounds of the Portuguese debt outstanding in response to the pressure put on their balance sheets by the Argentine debt default of 1891. 2 The republican regime, established in 1910, had declared goals of a balanced budget and a return to convertibility. However, the 1911 monetary reform was limited to a technical change concerning the creation of a new unit of account, the escudo (=1000 reais, the old unit of account). 3 In 1916, in order to finance war expenditures, the Portuguese authorities negotiated a grant by the Bank of England of a £2 million credit. An additional credit of £20 million was obtained during the year. The minister of finance at the time was Afonso Costa. 4 Other countries also experienced significant increases in the price level during this period. See, e.g. Marcelo Caetano (1931) and Friedman and Schwartz (1963). 5 Marcelo Caetano (1931) provides a detailed analysis of this period. 6 At the time, Portugal was under the Sidónio Pais dictatorship. 7 The Caixa Geral de Depósitos also played a major role in the financing of government deficits during this period, 8 The political regime defined by the republican constitution was parliamentary. The executive was assigned by Parliament on the basis of a majority voting rule. Since no political party had an absolute majority, parties had to enter coalitions to be able to form a government. These coalitions were highly unstable, even in the face of minor political events. Also, the occurrence of several revolutionary coups was a cause of political instability. Between 1919 and May 1926, Portugal had thirty governments. According to Macedo (1982), in the five-year period before the May 1926 coup, Portugal had fifteen governments and there were ten coup attempts. 9 Using the Bank of Portugal notes outstanding as a proxy for the money supply, the seigniorage revenues averaged, between 1918 and 1923, 5.8 per cent or real GDP. 10 The Granger causality concept stresses the role of temporal precedence as a criterion to ascertain whether a variable causes another. Given two variables X and

200 LAST TO JOIN THE GOLD STANDARD

Y, X is said to Granger cause Y if past values of X help to predict the current value of Y. Granger bivariate causality tests provided the following results:

The tests used eighteen lags for each variable in the regression. The degrees of freedom of the test are eighteen for the numerator and 137 for the denominator. The results suggest the existence of unidirectional Granger causality from the monetary issue of the Bank of Portugal to the price level, and two-way causality between the price level and the exchange rate. Similar results were obtained when using rates of change of the variables instead of their log levels. 11 The Portuguese authorities attempted to control the depreciation of the escudo by imposing several exchange controls and establishing, in December 1919, a Control Council. For example, a law was enacted in April 1918 according to which import tariffs had to be paid in gold. Initially, only a fraction of the tariffs had to be paid in gold. In August 1921 Parliament voted a law requiring its full payment in gold. It is worth noting that this exigency, by itself, revealed a lack of confidence of the government in the escudo, contributing therefore to speculation and a consequent increase of the nominal exchange rate. The control measures imposed were not successful at all and, during 1920, they were progressively withdrawn. 12 According to Telo (1993) there were, in the mid-1920s, two main views about the conduct of Portuguese macroeconomic policy. One view defended the immediate balance of the budget as a way to promote financial stability and, as a consequence, favor economic growth. The other view supported an appreciation policy of the escudo which was expected to bring about an inflow of capital abroad and increase productive investment. This increase in investment would increase production which, in turn, would eliminate the budget deficit. Alvaro de Castro, who favored the latter view, was the Prime Minister and Minister of Finance of Portugal between December 1923 and June 1924. Daniel Rodrigues succeeded Alvaro de Castro in the Finance Ministry and kept the same policy orientation. 13 As mentioned in Xavier (1950), almost two-thirds of the external debt was held by Portuguese citizens. The service of the debt was paid in gold and amounted to £1.2 million. With this decision, the Portuguese authorities were able to reduce the debt payments in foreign currency by almost two-thirds. 14 As Mata (1987) stresses, expectations played an important role in this reversal of the behavior of the exchange rate. 15 It should be pointed out that, despite the mentioned fluctuations in the real exchange rate, a more detailed analysis of the data for the period from 1918 to the first half of 1931 suggests that purchasing power parity held during this time span. In

PORTUGUESE CURRENCY EXPERIENCE 201

16

17

18

19

20 21

22

23 24

fact, unit root tests suggest that the real exchange rate was a stationary process and both Granger and Johansen tests suggest that the nominal exchange rate and the ratio of the Portuguese to the U.K. price level were cointegrated. It is worth mentioning that Salazar was appointed Minister of Finance of the new government, headed by J.M.Cabeçadas, formed in June 1926. He presented a stabilizing program which would impose a severe budgetary discipline on all the ministries and would assign important control functions to the ministry of finance. This plan was rejected and Salazar resigned on his fifth day in office. Sinel de Cordes did not favor an appreciation policy of the escudo. He supported a gradualist policy which would not impose severe short-run sacrifices. Such a gradualist policy would only have been possible on the basis of foreign loans. The high level of public deficits, past and current political instability, even within the victorious forces, might explain the lack of external credibility of the new regime. Indeed, these were the recommendations of the Monetary sub-Commission of the Brussels Conference (1920). In his monetary reform, Salazar adopted a statedependent rule to return to convertibility, according to which resumption of convertibility would occur at the current stabilized exchange rate as soon as the government budget was balanced and the floating debt was substantially reduced. The internal floating debt was fully liquidated in 1935. The new definition of the escudo accommodated its post-war depreciation at 1/24. 444 of its 1911 definition. Contrary to the experience of her major trade and financial partner, which adopted the pre-war parity, Portugal decided not to deflate. The appreciation that occurred under the Álvaro de Castro and D. Rodrigues ministries had not been well received by the business community and important political groups, a fact that also contributed to the 1926 military coup. So, in face of the ongoing international recession, Portuguese authorities were not willing to impose the hardship of such a deflation, which could have been politically costly for the still young regime. It is also of interest that in his monetary reform, Salazar closely followed the recommendations and suggestions of the international conferences promoted by the League of Nations. This might have reflected his concern with gaining international credibility. Salazar’s case for defining the new parity at 110 on the basis of ease in conversion between the escudo and sterling may not appear convincing. The choice of a parity of 100 would have made conversion even easier. Indeed, although 110 was close to the 108.25 current rate, its choice conceals the aversion to deflation as mentioned above. On May 22, 1931, the Comércio do Porto in its Financial Review announced that ‘The Minister of Finance has the intention of stabilizing the exchange rate at a price slightly above the current rate, with the laudable purpose to foster exports a little more.’ In fact, as Figure 7.5 shows, the establishment of the new parity, together with the deflation that had already occurred, implied a depreciation of the escudo/ sterling real exchange rate. This depreciation was somewhat above 12 per cent between May and July of 1931. Charter of the Bank of Portugal, July 8, 1931. In fact, due to the short duration of the gold standard, gold coins were never issued.

202 LAST TO JOIN THE GOLD STANDARD

25 A slight U shape in the line describing the stock of notes outstanding can be observed. This shape reflects the seasonal behavior of the series which peaks at the end of the civil year. 26 The indices are defined as follows: Index I—credit institutions, Index II— industrial companies, Index III—transportation companies, Index IV—insurance companies, Index V—colonial companies. 27 Some major European countries, such as France, Holland, Switzerland, etc., remained on the gold standard forming what was known as the gold bloc. 28 Salazar, in the report accompanying the 1930–31 public accounts, published in October 1931, explains the reason for this decision: ‘…namely,what we had to determine was the decision which, at the moment, would have the smallest impact on the national economy as a whole, which could best defend the vital interests of the nation, and which could more easily, although sacrificing a bit and bending a little to the violence of the storm, prepare even more solid conditions than those we had for the future of our currency.’ 29 In the report mentioned in the previous note, Salazar unambiguously reaffirms his commitment to a stable currency: ‘I don’t have to correct undertaken positions or former statements, substitute for others the principles adopted in the past; nor the needs fulfilled by the June reforms with respect to the currency and to the issuing bank, ceased to be, as a result of the current adversities and transient deviations, in the first plan of the Government’s preoccupations.’ 30 Figures reported in Branco (1958) p.155. 31 O Comércio do Porto, October 22, 1931. 32 Bordo and Kydland (1992) stress the role of the gold standard as a policy rule. ‘There are examples where society has overcome time-consistency problems by instituting commitment in the form of laws…. An institutional arrangement such as a gold standard may seem to have the potential for working as a well-understood and explicit rule. At times, it has been suggested that the gold standard, or, more generally, a commodity standard, be reinstated, presumably in order to discipline policy makers.’ 33 Despite the instability in international markets, Salazar probably never expected Britain to abandon the gold standard. Anyway, he could always argue that he had succeeded in his reforms, and if someone should be blamed, blame the British. Indeed, his speeches referring to the collapse of the British gold standard support this interpretation.

REFERENCES Amorim, Diogo Pacheco de (1946) ‘Preços’, Revista do Centro de Estudos de Economia, 2, pp.89–101. Banco de Portugal—Annual Reports, several issues. Barros, J.A.Correia (1938) O Problema do Estalão Monetário, Coimbra, Coimbra Editora Lda. Boletim da Previdência Social, several issues, 1921 to 1933. Bordo, Michael D. (1981) ‘The Classical Gold Standard, Some Lessons for Today’, Federal Reserve Bank of St. Louis Review, 63(5), pp. 2–17.

PORTUGUESE CURRENCY EXPERIENCE 203

Bordo, Michael D. and Kydland, Finn E. (1992), The Gold Standard as a Rule, Federal Reserve Bank of Cleveland, Working Paper No. 9205. Bordo, Michael D. and Santos, F.T. (1995), ‘Portugal and the Bretton Woods International Monetary System’, in James Reis (ed.) International Monetary Systems in Historical Perspective, London, Macmillan Press. Bordo, Michael D. and Schwartz, Anna (1984), A Retrospective on the Classical Gold Standard 1821–1931, University of Chicago Press. Branco, Carlos de Barros Soares (1958) Aspectos da Questão Monetária Portuguesa, Lisbon: Bank of Portugal. Caetano, Marcelo (1931) A Depreciação da Moeda depois da Guerra, Coimbra, Coimbra Editora Lda. Capie, Forrest and Webber, Alan (1985) A Monetary History of the United Kingdom Vol. 1, London, George Allen & Unwin. Comércio do Porto, daily issues, April to December 1931. Diério de Notícias, daily issues, April to December 1931. Diério do Governo (1931) Decreto Lei n0 19869 (June). Direcção Geral de Estatística (1929–1933), Boletim Mensal. Friedman, M. and Anna Schwartz (1963) A Monetary History of the United States 1867– 1960, Princeton University Press. Macedo, Jorge Braga de (1982) Portfolio Diversification and Currency Inconvertibility: three essays in international monetary economics, Lisbon: Universidade Nova de Lisboa—Faculdade de Economia. Mata, Eugénia (1987) ‘Câmbios e Política Cambial na Economia Portuguesa 1891–1931’, Cadernos da Revista de História Económica e Social, 8. Mata, Eugénia and Valério, Nuno (1994) História Económica de Portugal, uma Perspectiva Global, (Lisboa, Editorial Presença). Neves, João L.César (1993), ‘Long Term Series for the Portuguese Economy’, Universidade Católica Portuguesa, Departamento de Economia, Working Paper 44/ 93 (Lisboa). Pereira, H.Caeiro (1985) Banco de Portugal, do passado ao presente, Lisbon: Bank of Portugal. Reis, Jaime (1991) The Gold Standard in Portugal, 1854–1891, Conference on the Gold Standard in the Periphery, 1854–1939, Lisbon. Telo, António José (1993) ‘A Ditadura Financeira’ Como Caminho Para a Unidade Política—os Primeiros Anos de Salazar, II Encontro de História Económica Portuguesa, organized by the Instituto de Ciências Sociais, Curia, June 1993. Valério, Nuno (1983) ‘A Moeda em Portugal 1913–1947’, Cadernos da Revista de História Económica e Social, 5. Xavier, Alberto (1950) Memórias da Vida Pública, Lisbon, Livraria Ferin.

204

8 MONETARY STABILITY, FISCAL DISCIPLINE AND ECONOMIC PERFORMANCE The experience of Portugal since 1854 Eugénia Mata and Nuno Valério

AN OVERVIEW Table 8.1 presents a summary of the behavior of the main variables related to economic performance, monetary evolution, foreign accounts and public finance in Portugal between 1854 and 1990. Besides the whole period under consideration, phases defined according to the behavior of the above-mentioned variables are considered. Annual data are presented in Table 8.2, followed by Technical Remarks. The overall picture that emerges from this summary has five main features. First, all variables have long-term upward trends. Moreover, real per capita gross domestic product also presents a long-term upward trend, which means that Portugal has shared, at least during significant periods, the experience of modern economic growth. Second, phases of higher inflation show on average a slower economic growth, or even decreases of activity, while phases of lower inflation, or of stable or even declining prices, are associated with faster real growth. When the inflation rate decreases from one phase to the next, real gross domestic product rises faster, and when the inflation rate increases from one phase to the next, growth slows down. Third, phases of higher depreciation show on average slower economic growth, or even decreases of activity, while phases of lower depreciation, or of stable or even appreciating exchange rates, are associated with faster real growth. When the exchange rate has a higher rate of depreciation in one phase than in the preceding one, growth slows down, and when the exchange rate is more stable in one period than in the preceding one, real gross domestic product rises faster. There is, however, one exception: during World War II the exchange rate was as stable as during the preceding (and following) phases, but economic performance was clearly worse. The gold standard and ‘belle époque’ phases might seem to provide another such example, but the computed appreciation trend of the exchange rate during the ‘belle époque’ is misleading, as shall be seen below.

Notes a 1891–1990 only b stable under gold standard c statistically not different from 0 (F-test 1% confidence interval)

Table 8.1 Public accounts, monetary evolution and economic performance in Portugal 1854–1990: a summary

206 PORTUGUESE CURRENCY EXPERIENCE

STABILITY, DISCIPLINE AND PERFORMANCE 207

Fourth, phases of higher growth of the money supply show on average slower economic growth, or even decreases of activity, and phases of lower growth of the money supply are associated with faster real growth. When the increase of money supply accelerates from one phase to the next, growth slows down, and when the increase of money supply decreases from one phase to the next, real gross domestic product rises faster. There is, however, again one exception: the growth of money supply decreased from the gold standard phase (1854–91) to the ‘belle époque’ (1891–1913), but economic performance also deteriorated. Fifth, the behavior of the monetary variables seems to be clearly related to the behavior of public and external accounts. Prices and the exchange rate were both stable during the gold standard phase (1854–91) or when there were surpluses or small deficits in public accounts, but the gold standard monetary system was unable to survive in the long run because of deficits in public accounts, as shall be seen below. Whenever there were significant deficits in public accounts, prices rose, regardless of the way the deficits were financed. Surpluses in the balance of payments were responsible for inflationary pressures in the phases when deficits in public accounts were financed by borrowing in the market, as during World War II and the colonial war (1961–73). Whenever there were deficits in public accounts financed by borrowing from the Bank of Portugal, besides the price level rising, the exchange rate depreciated. We now consider in more detail what happened during each one of the seven phases. THE GOLD STANDARD PHASE (1854–91) Portugal adopted the gold standard in 1854 (law of the July 29) and abandoned it, provisionally from a legal point of view, but definitively for all practical matters, in 1891 (decree of the July 9). The decades before the adoption of the gold standard were decades of monetary instability (see Valério, 1991), and economic stagnation. The circulation of inconvertible paper money issued by the state between the 1790s and the 1830s and of inconvertible notes issued by the Bank of Lisbon between 1846 and 1854 disturbed the bimetallic monetary regime. The Peninsular War against revolutionary and imperial France, the independence of Brazil and several civil wars between the supporters of absolutist, liberal and democratic regimes disturbed the social and political situation. Despite the lack of data comparable to what is available for the years after 1854, it may be confidently stated that the money supply, the price level and the exchange rate all underwent significant fluctuations, and that real gross domestic product did not show any significant upward trend. In a certain sense, the gold standard adopted in 1854 was an instrument that put an end to the bad consequences of monetary instability, widely recognized by informed opinion in the country. A detailed analysis of the problems envolved in the process of adopting the gold standard may be found in Chapter 6 by Reis.

208 PORTUGUESE CURRENCY EXPERIENCE

During the gold standard phase, the bulk of the money supply in Portugal consisted of gold coins (to be precise, imported British gold coins). Its stock increased very significantly over time, as a result of balance of payments surpluses provided by the excess of emigrants’ remittances and capital inflows over the permanent trade deficit. It is unlikely that all the gold accumulation was really added to the circulating medium: an important part was almost certainly hoarded. The habit of gold coin hoarding, specially among the peasants of the north of the country, has been underlined by anthropological and sociological studies. Anyway, it must be stressed that this scheme of external payments balancing implied a close dependence of the Portuguese short-term economic situation upon the situation of its main emigrants’ destination (Brazil) and capital supplier (the United Kingdom). Subsidiary coins and bank notes complemented the circulation of gold coins. There was a big issuing bank in Lisbon (the Bank of Portugal) and several small issuing banks in the north of the country, resulting in quite different circulating media in the two main economic centers of the country (Lisbon and Oporto). Neither subsidiary coins, nor bank notes exceeded around one-tenth of the amount of gold coins. The amount of bank deposits was even lower, especially after several bank failures during the 1870s had shaken public confidence in the banking system. As there was no scope for devaluation under the gold standard, to maintain this monetary system it was necessary to balance public accounts. Portuguese governments of the period, however, opted for only a long-term balancing. Shortterm deficits to finance public investments, especially in a modern transportation network, to foster the economic development of the country, were judged unavoidable. To finance these systematic deficits by borrowing from issuing banks was out of the question, because it would have led to the collapse of the gold standard. Thus, the government turned to borrowing in the market. As the internal capital market did not have sufficient resources, there was also heavy borrowing in foreign markets, especially in the London market. Long-term balancing of the public accounts was expected to be a consequence of the increase of public revenue resulting from economic growth itself (and from the reduction of public investment during later stages of economic growth). Public accounts showed persistent deficits between 1854 and 1891. The amounts of these public deficits may be considered small when compared to later deficits. However, the inability to put an end to them proved fatal to the gold standard system. The absence of long-term balancing of the public accounts must be considered the consequence of the inability of the government to translate economic growth into higher public revenue and to curb public expenditure. There was a significant amount of public investment during the period, and economic growth did follow (even if it cannot be considered a consequence only of public investment). However, public revenue showed a low elasticity in relation to

STABILITY, DISCIPLINE AND PERFORMANCE 209

gross domestic product, and the reasons that prevented a significant increase of the tax revenue were mainly political (see Mata, 1993a). The increase in expenditure implied by the dynamics of accumulated debt (at a high interest rate because of the lack of confidence of lenders—see Valério, 1986) proved impossible to stop. The price level stagnated under the gold standard, in spite of significant shortrun fluctuations, which may be explained by the still traditional character of the Portuguese economy. Agriculture provided the bulk of gross domestic product and its output was highly dependent upon weather conditions. Output fluctuations were clearly mirrored in inverse fluctuations of agricultural prices, which were reflected in the index of the cost of living, where food had a large weight. The exchange rate had a similar long-run stable behavior, and higher short-term stability, because it was bounded by the gold points. Real gross domestic product showed a significant upward trend, also in spite of important short-term fluctuations. The long-term upward trend is traditionally explained as the result of a combination of several factors: first, the institutional reforms following the establishment of a constitutional political regime, begun during the 1830s but only completed during the 1860s; second, the policy of public investment, mainly in transport infrastructures; third, the stimuli coming from growth of the most developed European countries; fourth, the use of catching-up opportunities by Portuguese entrepreneurs; last (but perhaps not the least), monetary stability. Some of the short-term fluctuations may be related to the fluctuations of agricultural output—this is the case of the crisis of the mid-1850s, already visible in the data presented. Others must be related to the evolution of the economic situations of Brazil and the United Kingdom, because of the role they played as economic partners of Portugal as explained above—it is the case of the crisis of the mid-1870s and also of the crisis of the early 1890s that led to the suspension of the gold standard. Portugal managed to remain on the gold standard monetary system for almost forty years because the budget deficits were small enough to be easily financed by borrowing in the market. In the crucial fiscal years of 1890–91 and 1891–92, however, a general lack of confidence appeared. It was due to several causes. First, Brazil was undergoing a period of social and political instability, because of the abolition of slavery (1888) and the overthrow of the monarchy (1889). Second, there was a diplomatic conflict with Great Britain over several territories in Southern Africa, leading to a British ultimatum (January 14, 1890), demanding Portuguese withdrawal from the territories that today form the states of Malawi, Zambia and Zimbabwe. Third, Baring Brothers, the London banker of the Portuguese government, was for a while on the verge on bankrupcy, mainly because of its South American business. Fourth, a republican revolution broke out on January 31, 1891 in Oporto (this was promptly put down). Fifth, an attempt to issue a loan in foreign markets to provide external liquidity failed. A run on the bank to exchange bank notes for gold coins could not be stopped. A sixty-day moratorium was passed on May 9, 1891 during which the monetary

210 PORTUGUESE CURRENCY EXPERIENCE

authorities tried to ensure a return to normalcy by importing metal to be coined, but it was impossible to overcome the difficulties. As a consequence, the Portuguese gold standard collapsed. Thus, from a short-term perspective, the end of the Portuguese gold standard in 1891 was a consequence of exceptionally high budget deficits, of the reduction of emigrants’ remittances due to the Brazilian crisis, and of the inabilities of the London market (and especially of Baring Brothers) to help the Portuguese government through the international crisis. However, from a long-term point of view, it is possible to say that the gold standard could not have lasted anyway, because of the trend of increasing public expenditure due to the dynamics of accumulated debt, and the high trade deficit. A significant increase of public revenue and evolution towards a balancing of foreign trade would have been necessary to provide long-term stability to the Portuguese gold standard. THE ‘BELLE ÉPOQUE’ (1891–1914) The events that led to the suspension of the gold standard in 1891 were certainly a turning point in Portuguese monetary and financial history. Portugal was led to a phase of higher monetary instability and worse economic performance than it had experienced under the gold standard, though still far better than what was to come with World War I. The composition of money supply changed dramatically: gold coins were hoarded or exported, causing a very sharp decline of the money supply in 1891. The effect is certainly exaggerated in the data presented, because a significant quantity of gold coins were probably already out of circulation. Bank notes became the main part of money supply and gradually grew to amounts comparable to those of previously circulating gold coins. Subsidiary coins and bank deposits remained less important items of the money supply. Reis (1991a) argues that gold coins remained in circulation at least until World War I, but gold coins were commercially overvalued and tended to increase in trade value, which naturally led to hoarding. For the rest of the ‘belle époque,’ the Portuguese money supply included a majority of bank notes, and was mainly determined by government borrowing from the Bank of Portugal (which became the sole issuing bank, in the wake of the 1891 crisis). The rate of growth of the money supply appears to have been lower than under the gold standard, but this may be somewhat illusory because the hoarding of gold coins pre-dated 1891, and it certainly had more significant inflationary consequences. Classical standards remained the guidelines for the behavior of Portuguese monetary authorities throughout the ‘belle époque.’ Thus, they tried to balance public accounts and trade flows. To balance public accounts, an increase in revenue and a reduction in expenditures were needed. The increase in revenue was obtained by a tax on the interest of the domestic public debt (1891), and later by tax surcharges. The reduction in expenditure was obtained by suspending the payment of

STABILITY, DISCIPLINE AND PERFORMANCE 211

amortization and part of the interest in the foreign public debt (1892), and later by curbing expenditure in general. This had some political costs, which may have been decisive in fostering republican opposition which led to the overthrow of the monarchy in 1910. While deficits persisted for most of the period 1891– 1914, they were clearly smaller than during the gold standard years, and some surpluses began to appear, especially on the eve of World War I. Increased protection was imposed in an attempt to balance foreign trade. This proved a failure in the long run, and may have contributed to the economic stagnation that characterized this phase. It was also detrimental to the balancing of public accounts and to social and political stability. Emi-gration increased, as in other European countries. Due to increased emigrants’ remittances, there was no significant stress on the balance of payments. This did not prevent the appearance of some inflationary pressures and some depreciation of the Portuguese currency, since public deficits, however small, still had to be partially financed by borrowing from the Bank of Portugal. Between 1891 and 1898, the Portuguese public debt increased, with the price level, and there was some depreciation of the Portuguese currency. Between 1898 and 1907 there was no increase in the public debt, the price level stabilized, and the Portuguese currency appreciated until it attained its previous par. Between 1907 and 1914, there was no increase in the public debt, but prices rose, and again there was some depreciation of the Portuguese currency. The possibility of restoring the gold standard was never formally abandoned, but neither could it be seriously considered. Indeed, 1912–13 and 1913–14 were the only fiscal years to show two consecutive surpluses in public accounts. The gold reserves of the Bank of Portugal were too low to have sustained a possible run on the bank, should convertibility have been restored. The consequences of this monetary and financial evolution for economic performance are still being debated. The depreciation of the Portuguese currency and the protectionist measures provided short-term stimuli for economic activity during the 1890s. It has even been argued that the 1890s showed better economic performance than the preceding decades, but we are not persuaded by the evidence provided in Lains and Reis (1991) on the matter. Anyway, the appreciation of the Portuguese currency around the turn of the century and during the early years of the first decade of the twentieth century, certainly changed the picture. At the same time, the positive impact of the protectionist measures faded and there only remained their negative impact on the use of resources (especially in agriculture, where inefficient cereal production was encouraged at the expense of more efficient fruit and vegetable production) and on the cost of living (especially because cheaper foreign cereal imports were prevented from competing with domestic cereals). Thus, a stagnating trend is evident for the early twentieth century, and the renewed depreciation of the Portuguese currency on the eve of World War I failed to foster economic growth.

212 PORTUGUESE CURRENCY EXPERIENCE

In the short run, the inverse association between economic performance and inflation remained valid, for the same reasons as during the gold standard phase. WAR AND POST-WAR INFLATION (1914–24) World War I and the immediate post-war years were a period of inflation and economic distress for Portugal. The composition of the money supply remained quite similar to what had prevailed during the ‘belle époque,’ and its increases mainly came from the same source: government borrowing from the Bank of Portugal in order to finance budget deficits. The increase in expenditure resulting from military activity (mainly to defend Angola and Mozambique against German attacks) consequently implied higher increases in the money supply. Together with supply problems due to the war, this was the cause of the rise of prices. The rise in the velocity of circulation triggered by the rise of prices itself provided a feedback mechanism to inflation. The depreciation of the Portuguese currency remained moderate during the war, because borrowing from the Bank of England (to pay for Portuguese troops sent to Flanders) avoided incurring huge balance of payments deficits, but depreciation accelerated as soon as it became clear that the implementation of post-war financial arrangements (especially concerning war debts and war reparations) would not be made according to the agreed schedule. Though there was never a complete loss of control, the country lived for a while on the verge of hyperinflation. The size and mechanisms of the production crisis that accompanied the war and the immediate post-war years need some elaboration. The size of the crisis is overestimated in the series presented. The series are based on data from the monetized sector of economic activity, which certainly suffered far more than the still significant self-consumption sector. The mechanisms of the crisis were linked to Portuguese dependence upon foreign supplies of several vital commodities (among them cereals and fuels) and upon foreign ships (mainly British) for their transportation, and to the impact of Germany’s submarine war. At the same time, there was no significant demand pull, because Portuguese industries were unable to provide the armaments production needed to fight the war. Of course, these problems disappeared with the end of the war, but monetary disarray stunted a prompt recovery. The short-run inverse relation between prices and economic performance disappeared during this phase. This is not surprising, given the fact that price behavior was no longer mainly linked to harvest fluctuations. STABILIZATION (1924–39) The indebtedness of the government to the Bank of Portugal ended in 1924. A clear schedule for the payment of war debts and reparations was also set up by the

STABILITY, DISCIPLINE AND PERFORMANCE 213

mid-1920s. This promised an end to the internal and external depreciation of the Portuguese currency. Yet the achievement remained shaky until the public accounts deficit finally disappeared in the fiscal year of 1928–29. Monetary stabilization resulted from the ending of government borrowing from the Bank of Portugal. For a while remaining public deficits were financed by borrowing in the capital market. Then, from the late 1920s on, public accounts began to show surpluses, putting an end to the threat of monetary problems caused by public accounts deficits. As a consequence of these changes, between the mid-1920s and World War II the monetary base was increased due to an inflow of international funds produced by surpluses in the balance of payments. At the same time, there was a relative increase in bank deposits, and a significant widening of the use of bank deposits as a means of payment. In spite of the resulting increase in the money supply, prices did not exhibit signs of an upward trend. This may be attributed to economic growth, and to a reduction in the velocity of circulation, which was itself a consequence of the end of inflation. The end of the depreciation of the Portuguese currency in 1924 was obtained by means of active government intervention (for a detailed description of this policy, see Mata, 1987). From then on, the exchange rate remained stable, as a result of the surpluses of the balance of payments and of the intervention of the monetary authorities. On July 1, 1931, there was even an attempt to adopt the gold exchange standard system, by restoring full convertibility of the Portuguese currency with the British currency, which was itself convertible in gold. However, less than three months later, the United Kingdom suspended the convertibility of the pound into gold. The Portuguese monetary authorities faced a dilemma: to abandon gold would have been a political defeat, but to stick to gold would have meant losing competitiveness in foreign markets. The Portuguese government chose to peg the Portuguese currency to sterling, and subsequently to the dollar and the franc, to avoid both inflationary pressures and loss of competitiveness of Portuguese exports in international markets. For a detailed account of this episode, see Chapter 7 by Santos. The economic recovery of the late 1920s was certainly helped by international circumstances. The situation changed with the Great Depression. Nevertheless, Portugal was able to achieve some economic growth even during the 1930s. From a structural point of view, this was a consequence of the overcoming of some crucial problems that had hitherto checked sustained economic growth. Among them, the quality of human resources and the efficiency of the banking system were perhaps the most important. As a matter of fact, literacy and school enrollment rates attained what some authors consider critical thresholds for modern economic growth (for a discussion of this point, see Nunes, 1993) and general confidence in the banking system allowed a much better use of available savings. From a short-term point of view, credit must be given to a skillful policy mix that combined some curbing of private and public consumption by means of

214 PORTUGUESE CURRENCY EXPERIENCE

balanced current public accounts, some stimuli to investment by means of lowering interest rates and launching public investment programs, and a careful competitive devaluation of the Portuguese currency. Price changes and economic activity are negatively associated during most of this phase, but this relation is replaced by a positive association during the early years of the Great Depression. It is possible to suggest that the old negative relation due to the dominant role of agriculture in the Portuguese economy was being replaced by a new positive association due to the reaction of economic activity to short-term stimuli provided by price increases. WORLD WAR II (1939–45) In a certain sense, World War II was similar to World War I so far as the Portuguese economic situation is concerned: inflation and a scarcity crisis put an end to monetary stability and economic growth. However, differences were as striking as similarities. In spite of the fact that Portugal did not participate as a belligerent country in World War II, the consequences of the conflict to public accounts were quite similar to those of World War I. The need to increase military expenditure and the reduction of revenue due to a reduction of economic activity implied significant deficits. Their financial consequences, however, were quite different. The Portuguese government was able to finance the budget deficits by borrowing in the market. This was a consequence of the confidence generated during the previous phase, and of the availability of capital, due to huge surpluses in the balance of payments resulting from the export of tungsten ores and from capital brought in by refugees fleeing from Nazi-occupied Europe (sometimes en route to America, sometimes as a definitive refuge) and from remittances sent to them (mainly from the American Jewish community). Thus, the economic environment of World War II brought some benefits to Portugal. In the monetary field this implied an increase of the money supply (triggered by the surpluses of the balance of payments), an inflationary process (which the government tried to fight by borrowing in excess of its needs, to finance the deficits of public accounts) and stability of the exchange rate (indeed, there was even some appreciation against sterling). Neutrality also avoided critical supply problems when compared with what had happened during World War I. At the same time, the boom of tungsten ore exports and the remittances sent to refugees provided some stimuli to production. Thus, the recession of economic activity was much milder than during World War I.

STABILITY, DISCIPLINE AND PERFORMANCE 215

THE POST-WAR PROSPERITY (1945–74) Between the late 1940s and the early 1970s, Portugal shared the relative monetary stability and economic prosperity of the international economy, and made its true entrance in the era of sustained modern economic growth. The evolution of the money supply remained quite similar to what had happened during the preceding phases. The increases in the monetary base were mainly the result of balance of payments surpluses, rather than government borrowing from the Bank of Portugal. Bank deposits reinforced their quantitative and qualitative role. Between the end of World War II and the beginning of the colonial wars in 1961, the situation was rather similar to the period 1928–39 in regard to public accounts. Most of the years closed with surpluses. When deficits appeared, they were small and easily financed by borrowing in the market. At the same time, the balance of payments presented systematic deficits. The money supply increased smoothly, there was price stability and the exchange rate behaved according to the Bretton Woods rules. After 1961 the situation changed completely, though Portugal only became a member of the International Monetary Fund and of the World Bank in 1962. The colonial wars against the independence movements in Angola, GuineaBissau and Mozambique were responsible for high military expenditures and for systematic budget deficits. These deficits were financed by borrowing in the market and were not the cause of any monetary pressure in the short run. At the same time, the balance of payments began to show significant surpluses, mainly because of a boom in tourism and an increase in emigrants’ remittances that followed a spurt of emigration to the most developed countries of Western Europe, especially France. This implied a higher increase of money supply and some inflationary pressure, but the exchange rate of the Portuguese currency remained stable. The escudo even appreciated slightly in the readjustments that followed the declaration of inconvertibility of the U.S. dollar in 1971. From a short-term perspective, Portuguese economic growth during this phase may be seen as a consequence of international prosperity and of Portuguese participation in the process of European economic integration, first as a member of the European Organization for Economic Cooperation (though no significant help came from the Marshall Plan, because of Portugal’s comparatively good economic situation after the war), later as a founder member of the European Free Trade Association, and as an associate (later full) member of the European Community from 1972 on. It is also worth noting that during the 1960s the Portuguese government was able to come back to the international financial markets as a borrower. From a long-term perspective, the structural changes already visible in the inter-war period were certainly equally decisive for this period of Portuguese economic growth, and especially for the catching-up process it implied.

216 PORTUGUESE CURRENCY EXPERIENCE

At the same time, a positive relationship appeared between increases in the price level and increases in economic activity. This suggests that, as a consequence of its increasingly non-agricultural character, the Portuguese economy began to respond positively to inflationary stimuli. RECENT YEARS (SINCE 1974) The first oil shock between the autumn of 1973 and the spring of 1974 hurt the Portuguese economy very badly, as it was coupled with a transitory embargo by Arab countries against Portugal. This was in retaliation for the authorization given to American aircraft to use Portuguese bases to send military supplies to Israel during the Yom Kippur war. Within a few months emigration stopped, balance of payments surpluses turned into deficits, the authoritarian regime that had ruled the country since the mid-1920s was overthrown, and a period of political and social unrest started. There followed attempts at deep institutional reform (agrarian reform, nationalizations and so on), and decolonization brought to the country around half a million refugees (more than 5 per cent of the total population). The end of the colonial wars and of the military expenditure they implied did not restore the Portuguese public account to surplus. On the contrary, deficits increased as the new democratic regime tried to catch up with the more developed European countries in developing the structure of a welfare state. Moreover, as it was impossible to finance the increased deficits by borrowing in the market, huge borrowing from the Bank of Portugal took place. A significant increase in the monetary base and in the money supply ensued, with a significant impact on the price level. Inflation accelerated. The Portuguese currency began to depreciate in the exchange markets from 1977 on. This was partly a spontaneous evolution, partly an attempt by the Portuguese monetary authorities to ensure some short-term gains from the competitiveness of Portuguese exports. In the late 1970s, stabilization and recovery eventually came with the improvement of the monetary situation. This was partly the consequence of a stabilization program negotiated with the International Monetary Fund. The second oil shock, the appreciation of the American dollar and the rise of international interest rates brought new trouble to the Portuguese monetary and economic situation during the early 1980s. There was another stagnation period, inflation accelerated, external payments problems arose again, and it was necessary to negotiate a second stabilization program with the International Monetary Fund. These trends were mitigated during the late 1980s as a consequence of a better external background for the Portuguese economy. Portugal had become a full member of the European Community, and international interest rates, the dollar exchange rate and the oil price decreased. This allowed some reduction of the deficits of the public accounts, which could be financed again by borrowing in

STABILITY, DISCIPLINE AND PERFORMANCE 217

the market, a deceleration of the inflation, and a stabilization of the exchange rate of the Portuguese currency. Moreover, economic growth came back again. In 1992, the Portuguese government even thought the moment had come for the Portuguese currency to be integrated in the European Monetary System. The decision was taken to join during the first half of the year, a few months before the crisis that hit the system in September of that year. In spite of this crisis, full exchange convertibility of the Portuguese currency was restored in December 1992. This period of improved economic performance lasted until the present international recession. At the same time, most of the economic institutional reforms of the mid-1970s were undone by an ongoing process of privatization. The reduction of long-term growth since 1974 is, of course, quite natural given the international situation. Nevertheless, the Portuguese economy was able to maintain a gradual catching-up process with its most developed partners of the European Union, although it is still very far from attaining a similar per capita gross domestic product. The relationship between the evolution of real gross domestic product and of the inflation rate changed in the late 1970s. It is clear that the Portuguese economy ceased to respond to inflationary stimuli in the same positive manner as during the 1950s, the 1960s or even the 1970s. Monetary stability is now the key for a good economic performance. CONCLUSION Concerning the relationship between public finance and monetary evolution, the Portuguese experience since 1854 may be summarized by saying that fiscal discipline was always crucial for monetary stability. Public accounts deficits financed by borrowing from the Bank of Portugal were always associated with inflation and currency depreciation in the short run. Deficits of the public accounts in general, even when financed by borrowing in the market, usually triggered inflation and currency depreciation, at least in the long run. Concerning the relationship between monetary evolution and economic performance, the Portuguese experience since the mid-nineteenth century confirms that money is clearly detrimental to economic performance when it functions badly. The inflationary phases of World War I, the interwar years, World War II, and even of part of the last quarter of the twentieth century show this. To form some conclusions about the effects of stable money on economic performance, we must, first of all, distinguish the phases before and after the take-off into sustained modern economic growth. Before the take-off into modern economic growth, the gold standard performed better than other monetary systems in the long run. There was an inverse relationship between economic performance and prices, due to the traditional mechanism of good and bad harvests and their opposite impact on economic well-being and prices.

218 PORTUGUESE CURRENCY EXPERIENCE

After the take-off into modern economic growth, the economy became sensitive to inflationary stimuli to growth. Yet it is doubtful how long and how far such a mechanism may be used without perverse effects. Inflationary stimuli fostered economic growth during the 1950s, 1960s and even 1970s, but the mechanism faded out during the late 1970s. During the 1980s and early 1990s, monetary stability was a necessary condition for economic growth. To sum up: the Portuguese experience suggests there are strict limits to the use of inflationary stimuli in small and open economies because they often lead to balance of payments problems. STATISTICAL APPENDIX

Technical remarks Calendar years and fiscal years Data for real gross domestic product, gross domestic product deflator, money supply, exchange rate and overall balance refer to calendar years. Data for public debt at the Bank of Portugal and for budget surplus or deficit refer to fiscal years. The fiscal years from 1854–55 to 1933–34 began on July 1 of each calendar year and ended on June 30 of the next calendar year. The figures for public debt at the Bank of Portugal and for budget surplus or deficit presented for the years 1854 to 1934 refer to the fiscal years that ended on that calendar year. The fiscal year 1934–35 began on 1 July 1934 and ended on 31 December 1935 (that is to say, it lasted for eighteen months). The figures for public debt at the Bank of Portugal and for budget surplus or deficit presented for the year 1935 refer to the fiscal year 1934–35. The fiscal years from 1936 and on coincide with calendar years. Gross domestic product The series of real gross domestic product is based on the series published in Nunes, Mata and Valério (1989). For the years 1947 and on, it is the official series for the continental part of the country, adjusted to include the activity of the archipelagos of Madeira and the Azores. For the years before 1946, it is an estimate based on exports, imports, fiscal revenue and public expenditure as proxy variables. This series has been the subject of some controversy. On the resulting debates about Portuguese economic performance, see Lains (1990), Lains and Reis (1991), Nunes, Mata and Valério (1991), Nunes, Mata and Valério (1992),

Table 8.2

STABILITY, DISCIPLINE AND PERFORMANCE 219

220 PORTUGUESE CURRENCY EXPERIENCE

STABILITY, DISCIPLINE AND PERFORMANCE 221

222 PORTUGUESE CURRENCY EXPERIENCE

Source: Mata and Valério (1994) and their sources. Notes * Data not available. a Par value b The evolution of these figures reflect the withdrawal of gold coins from circulation—see technical notes. c Not homogeneous with previous figures, because of differant coverage of banking system —see technical notes. d Previous figures rough estimate; subsequent figures official data—see technical notes. c Previous figures pound sterling exchange rate; following figures U.S. dollar exchange rate. The pound sterling exchange rate in 1940 was 102.726; the U.S. dollar exchange rate in 1939 had been 24.944.

STABILITY, DISCIPLINE AND PERFORMANCE 223

224 PORTUGUESE CURRENCY EXPERIENCE

Esteves (1993), and Marques and Esteves (1994). The main doubts raised during these debates concern the poor performance of the Portuguese economy during the 1890s, World War I and interwar years, and during World War II. We would acknowledge that the size of the interwar crisis is overestimated (especially in what concerns the 1921 figure), but do not find the arguments about the other periods convincing. This series was the basis of the statistics presented in Neves (1994). Gross domestic product deflator The gross domestic product deflator is based on the consumer price index. The series of the consumer price index is based on an official series of an index of cost of living (for the period 1900–81, extended for the period 1854–99 using data from Justino 1988–89), and on a consumer price index (for the period 1981–90), duly linked using the 1981 levels. This implies that the series is not homogeneous, because it has several breaks, due to a different coverage of goods, mainly between 1899 and 1900 and between 1980 and 1981. Money supply The series of money supply is based on the series of coins published in Sousa (1991a) (for the period 1854–90), and on the series of money supply published in Reis (1990) (for the period 1854–1912), and Valério (1984) (for the period 1913– 45), and by the Bank of Portugal (for the post-World War II years). The series is not homogeneous, because it has a break, due to a different coverage of the banking system, in 1912. For the period 1891–1912, we subtracted gold coins from the figures presented in Reis (1991a), because we do not believe they were really used as a means of payment, but only as a store of value, because their trade value was clearly above their nominal value. This implies a sharp drop in the money supply from 1890 to 1891 (due to the hoarding of gold coins) and a high increase from 1891 to 1892 (due to an exceptional issue of bank notes by the Bank of Portugal). We believe that a rise in the velocity of circulation (which may be only partly apparent because a significant fraction of the gold coins might already have been hoarded before 1891) compensated for the bulk of the potential disturbances resulting from these changes. Exchange rate The exchange rate presented is the pound sterling exchange rate for the period 1891–1940, and the dollar exchange rate for the period 1940–90. This corresponds to the exchange rate that mainly concerned the monetary authorities in each period.

STABILITY, DISCIPLINE AND PERFORMANCE 225

Overall balance The overall balance figures for the years prior to 1939 are the estimates presented in Reis (1991), Mata (1987), and Valério (1982). The overall balance figures for the years from 1939 on are official figures. Public debt at the Bank of Portugal The figures for public debt at the Bank of Portugal were computed adding the outstanding amounts of all types of debt at the end of each fiscal year. Figures of total internal and external debt may be found in Mata and Valério (1994). Budget surplus or deficit The budget surplus or deficit was computed by adding fiscal revenue and other effective revenue and subtracting effective expenditure (excluding amortization of public debt). REFERENCES Barbosa, António Manuel Pinto (1983) Problemas monetários internacionais da actualidade, Lisbon, Academia das Ciências. Branco, Carlos Soares (1950) Aspectos da questão monetária portuguesa, Lisbon, Bank of Portugal. Caetano, Marcelo (1931) A depreciação da moeda depois da guerra, Coimbra, Coimbra Editora. Eichengreen, Barry (1985) The Gold Standard in Theory and History, New York, Methuen. Eichengreen, Barry (1988) ‘Resolving debt crisis: an historical perspective’, discussion paper no. 239, Center for Economic Policy Research. Eichengreen, Barry (1989a) ‘The gold standard since Alec Ford’, discussion paper no. 347, Center for Economic Policy Research. Eichengreen, Barry (1989b) ‘International monetary instability between the wars: structural flaws or misguided policies?’, discussion paper no. 348, Center for Economic Policy Research. Eichengreen, Barry (1989c) ‘The comparative performance of fixed and flexible exchange rate regimes: Interwar evidence’, discussion paper no. 349, Center for Economic Policy Research. Esteves, Paulo (1993) ‘Portuguese prices before 1947: inconsistency between the observed cost of living index and the GDP price estimation of Nunes, Mata, Valério (1989)’, Lisbon, Bank of Portugal, working paper no. 19/93. Flanders, June M. (1990) ‘Rules versus rules under the gold standard: The Bank Act of 1844, the Cunliffe Committee and the Bank of England’, European Economic Association, working paper no. 18–90, Lisbon. Hayek, Friedrich (1991) Economic Freedom, Oxford, Basil Blackwell. Justino, David (1988–89) A formação do espaço económico nacional, Lisbon, Vega.

226 PORTUGUESE CURRENCY EXPERIENCE

Kindleberger, Charles (1984) A Financial History of Western Europe, London, George Allen & Unwin. Lains, Pedro (1990) A evolução da agricultura e da indústria em Portugal 1850–1913. Interpretação quantitativa, Lisbon, Banco de Portugal. Lains, Pedro and Reis, Jaime (1991) ‘Portuguese economic growth 1833–1985: some doubts’, The Journal of European Economic History, vol. 20, no. 2, fall. Macedo, Jorge Braga de (1982) Portfolio Diversification and Currency Inconvertibility: Three Essays in International Monetary Economics, Lisbon, Universidade Nova de Lisboa. Marques, Carlos and Esteves, Paulo (1994) ‘Portuguese GDP and its deflator before 1947: a revision of the data produced by Nunes, Mata, Valério (1989)’, Lisbon, Bank of Portugal, working paper no. 4/94. Mata, Eugénia (1987) Câmbios e política cambial na economia portuguesa 1891–1931, Lisbon, Sá da Costa. Mata, Eugénia (1988) ‘As três fases do fontismo: projectos e realizações’, Estudos e ensaios em homenagem a Vitorino Magalhães Godinho, Lisbon, Sá da Costa. Mata, Eugénia (1991) ‘Exchange rate and exchange policy in Portugal 1891–1931 revisited’, Estudos de Economia, vol. XII, no. 1. Mata, Eugénia (1993a) ‘Order and progress or order versus progress: the Portuguese social dilemma of the mid-19th century’, Estudos de Economia, vol. XIII, no. 2. Mata, Eugénia (1993b) As finanças públicas portuguesas da Regeneração à Primeira Guerra Mundial, Lisbon, Bank of Portugal. Mata, Eugénia and Valério, Nuno (1994) História económica de Portugal—uma perspectiva global, Lisbon, Presença. Neves, João César das (1994) The Portuguese Economy—A Picture in Figures. XIX and XX Centuries, Lisbon, Universidade Católica Editora. Nunes, Ana Bela (1993) ‘Education and economic growth in Portugal: a simple regression approach’, Estudos de Economia, vol. XIII, no. 2. Nunes, Ana Bela, Mata, Eugénia and Valério, Nuno (1989) ‘Portuguese economic growth 1833–1985’, The Journal of European Economic History, vol. 18, no. 2, fall. Nunes, Ana Bela, Mata, Eugénia and Valério, Nuno (1991) ‘Portuguese economic growth 1833–1985,’ some comments on Jaime Reis’ and Pedro Lains’ doubts, The Journal of European Economic History, vol. 20, no. 2, fall. Nunes, Ana Bela, Mata, Eugénia and Valério, Nuno (1992) ‘O que sabemos sobre o crescimento económico português entre meados do século XIX e meados do século XX ?’, Estudos de Economia, vol. XII, no. 2. Reis, Jaime (1991a) A evolução da oferta monetária portuguesa, Lisbon, Banco de Portugal. Reis, Jaime (1991b) ‘The gold standard in Portugal, 1854–1891’, paper presented at the international conference on the gold standard in the periphery 1854–1939, Lisbon. Salazar, António de Oliveira (1915) O ágio do ouro, sua natureza e suas causas, 1891– 1915, Coimbra, Imprensa da Universidade. Samuelson, Paul (1976) Economics, 10th edition, New York, McGraw-Hill. Schumpeter, Joseph (1954) History of Economic Analysis, London, Oxford University Press. Sousa, Rita (1991) ‘Money supply in Portugal 1835–1891’, Estudos de Economia, vol. XII, no. 1. Valério, Nuno (1984) A moeda em Portugal (1913–1947), Lisbon, Sá da Costa.

STABILITY, DISCIPLINE AND PERFORMANCE 227

Valério, Nuno (1986) ‘Expectativas dos credores externos sobre a solvabilidade do estado português 1881–1910’, Revista de História Económica e Social, no. 18. Valério, Nuno (1987) ‘Aspectos das finanças públicas portuguesas’, in O Estado Novo— Das origens ao fim da autarcia (1926–1959), vol. 1, Lisbon, Fragmentos. Valério, Nuno (1991) ‘Periodização da história monetária de Portugal’, Estudos de Economia, vol. XII, no. 1. Valério, Nuno (1994) As finanças públicas portuguesas entre as duas guerras mundiais, Lisbon, Cosmos.

228

COMMENT Pablo Martin-Aceña

Portuguese currency history during the nineteenth and twentieth centuries is rich and varied. Portugal is a peripheral country, a late developer closely integrated in the European economy, through its priviliged relations with the United Kingdom, and with special links with Brazil. The reading of Portugal’s experience reveals many common features with international experience, but also considerable differences, which makes Portugal an interesting and intriguing case study. Curiously enough, Portugal was the first nation to join the United Kingdom on the gold standard in 1854 and the last to restore gold convertibility in 1931. Although the three papers included in Part III of this volume deal with the monetary history of Portugal, they are different in kind. Chapter 8, by Mata and Valério, is an overview of 150 years of the nation’s currency events, which basically underscores the importance of the international monetary regime for a country’s economic development, while the other two papers confront the issue of Portugal’s decisions to join the gold standard in 1854 and 1931. As indicated by Mata and Valério, their paper tries to assess the relations between public finance, monetary evolution and economic performance in the long run (from 1854 to present times). To accomplish such an imposing task, they divide the period into several shorter phases, in order to offer a detailed account of what happened to each of the economic and financial variables they have selected (GDP, prices, the exchange rate, the money supply, the balance of payments and the budget). They present the data in Table 8.2 and discuss the sources and the problems of the series under Technical Remarks. Mata and Valério point out three main features of Portugal’s monetary experience: first, that all series—GDP, money, prices and the exchange rate— exhibit a long-term upward trend; second, that growth seems to have been inversely associated with inflation, and hardly related to the evolution of the money supply or the exchange rate; and third, that monetary variables maintained a close association with both the budget deficit and the position of the external account. As a final conclusion they add that the ‘Portuguese experience since the mid-nineteenth century confirms that money is clearly detrimental to economic performance when it functions badly’, and also that ‘fiscal discipline was always crucial for monetary stability.’

230 COMMENT

The authors have amassed a large amount of data and made an attempt to link real and monetary variables. The description of each of the series for each of the periods is clear, but what remains undisclosed is their exact relation; they tell us the evolution of the money supply or that of the exchange rate, but one would like to know the contributions of both the fiscal deficit and the balance of payments to the creation of money To go beyond the description a model is needed, and although such a model might be in the mind of the authors, it is not spelled out in the text. Unfortunately some of the sections are too short to obtain a clear picture of what happened, and perhaps rather than focusing on each of the variables separately it might have been better to have offered a general interpretation of Portuguese financial history over the period. Furthermore it may have been of some interest to comment on the overall picture that emerges from Table 8.1, which allows for a more elaborated set of conclusions than those offered in the text. I must also add that some of the data presented in Table 8.1 are puzzling. With regard to GDP, there seems to be some discrepancy—and an intense debate— between the authors’ figures and those assembled by Lains and Reis. This is not the place to enter into, still less to discuss, the controversy, but even if we go along with Mata and Valério’s figures, they do not offer a convincing explanation of why the 1890s exhibited poor performance and why the economy stagnated between 1900 to 1914. The evidence presented in Table 8.1 casts some doubts on these allegations. Prices remained relatively stable, contrary to the assertion of Mata and Valério, and the public deficit was moderate and tended to decrease as a percentage of GDP; besides, nothing distressing is said about the trade balance, and the exchange rate hovered between 6.0 and 4.5 without significant appreciation or depreciation. Portugal may have stagnated from 1890 to 1914 (in fact it grew at a moderate rate of 1.3 per cent, according to figures for GDP in Table 8.1), but no convincing arguments are put forward. The other statistical query refers to the remarkable drop of the money supply from 152,000 contos in 1889–90 to 78,000 contos in 1891–92, again without any significant change in prices or GDP. This is the result of a sly trick whereby from Reis’ quantity-of-money figures, Mata and Valério have decided to subtract all gold in circulation because they do not ‘believe [gold coins]…were really used as a means of payment, but only as a store of value.’ However, no proofs or satisfactory arguments for this assertion are offered in their paper. There must exist an ample contemporaneous literature in the form of books, articles and pamphlets that tell us how gold was considered and whether it circulated or not. To my knowledge, Reis has done a thorough historical study of the Portuguese quantity of money, and I think we must stick to his figures. But even if we admit that the Mata and Valério assumption is correct and therefore that the deduction is warranted, one would expect prices and/or GDP to adjust a to fall of 100 per cent in the money supply. They suggest that the adjustment occurred through a rise in the velocity of circulation, but again such an increase would be extremely

PORTUGUESE CURRENCY EXPERIENCE 231

sharp, it would not have passed unnoticed at the time, and, besides, it must be substantiated with historical evidence. The contributions by Reis and Teixeira dos Santos tell us how Portugal joined both the pre-war gold standard as well as the system reconstructed after War World I. The papers reveal the contrast between the decisions to adopt the gold standard in 1854 and in 1931. The first of these was, as Reis argues, a pragmatic response to the confused Portuguese monetary situation of the mid-1850s; in contrast, in 1931 the decision to introduce convertibility was part of a grand policy design, as Teixeira dos Santos suggests. In both cases, the decision to join the gold standard interacted with politics. The explanations for Portugal’s double desertion of the gold standard are similar. The suspension of convertibility in 1891 came as a consequence of the international financial crisis of that year, and the abandonment of 1931 was also due to an international crisis, that generated by the suspension of the convertibility of sterling into gold. Reis’s paper addresses two separate although interrelated issues: Portugal’s early adoption of the gold standard and the so-called ‘silver problem’ (the scarcity of silver and the disorderly pre-1854 monetary regime). The latter question, common to many other European nations, forced the authorities to attract and legalize the circulation of foreign coins. However, this policy generated more problems than it resolved; agents faced high transaction costs and instead of a single official gold-silver ratio, there were several, depending on which pair of coins was being considered. A rapid solution to end the lack of silver would have been to raise its mint price to the market ratio when needed. But the authorities only hesitantly altered the official price. All this is explained in detail in Chapter 6. Reis offers an explanation for the striking fact of Portugal’s early adherence to the gold standard and explains in detail how the decision came about. He finds that some recent arguments for the introduction of an international monetary standard do not serve to account for Portugal’s behavior. First, it was not a rich and industrialized nation; second, no social group favored clearly the gold option; and thirdly, no ‘ideology of gold’ prevailed in the country. For Reis, the decision of 1854 ‘was largely a matter of historical accident’; it was a ‘pragmatic measure’ since by that year Portugal was de facto on gold and the authorities’ move made that a de jure situation. Although Reis’s argument (that Portugal’s adoption of the gold standard was a question of ‘practical nature’) is essentially correct, it overlooks certain fundamental economic facts that, in my view, help to explain the decision of 1854. One could easily argue that the introduction of the gold standard in Portugal was not an accident but rather the result of trends in the Portuguese balance of payments. While in this paper Reis tells us little about Portugal’s trade flows, we know that the country received large quantities of gold from Brazil. By the mid-1850s, the special political and commercial ties that Portugal had developed with Brazil could hardly be called an accident; on the contrary, Brazil played a significant role in the economic fabric of Portugal. A second

232 COMMENT

fundamental argument flows from the special trade connections with the United Kingdom, which went back to the eighteenth century and made Portugal a unique case in Europe. The fact that half of Portugal’s foreign trade went and came from the United Kingdom impinged on the overall performance of its economy. This was no historical accident. Since the United Kingdom was already on the gold standard, the correct policy for Portugal was also to adopt gold monometallism. Finally, I find unconvincing Reis’s argument that the question of the monetary regime was irrelevant to foreign investors. Rather than argue that ‘there seem to be no grounds for supposing that the adoption of the gold standard was motivated by a desire for easier access to the international capital market,’ the author should propose and answer the following question: had Portugal not adopted the gold standard, before or after 1854, would it have received as much foreign capital as it obtained from the United Kingdom? Only if the answer is affirmative, would I then agree with Reis. In Teixeira dos Santos’ paper we learn the details of the Portuguese currency experience after War World I. Gold convertibility was a major goal, as in other European countries. However, between 1920 and 1924 budgetary difficulties and political instability prevented any successful return to the gold standard. The situation improved in the following four years, when in particular the fluctuations of the escudo exchange rate were maintained within narrow margins. But not until the appointment of Salazar as Minister of Finance was a clear move towards financial stabilization made. In 1928 the authorities introduced a full program to balance the budget, reduced the floating debt and stabilized the exchange rate. Despite the obvious intention of the Portuguese authorities, de jure convertibility was not declared until July 1931. Although the author offers a convincing account of the problems experienced by the Portuguese economy during the 1920s, including the policy implemented at the end of the decade, he does not explain why it took three years (from 1928 to 1931) to introduce the gold standard. From Teixeira dos Santos one can deduce that Portugal was ready to go into convertibility well before 1931; nevertheless, the decision was postponed. I think that in order to understand this puzzling fact, we need to learn something about Portugal’s external accounts during the period, as well as something about the process of the adjustment of the domestic economy to changes in the balance of payments. What was driving the exchange rate before and after 1931? It is also curious that the authorities took the decision to join the gold standard when the European financial crisis unleashed by the collapse of the Kredit-Anstalt was already underway. By July 1931 news of the crisis would have been well known in Lisbon. Why did this not alter the government’s plan? A logical course of action would have been to adopt a ‘wait and see’ policy, postponing any definitive solution with regard to the currency regime. In addition, did the authorities discuss the benefits and costs of the gold standard versus a flexible escudo exchange rate, taking into account that the Depression was already two years old? Had the Bank of Portugal any commitment with other European central banks that forced it to introduce the

PORTUGUESE CURRENCY EXPERIENCE 233

gold standard? Was it rather a politically motivated decision taken without any consideration of the economic consequences? These are questions that should be answered. Although Portugal took a full decade after World War I before adopting the gold standard, its abandonment was only a question of a few months. In September, when the Bank of England suspended the convertibility of sterling, the Portuguese authorities decided to peg the escudo to the pound and later, in December, formally declared inconvertibility. Commercial relations with the United Kingdom and de facto suspension of payments from Brazil left the Portuguese authorities with no escape. Teixeira dos Santos explains that despite being off gold, the authorities remained committed to a stable currency. Thereby, when sterling depreciated the escudo was unpegged and linked to the dollar; later, after the devaluation of the dollar in 1933, the escudo was pegged to the franc. That was certainly a peculiar arrangement and it would be interesting to know about its impact on the real economy and the balance of payments. It would be also interesting to know about the monetary and fiscal policy followed by the Portuguese authorities after September 1931. Teixeira dos Santos leaves these questions unanswered, although I hope he will address them in a future paper.

234

COMMENT Gianni Toniolo

The papers by Reis, dos Santos, Mata and Valério can easily be read, and discussed, together: in fact, they seem to form the core of a comprehensive monetary history of Portugal. Or at least that is how I see them: my comments will be framed accordingly. From the authors’ accounts—particularly that by Mata and Valério—the monetary history of modern Portugal appears to be broadly divided into nine main periods. 1. The opening gambit or ‘first to join’ move. 2. The gold standard years (1854–91). 3. The 1891–92 crisis (which, however, I would argue should be seen as part of the previous period). 4. A series of attempts at price and exchange rate stabilization (1891–1918). 5. Inflation and devaluation (1919–28). 6. Stabilization (1928–31). 7. Pegging (1931–45). 8. First mimicking, then joining the Bretton Woods Agreements (1945–74). 9. Inflation and stabilization (1974–90). Reis and dos Santos deal with the two major pre-World War II episodes of exchange rate stabilization, while the other two authors take the long-run approach. Jaime Reis has produced a very clear and informative paper, to which I can add very little. He sees three main causes for the success of monetary stabilization and the introduction of (gold) convertibility in 1854: 1. a shift in world prices that made Portuguese bimetallism untenable, given the high legal ratio of silver to gold price, and the attendant realization (by 1853) that the problem would not disappear; 2. the fact that the problem of inconvertible paper money began to find a viable solution; 3. the virtuous role played by expectations once the monetary reform became a likely eventuality.

236 PORTUGUESE CURRENCY EXPERIENCE

These are convincing reasons: if the system stood in balance between bimetallism and gold standard, they certainly provided what was necessary to tip the scale in favor of the latter. The author also provides satisfactory explanations for why the particular English form of bullion standard was adopted. But was the adoption of the gold standard merely a technical decision? Jaime Reis seems to think so, but fails to fully convince the reader that this was the case. He argues that up to the 1840s the performance of the bimetallic system was not unsatisfactory: with parity adjustments (in 1835–41) it seemed to perform its monetary functions rather smoothly. In the early 1850s the problem of the silver/ gold market price became acute: why couldn’t this be solved again by altering the legal ratio, which was higher than that prevailing in other countries? After all, the Latin Monetary union was formed in 1865 and the heyday of international bimetallism would not be over until the early 1870s. In explaining the regime change that took place in 1854, the author rules out both the role of a non-existing new ‘urban-industrial class’ and that of a more general ‘gold ideology.’ He is probably right to do so. However, one would like to understand better the coincidence of the monetary reform with a changing political situation, namely the liberal turn taken by government in 1852 which, later, gave birth to rotativismo. In the short run, some kind of political stability seems to be one of the prerequisites to far-reaching monetary reforms and one wonders whether this was the case also in Portugal. Taking a longer-run perspective, one is struck by the fact that the gold standard was introduced precisely at a time when some kind of ‘modernization’ policies were gaining momentum (e.g. in the form of social overhead capital formation): these required foreign investment that, in the case of Portugal, was likely to come from Britain rather than from France. In this situation, the adoption of the gold standard seems to have been a rational decision, as it was later in the case of other ‘peripheral’ countries. For a peripheral developing country, thirty-seven years of convertibility (except for a brief, wise, moratorium in August 1876) are undoubtedly a remarkable success. According to Mata and Valério, state budget deficits explain both convertibility and its demise. They were small enough to be easily financed in the market, thus providing the flow of gold that was necessary to maintain convertibility. In the long run, however, they allowed the accumulation of a stock of debt too large to survive a confidence crisis. This may indeed be part of the explanation which, however, leaves out the crucial role played by emigrant remittances in creating a current account surplus that, in turn, reinforces the credibility of the commitment to convertibility. Indeed, emigrant remittances and foreign lending seem to be mutually related, in that both reinforced expectations about long-run exchange rate stability. Here we are confronted with a standard problem in the economic history of several countries during this period: was it wise to subject a weak but growing economy to the rigors of the gold standard? A positive answer to this question

COMMENT 237

depends crucially on the relevance of emigrant remittances and foreign lending in the financing of social overhead capital formation. The wisdom of sticking to gold might also be questioned in view of the longrun effects of the 1891–92 crisis, which crippled Portugal’s ability to borrow abroad. This crisis, part of a worldwide phenomenon, was a major shock to the country’s monetary system and economy. Political mismanagement fed into economic fragility. Portugal could not afford to quarrel over Africa with its main trading partner and provider of capital: Lord Salisbury’s ultimatum was not only a national humiliation but an economic disaster as well, with obvious implications for the exchange rate. However, if it was in the government’s power to avoid a diplomatic confrontation with London, it had no control over two other major events: the Barings crisis and Brazil’s change of political regime, the latter affecting expectations about emigrant remittances. These were both exogenous shocks which the Portuguese economy met in conditions of extreme fragility: here again one wonders if an early suspension of convertibility might have allowed its resumption after the end of the storm. Having said that, the net long-run effects of the crisis on the real economy are not clear. I shall return to this point. More generally, the whole experience of thirty-seven years on the gold standard must be judged also in the light of its outcome in 1891–92, an analysis that is still lacking in the papers under review. The paper by dos Santos tells another interesting story about an event not widely known outside Portugal, that deserves to be added to the list of case studies of post-World War I monetary stabilization. It is interesting both for the similarities with episodes in other countries and for its peculiarities. Here we have an account of what is probably the shortest-lived ‘goldexchange standard’: the yen’s convertibility was assured for less than a year, that of the escudo for only eighty-two days. The standard questions scholars have asked about the post-war instances of returning to the gold standard are: 1. What were its preconditions? 2. How do we account for the timing of the regime change? 3. What were its consequences for the real economy? In the case of Portugal the first two questions can be unified in asking: why was stabilization so long delayed? The part of the story that sounds familiar runs as follows: in Portugal, as in most other former belligerent countries, post-war inflation fed into exchange-rate depreciation and into new inflation. Exchange-rate depreciation reflected anticipations of changes in purchasing power parity. However, state budget deficits (which do not look colossal when seen in international perspective) were indeed drastically reduced without producing the expected results on inflation. This was apparently due to the reluctance of savers to re-subscribe maturing debt, which led to the latters monetization. It seems that the governments of the time put themselves in a most uncomfortable situation: they took the blame for

238 PORTUGUESE CURRENCY EXPERIENCE

socially painful policies without being able to claim success on the monetary front. Why so? Why weren’t interest rates on state bonds increased to a level where supply and demand would match? Or, alternatively, why was short-term debt not consolidated, as in Fascist Italy? The political economy implications of this situation are not entirely clear. What seems to appear from dos Santos’s account is that de Castro’s initial success was frustrated by the incompetence of the military after 1926. However, is there any link betweeen the ineffective financial policies of the democratic governments between 1924 and 1926 and the success of the 1926 military coup? The actual timing of the stabilization seems to be dictated by Salazar’s availability: he first refused and then, in 1928, accepted the finance portfolio. It must be noted, incidentally, that his conditions for taking the job in the summer of 1936 recall the requests made by Poincaré a few weeks later: full power to rule by decree in financial matters. Salazar’s fiscal and monetary policies were pretty effective (however, our curiosity about the actual measures taken to reduce the floating debt remains unsatisfied). Nevertheless, it must be noted that the timing of the resumption of convertibility was entirely dictated by domestic rather than international considerations (the paper does not mention negotiations for the customary international loan). The timing was also quite ill chosen. While the first moves to stabilize the escudo were made at the beginning of May, when few knew the truth about the Kredit Anstalt’s position, by July 1 not only had the latter crisis exploded into a major currency problem but it was clear that the situation of the major German banks was, to put it mildly, extremely delicate. Why didn’t Salazar revise his stabilization plans? Was he unaware of what was going on outside Portugal, or did he prefer to gamble rather than to go back on his word? The decision to follow the United Kingdom off gold is witness of Salazar’s wisdom and even courage. Management of this crisis was in sharp contrast with that of 1891–92. With the decision to peg to sterling, Portugal’s monetary history seems to have completed a full circle, getting back to where it had begun in 1854. Judging by the subsequent performance of the economy, the courage shown in September 1931 was duly rewarded. The long-run approach of Chapter 8, by Eugénia Mata and Nuno Valério, conveys one main message: Portugal’s monetary history shows that fiscal discipline and exchange rate stability go hand in hand. With some qualifications, few would disagree with this general statement. Disagreement may begin when discussing cause and effect. Mata and Valério’s point is that sound financial policies produce stable exchange rates. However, quite a few macroeconomists would argue the reverse: that the gold standard or any other commitment to external stability would provide the needed exogenous constraints capable of inducing fiscal discipline. This is, of course, not a trivial point, and it should be more carefully taken into account when discussing individual episodes in Portugal’s monetary history. The 1930s and the 1980s, in particular, do not offer obvious insight into the direction of causality.

COMMENT 239

While perhaps satisfactory for a macroeconomics textbook, an interpretation of one and one half centuries of monetary history entirely based on the fiscal prerequisite, although interesting, cannot entirely satisfy the economic historian. For instance, in Portugal, as in most other countries, periods of high inflation included World War I and its aftermath, World War II and the oil-shock years. The escudo seriously depreciated on foreign exchange markets in only two of them: 1914–24 and that following 1974. These, however, are cases in which exchange rate depreciation might have been caused by a host of factors (e.g. expectations about the outcome of the war) and when fiscal discipline might not have been desirable or, at any rate, possible. By focusing on the budget-exchange rate relation one is, at best, considering only the proximate, and possibly the least interesting, cause of inflation and depreciation. Mata and Valério’s paper is perhaps the most ambitious of the three discussed here. It addresses the question of the relation between monetary variables on the one hand and the performance of the real economy on the other. Needless to say, this is a bold attempt, since we hardly possess a theory on which we can safely rely when tackling such a crucial issue. However, the paper takes the general view that monetary stability is good for growth. The authors illustrate the point ex-post by observing that growth was most remarkable under the gold standard and after Salazar’s stabilization, while it was sluggish in 1891–1924 and during the oil shocks and their aftermath. Let’s briefly examine each of these periods in turn. As mentioned above, it may plausibly be argued that the gold standard was a blessing for the growth of a peripheral economy if it can be argued that it was an indispensable prerequisite for foreign lending that was put to profitable use in the real economy, e.g. by allowing faster social and private real capital formation. If we take a clue from Witte’s Russia, this is certainly a possibility and one that has been extensively discussed by economic historians. However, in order to prove this point, Mata and Valério should show that the available monetary alternative —‘forced’ savings through inflation—would not have produced the same results and/or would have been politically impracticable. Mata and Valério proceed by arguing that the inability or unwillingness to tax brought about the financial crisis of 1891–92, with dire results for the economy. However, the net effect of the crisis is not clearly assessed. In order to do so, one should weigh the subsequent inability to borrow against the advantages of a default which, by and large, would have boiled down to a free ride on foreign investors for part of the previously accumulated physical capital. If, again, we consider the lessons of subsequent Russian experience, we see that full service and repayment of the debt entailed levels of taxation, particularly on peasants, that were incompatible with economic incentives to produce. A crisis in the economy followed, triggering a financial crisis as well: their combined effect put an end to the so-called Witte experiment. In the case of Portugal, fiscal pressure increased considerably during the gold standard years and, by the end of the period, over 50 per cent of government revenue was allocated to debt servicing.

240 PORTUGUESE CURRENCY EXPERIENCE

It may very well be the case that these levels were the highest compatible with effective economic incentives and social stability. If so, partial default after cashing in the benefits of foreign investments was the best available alternative, at least in the short run. Between 1891 and 1924 Portugal experienced monetary instability and sluggish growth. Mata and Valério see the latter as being caused by the former. However, a glance at the history of the period makes political mismanagement a good candidate as a cause for both phenomena. The same social and political instability that made the resumption of convertibility impossible is likely to have been responsible for Portugal not sharing in the international prosperity during the ‘belle époque.’ The political record of the first republic is not one of firm leadership. The decision to enter World War I was a colossal mistake. Given the number of aborted pronunciamientos, it is not difficult to believe that the country might have looked unstable to foreign and domestic investors alike. Weak governments easily reach for the printing press, but blaming inflation for slow investment and growth is just considering the last ring in the chain of causation. The same point can be made about the 1970s. The oil shocks were bad enough for much more developed, large and self-sufficient economies than Portugal’s. On top of that, Portugal had to absorb the shocks of a revolution and a massive wave of immigration. No wonder output and prices were affected. If anything, the questions to be asked should concern the success in maintaining a stable exchange during the revolution and, more generally, in avoiding hyperinflation. For one who is admittedly rather unfamiliar with recent Portuguese economic history, it is not easy to understand why the authors of this paper fail to appreciate the resilience of the Portuguese economy under the formidable shocks of the 1970s. The potential for growth shown in the 1980s derived from the basic solidity of the society and the economy, a solidity that is, perhaps, easier to grasp in comparative terms from the foreigners detached viewpoint than from that of understandably emotionally involved Portuguese scholars.

Part IV IMPLICATIONS FOR EUROPE IN THE 1990s

242

9 CONVERGING TOWARDS A EUROPEAN CURRENCY STANDARD Convertibility and stability in the 1990s and beyond* Jorge Braga de Macedo

INTRODUCTION This chapter assesses the progress and prospects for European monetary unification in light of the historical experiences analyzed in the rest of this book. The removal of restrictions on capital movement and the stabilization of exchange rates have been goals of European policy ever since the adoption of the Single European Act in the 1980s. Indeed, the roots of Europe’s effort to create a stable, convertible currency system can be traced back to the early 1970s and the creation of the European Snake. The most recent and successful arrangement established in pursuit of this goal is the European Monetary System (EMS). The EMS, established in 1979, stabilized the exchange rates of its members within narrow bands subject to periodic realignments. The pace of monetary integration then accelerated in the second half of the 1980s. Starting in 1987 there were no further realignments of EMS currencies. At the Madrid European Council meeting in 1989, the report of the Committee of Central Bank Governors, chaired by European Commission President Jacques Delors, was accepted as a basis for economic and monetary union (EMU). The Delors Report urged the establishment of a single European currency and recommended completing the transition in three stages, with Stage I starting on July 1, 1990. The single currency was to be named after the European Currency Unit. We may thus refer to EMU as an ECU standard, even if the name of the currency is ‘euro’ instead. The blueprint for the transition to EMU was agreed upon at the Maastricht meeting of the European Council in 1991. According to the Treaty on European Union signed at Maastricht, which took much of its inspiration from the Delors Report, Stage II, when final preparations for economic and monetary union were to be made, was to commence on January 1, 1994. Stage III, full economic and monetary union, was to begin after the treaty was updated at the 1996 Intergovernmental Conference, although the precise date would depend on the pace of progress toward convergence and cohesion. Given the progress on meeting the inflation, exchange rate, interest rate and fiscal targets of the treaty, Stage III might be delayed beyond the end of the century. But the convergence

244 IMPLICATIONS FOR EUROPE

criteria of the treaty had built-in flexibility: they were to be interpreted in terms of medium-term policy outturns, rather than year-by-year variations in macroeconomic conditions. In the spring of 1992, when Portugal joined the EMS, all EC currencies except the Greek drachma were in the Exchange Rate Mechanism (ERM) parity grid. The EMS bloc effectively encompassed the members of the European Free Trade Association (EFTA), most of whom pegged their currencies to the ecu or the Deutschmark. It seemed as if the transition to monetary union was on course. But the link of the pound sterling with these arrangements was weak. The dollars depreciation in 1991–92 aggravated the competitive problems of European countries and intensified strains within the emerging ECU bloc. In August and September of 1992, crisis erupted. Political instability interacted with the most severe recession and highest unemployment the EC had ever experienced to undermine the credibility of the Community’s commitment to establish an ECU standard. The British and the Italian currencies were attacked and forced out of the EMS. Between November 1992 and May 1993, Ireland, Spain and Portugal were all subjected to speculative attacks and had to realign. Attacks on the French franc and other ERM currencies made it impossible to defend the narrow bands of the parity grid. ERM fluctuation bands were widened from 2¼ to 15 per cent on August 2, 1993, in principle allowing greater exchange rate fluctuations around parities than ever before. This experience might be thought of as throwing cold water on the prospects for an ECU standard in the 1990s and beyond. But a perspective grounded in the history of the international monetary system suggests that prospects for EMU may not be so dim. In the next section, I discuss lessons from the history of the gold standard and convertible currency regimes generally that place in perspective the outlook for EMU. The point is that geography is associated with a financial market hierarchy. Financial markets’ perceptions of geography and hierarchy only change as policies at the periphery are sustained and become credible in the medium term. The major difference between an ECU standard as envisaged in the Union Treaty and the implications from history has to do with the willingness and the ability of union states to engage in multilateral surveillance procedures which involve political peer pressure about policy rules and may therefore reassure market sentiment. These procedures are to be reinforced during the current stage of EMU. Moreover they have been supported by all member states, including those —such as Britain—which are not committed to entering Stage III. The current emphasis on the excessive deficit procedure may therefore bring about a positive, rather than negative, interaction between geography and credibility. The lessons are consistent with the realignment of the peseta and the escudo in March 1995. In the last section I focus on the convertibility and stability decisions taken by Portugal in 1854, 1931 and 1992. The chapter ends with a conclusion stressing

EUROPEAN CURRENCY CONVERGENCE 245

the public good element in international monetary standards, including the ECU standard. IMPLICATIONS FROM HISTORY A first implication of history derives from the geography of the international monetary system. Substantial regional variations have always characterized the operation of international monetary arrangements. In Chapter 2, Michael Bordo and Anna Schwartz provide evidence to this effect for twenty-one countries spanning 110 years. They define the core as including the present-day G-10 (today’s seven largest industrial countries along with Belgium, the Netherlands and Sweden), plus Switzerland. In the periphery they place Latin America, Australia, Denmark, Greece, Finland, Norway, Spain and Portugal. They find that the gold standard (and its post-World War II successor, the Bretton Woods gold-dollar standard) functioned more adequately at the core, with the reserve currencies on each side of the North Atlantic. Therefore, Atlantic currency standards prevailed since the 1870s, except during the two world wars, and the recent floating exchange rate system. The gold standard featured several distinct currency areas, each with its own center and corresponding periphery: might one similarly expect to see the development of an EMU core and periphery? This question is clearly related to the debate about whether economic and monetary union should proceed at one or several speeds. In the case of monetary union, a variable geometry is probably inevitable. Even among the present members of the European Union, substantial degrees of economic divergence exist and are likely to persist for some time. As the EU is expanded to encompass Eastern Europe, divergence will be magnified. Some EU member states will have to drastically reform their economic, financial and fiscal policies in order to achieve the requisite convergence. Not all will succeed. Indeed, not all may try. Linked to the geography of the single European currency is the hierarchy of financial markets. Where local financial monopolies exist, differences between interest rates at the core and at the periphery may endure, even in the presence of full currency convertibility and perfect capital mobility among core markets. Belonging to the convertibility and stability club is nevertheless useful, to the extent it signals to market participants that the country is keen on achieving external credibility without relying only on instruments it could control—and might therefore manipulate. In this sense therefore it is also attempting to buy domestic credibility for its efforts. This is the only way in which the national authorities could escape the adverse selection bias from which new participants in the internal capital markets have been shown to suffer.1 This is why the notion of medium-term policy credibility helps in the evaluation of how the regime in the Treaty combines convergence and cohesion. The European construction foresees an EU characterized not just by the convergence of economic policies but of increasing ‘cohesion’—declining

246 IMPLICATIONS FOR EUROPE

disparities in levels of economic development and national income across member states. But even with convergence of economic policies, considerable time may still be required to bring about the desired level of cohesion. This is evident in the regional disparities that persist within EU member states. In Spain, half the population lives in regions as poor as the Greek or Portuguese national average, the other half in regions where incomes approach Northern European levels, In Italy and Germany, dramatic disparities run along North-South and East-West lines respectively. None of these disparities is eroding rapidly under the pressure of market forces. Looking further afield toward Eastern Europe and the southern shores of the Mediterranean makes the mix of national and regional economic structures and outcomes all the more diverse. This suggests that the historical model of several distinct currency areas or competing arrangements is likely to apply to the Europe of the future as well. A second implication of history concerns fiscal and monetary policy rules. Bordo and Schwartz also emphasize the role that fixed rates can play in enhancing the credibility of policy and the effectiveness of external adjustment by national economies. They argue that well-defined contingencies under which escape clauses could be invoked were critical to the viability of fixed exchange rate rules. They were an integral feature of all exchange rate rules that functioned smoothly in the past.2 But EMU will include no such escape clauses, for it entails an irrevocable commitment to a single currency. It will therefore require greater cohesion than its predecessors in order to avoid systemic instability. Policy rules, in turn, raise the issue of politics and market institutions. The chapters in Part II emphasize the importance of politics in the origin and operation of the nineteenth-century gold standard. The rise of the gold standard, Alan Milward argues in Chapter 3, was a corollary of the constitutional rule of the middle class. The gradual displacement during the second half of the century of the bimetallic system proposed by France is based on the perception of the gold standard as a dynamic system, rather than as one promoting convergence at the expense of cohesion and real economic growth. The way in which German public opinion changed in the 1870s from favoring silver to favoring gold is illustrative of the current debate about the role of the Deutschmark in an ECU standard.3 Cultivating popular support will also be important for the success of the EMU project, as the Danish and French referenda on the Treaty made painfully evident in 1992. Courting such support will require winning the confidence not just of the public but also of the markets. Marcello de Cecco, in Chapter 4, highlights the applicability of this insight to the nineteenth-century gold standard. Few current-day European policy-makers would deny its relevance in the wake of the 1992–93 crisis. In Chapter 5 Barry Eichengreen and Marc Flandreau also show the political element behind the geography of the gold standard. The institutional responses to speculative attacks on the French franc and the Benelux currencies differed from those on more peripheral currencies in the EMS.

EUROPEAN CURRENCY CONVERGENCE 247

THE GEOGRAPHY OF THE ECU STANDARD In this section I review the issues that the European Union will have toaddress if it is to complete the transition to monetary union. I do so fromthe geographical perspective suggested by this volumes historical studies ofthe international monetary system. Prerequisites for convergence Completing Europe’s internal market is a prerequisite for creating the currency union called for by the Treaty. This connection was emphasized by the European Commission in its report One Market, One Money (European Commission, 1990). Achieving free trade in goods, services and financial assets, along with free movements of people, was the primary objective of the Single European Act. The abolition of internal borders created a desire for stable exchange rates and therefore for a single currency. But the single currency, if it is to become the monetary standard of the European Union, must be stable. And monetary stability requires stable and consistent fiscal and financial policies on the part of member states. The treaty therefore requires EU member states to reduce public sector debts and deficits before the single currency is established. Political stability and social consensus are essential to the successful completion of this task. Consensus requires that the social partners and the public understand and accept the stance of policy. Price stability will not be durable in the absence of wage moderation. Trade unions must recognize this fact and that the poor and unemployed are disproportionately hurt by the destabilizing interaction of wage and price increases. The government can encourage distributional moderation by taking the lead in wage negotiations with public sector employees and by facilitating cooperation between the social partners. Article B of the treaty therefore refers to ‘strengthening economic and social cohesion’ as an instrument of ‘economic and social progress which is balanced and sustainable.’ But political stability and social consensus are required at the level of the EU as well as that of the member state. This points to the need for income redistribution across states to cement the political and social solidarity that is a prerequisite for EMU. In European Commission (1993) it is shown that, contrary to widespread belief, significant redistribution could be undertaken without increasing the size of the EU budget, so long as a program of intra-EU transfers is tailored to the purpose. That scheme for increasing cohesion presupposes effective procedures of multilateral surveillance will be developed to prevent member states from abusing the transfer mechanism.4

248 IMPLICATIONS FOR EUROPE

The state of convergence The convergence criteria of the Treaty set inflation and interest rate ceilings in terms of deviations from the inflation and interest rates of the three best performers in the European Union. Consumer price inflation rate must be within 1.5 per cent of that of the three best performers, long-term interest rates within 2 per cent. In 1994 this implied inflation rates of about 3 per cent and long-term interest rates of 7 per cent. A protocol to the treaty specifies reference values for government budget deficits and public debts of 3 and 60 per cent of GDP, respectively. Member states’ conformance with these guidelines is to be monitored at the national and EU levels. These convergence criteria are demanding under the best of circumstances. But they became especially difficult to meet in 1992–93 as Europe descended into the most severe recession of the post-war period. Negative output growth increased public expenditure relative to taxes, raising the debts and deficits of most member states above their reference values. The union averages for deficits and debts in 1993 were 6 and 66 per cent, respectively. Table 9.1 presents the convergence performances and divides them into three categories, according to whether convergence was high, medium or low in 1994. Only the four high-convergence states fulfilled the deficit criterion. The lowconvergence member states, except for Sweden, failed the interest rate criterion. Of the six medium-convergence states, three failed at least two criteria while three failed one. Table 9.1 suggests that it is inappropriate to group together the low income member states. Greece is an outlier because its inflation rate remains in double digits. Spain and Portugal, in contrast, are much closer in terms of inflation, although their growth and employment performance continues to differ. Spain has a higher unemployment rate and more pronounced regional disparities than Portugal. Portugal’s reputation for financial stability is more recent than that of Spain, on the other hand. Instability in Stage II of EMU Against this background Europe embarked on the three-stage transition sketched in the Treaty. The outlook for economic growth was favorable at the outset of Stage I. In 1990 there was no reason to anticipate the financial turmoil and recession that would follow. Indeed, the forecast errors made when the European authorities projected growth rates for 1992 and 1993 were among the highest on record. To be sure, the demise of the Soviet Union was a shock to the European economy; the uncertainty this event created could hardly have been anticipated. Nor were the economic and financial consequences of German unification adequately appreciated at the time. The severity of the 1992–93 recession was unprecedented in the history of the European Community. For all these reasons,

EUROPEAN CURRENCY CONVERGENCE 249

Table 9.1 Convergence criteria

Source: UBS Group Economic Research, Political Focus, March 1994 * EC Update, spring 1994 (for EU 12 only). Note: Within each one of the three categories, countries are ranked by alphabetical order in home language. Shaded cells: Convergence criterion not satisfied.

few if any observers anticipated the events that would strain the second half of Stage I.

250 IMPLICATIONS FOR EUROPE

Each of the events described in the preceding paragraph played a role in the spectacular foreign exchange crises of the summer of 1992. The convergence criteria of the treaty require countries to hold their exchange rates within the ‘normal fluctuation margins’ of the EMS for at least two years prior to the inauguration of monetary union. National central banks are vested with carrying out this responsibility. But their ability to do so is dependent on economic and political conditions. It can be helped or hindered by accommodating adjustments in policy by Germany, the country that issues the anchor currency of the EMS. It depends on the relationship of the EMS to the major currencies of the nonEuropean world such as the dollar and the yen. In the final analysis, the behavior of the exchange rate reflects the credibility of domestic policy. Unfortunately, credibility signals may be difficult to discern in practice. Financial markets tend to amplify rather than dampen noise. This is one interpretation of the situation in 1992–93, when the entire EMS parity grid came under attack. The markets had trouble distinguishing credible from incredible policies. Thus, the speculative attacks on the Irish punt and Portuguese escudo may have had less to do with the credibility and medium-term resolve of the national authorities in those countries than with turbulence in larger neighboring markets, like those for sterling and the peseta. Developments in large markets have a tendency to spill over to smaller neighboring ones. The attacks suffered by Ireland and Portugal can thus be described in terms of ‘geographic’ rather than ‘economic’ fundamentals.5 If this view is correct and there can be contagion within the EMS, then the intermediate stages of the transition to monetary union may be inherently unstable. Countries running the ‘right’ policies may nonetheless lack immunity from destabilizing shocks. Indeed, the scope for contagious speculative attacks may be greatest when the authorities attempt to lock in the prevailing grid of exchange rates in the final phases of Stage II. Under these circumstances, speculators will be offered a one-way bet (currencies can be significantly devalued but not revalued), magnifying their incentive to challenge the resolve of authorities of neighboring countries. With sufficient exchange rate flexibility, in contrast, speculators are no longer offered a one-way bet. This lesson was powerfully imparted to European policymakers in 1992–93. That it was taken to heart became evident on August 2, 1993, when the normal fluctuation margins of 2¼ per cent were widened to 15 per cent. The question is whether European policy-makers will wish to restore narrow bands at some future date, and if so, whether this would reopen the scope for instability like that experienced in 1992–93. Financial markets Over and above the instability of the transition to EMU, a single European currency may aggravate the financial weakness of local business firms and hinder their global competitiveness. This is more likely to happen if banking

EUROPEAN CURRENCY CONVERGENCE 251

competition, transparency of interest rates and consumer protection are not sufficient to give European firms competitive access to credit, European depositors competitive reward to savings, or both. As the macroeconomic benefits of a fixed exchange rate materialize first in financial centers, these effects are more likely at the periphery of the union. Therefore, appropriate structural policies must succeed in combining convergence with cohesion to avoid a loss of competitiveness of European business, taking into account the hierarchy of financial markets. Aside from the spatial effects derived from economic distance, there are unfavorable effects for the periphery due to the time,profile of policy changes. The need for financial stability may imply initial costs to the cohesion objective which are more than offset by subsequent benefits. The question is whether reforms will be reversed by the voting patterns. The acquisition of a reputation for financial stability, by anticipating some of the benefits, may be the best way to overcome this unfavorable time sequence. An independent central bank is a reputation-building signal, even though it does not by itself connect the local financial market to the center. Effective prudential supervision of the financial system, together with other structural policies, such as the defenses of competition and consumer protection, are also required to minimize the initial costs of nominal convergence. Since these policies will lower the difference between long-term interest rates prevailing at home and in the financial markets at the center, their success implies what we call ‘rental moderation.’ They are therefore essential to ensure medium-term policy credibility at national and union levels, especially if average interest rates are higher in the Union than in the United States or Japan, as has been the case in the 1990s. The issue of democratic accountability raised by the independent central banks at the national and union levels before the third stage of EMU may also be found in connection with independent courts in the framework of the cooperation in matters of justice and internal affairs envisaged in Title VI of the Treaty. The relation between the two in matters of money laundering is clear enough. Insider trading and banking supervision suggest another topic where the accountability of central bankers and judges may interact, but where the Treaty does not espouse any particular solution at the Union level. In the German model, banking supervision does not belong at the central bank because it could endanger its independence with respect to monetary and exchange rate policy. The choice remains open in public discussion both within and outside the Union, and especially in the United States and Japan. Recent changes in central bank statutes adopted in France and Spain suggest that the socalled ‘national superbank’ model—including monetary and exchange rate policy as well as banking supervision—remains attractive, in spite of the weight of the Bundesbank tradition. Reflecting the hierarchy of financial markets, the lack of accountability on the part of regulatory institutions explains why reputation-building at the periphery

252 IMPLICATIONS FOR EUROPE

tends to be slower than at the center. It is in the periphery, however, that the acquisition of a reputation for financial stability is most required to sustain a regime change in the direction of price stability. Moreover, in nations whose currencies are inconvertible, this hierarchy may also affect the foreign exchange market and slow down the dismantling of exchange controls, thereby insulating the local market from the single market in financial services existing since at least January 1, 1993. The volatility of interest rate differentials shows that financial development and trade in assets tends to magnify the forces of real convergence, or the widespread expectation thereof, but that they are no substitute for economic development and trade in goods and services, especially in local markets not well connected to the center due to controls or inadequate information. Deepening and widening Inevitably, the transition to a single currency will involve a process of simultaneous deepening and widening of the European Union. There is reason to hope that these dimensions will not be incompatible. Perfecting the Single Market promises to create additional trade, which will help to offset the scale diseconomies that have traditionally hampered growth in Europe’s smaller and more peripheral regions, and cohesion throughout the EU. The increased labor mobility associated with the Single Market will speed wage convergence and, by inference, economic convergence generally. Indeed, there has been evidence of wage convergence during Stage I, in spite of recession and turmoil in the foreign exchange markets. Still, labor mobility across Europe’s internal borders remains limited, by the standards of continental economies like the United States. This results in concentrations of unemployment that threaten economic and social cohesion within and between EU member states. It demands renewed efforts at the Union level to secure economic and social cohesion. To the extent that nations forgo the use of autonomous national monetary and exchange rate policies in Stages II and III of the EMU process, direct transfers become the only way of achieving this goal. Cohesion has been a long-standing goal of EU policy. It has been politicized since the reform of the EC’s Structural Funds in 1988. The question of whether to further expand the structural funds reappeared during the enlargement negotiations of 1994, when Spain linked the admission of Austria, Finland, Norway and Sweden to additional transfers. The complexity and delicacy of these negotiations was magnified by the fact that they were superimposed on the difficulties of ratifying the Treaty and on the European recession which heightened unemployment. The Treaty seeks to make intra-EU redistribution enduring. It provides for intra-EU redistribution directed to nations rather than regions, as was the case previously under the structural funds. The protocol on economic and social

EUROPEAN CURRENCY CONVERGENCE 253

cohesion specifies that interventions under the Cohesion Fund are to be reserved for member states with incomes per head less than 90 per cent of the union average (thereby excluding low-income regions within high-income member states). It also makes access to the Cohesion Fund conditional upon a country’s fulfillment of the convergence criteria for monetary union, to which the next section turns. MANAGING THE TRANSITION In this section I consider how the issues raised in the previous section should be managed by policy-makers. Cooperative responses in the monetary and exchange rate field can be elicited by progress in political areas of European integration, especially in the run-up to the Intergovernmental Conference of 1996, as long as they allow peer pressure and variable geometry to work towards an enduring ECU standard. Exchange rate stability and surveillance Multilateral surveillance procedures designed to ensure the convergence of national economies toward price stability and sound public finances become operational during Stage II. The ‘excessive deficits procedure’ is used to determine whether or not member states qualify for EMU. Based on this procedure and on the convergence achieved by a majority of member states, then the Council of Finance Ministers, the ‘ECOFIN Council’, decided Stage III will begin in 1999. The forthcoming revision of the Union Treaty may still modify this schedule, allowing the calendar for Stage III to slip into the twenty-first century. The recession which afflicted the European economies in the first half of the 1990s is thought to have made such an outcome more likely by increasing the difficulty of meeting the convergence criteria. So is the foreign exchange market turmoil that erupted in 1992. That upheaval began in August, when interest rates in the United States fell substantially. German interest rates remained high, leading to flight into the Deutchmark, at the expense of other ERM currencies. Pressures for wage increases within Germany heightened the reluctance of the Bundesbank to acknowledge that a Europe-wide recession was imminent. This policy imbalance forced the realignment of the Italian lira and the Spanish peseta and then the exit of the lira and the pound sterling from the ERM. Turbulence persisted until August of 1993. The Spanish authorities requested two additional realignments of the peseta, each of which also involved the Portuguese escudo. There were speculative attacks on the Irish punt, the French franc, the Belgian franc and the Danish krona. The July 1993 attack prompted an emergency meeting of the ECOFIN Council including central bank governors. Officials were left with no alternative to a widening of ERM fluctuation margins from 2¼ per cent to 15 per cent.

254 IMPLICATIONS FOR EUROPE

The 15 per cent band has not been fully utilized by most central banks. (Political instability in conjunction with high unemployment did, however, cause the Spanish peseta to fall to the bottom of its wide band in the winter of 1993–94 and a realignment to take place in March 1995.) For most countries, then, the new system has successfully eliminated the problem of one-way bets without condemning them to significantly greater exchange rate instability. However, the external discipline provided by the narrow-band system is now absent, and each central bank must decide unilaterally whether to intervene to hold its currency within the old fluctuation bands. With the passage of time, the possibility of a return to the old narrowband system, perhaps through a gradual narrowing of fluctuation margins, that was a widespread presumption in August, has become less likely. That wider bands have become an economic and political fact of life is a mixed blessing. The reluctance to restore narrow bands raises the danger of a delay in the calendar agreed to at Maastricht if countries like Germany resist the idea of jumping directly from wide bands to monetary union. Skeptics like the German Constitutional Court will have to be reassured by a rigorous interpretation of the convergence criteria. To be effective, though, multilateral surveillance must build on the EMS code of conduct as it has emerged over the years. But the EMS is not an end in itself. It is merely an instrument of convergence towards the single currency, which must be flanked by other instruments. The Treaty provides these instruments and its implementation has already started. The framework for economic surveillance is consistent with the ‘excessive deficit procedure’ and calls for the adoption of appropriate national budgetary procedure. Convergence programs were of course helpful in making fiscal issues more visible. At the same time, progress towards independence of national central banks in many countries is impressive, as is the fact that the public sector can no longer be financed by central banks or by privileged access to financial institutions. Signaling the change in regime Enhancing price and exchange rate stability and buttressing the soundness of public finances is a formidable task at the fringes of the EU. In countries with histories of high inflation, neither the social partners nor public employees automatically appreciate the benefits of the regime change that the policy-makers are attempting to engineer. Because information about the change in regime is not readily available to international financial markets, errors in policy appraisal can unduly raise the costs of reform. Repeated market tests of the authorities’ commitment to exchange rate stability may result. If these tests of the authorities’ resolve greatly increase the cost of defending the exchange rate, they can lead to policy reversals. This is one reason why Article 102 of the Union Treaty instructs the European Council to issue broad policy guidelines against which policy and performance in

EUROPEAN CURRENCY CONVERGENCE 255

the member states are to be gauged. With such guidelines at hand, the markets will be better able to appraise the progress of policy reform.6 How long will it take for the public and the markets to recognize that the policy regime has changed? Ten years is often taken as the time needed for a nation to acquire a reputation for financial probity. This answer suggests that it may be better to take time and set on foot a self-reinforcing process of reform than to attempt a succession of overly ambitious and excessively drastic measures that will ultimately fail, damaging policy credibility. Credible signals that the authorities are committed to reform may be needed to construct a social consensus domestically. One such signal is the provision in the Treaty directing member states to buttress the independence of their central banks.7 Bringing national bank statutes in line with the Treaty serves to demonstrate that the authorities are prepared to resist the temptation of excessive policy activism. It should thereby induce supportive reactions on the part of the private sector. Initial and terminal conditions Deepening and widening will facilitate cohesion only if there is a change in regime in poorer nations and regions. Such a change presupposes structural reforms which involve substantial redistributions of income and wealth among social groups. If stable democratic governments succeed in implementing reforms which sustain this process, helping to achieve convergence between poorer and richer member states, they can set off a self-reinforcing virtuous cycle of stability and growth. On the other hand, there will be a vicious cycle if shortlived governments, fearing the social conflicts associated with reforms, delay implementation and impair convergence. The initial conditions a government inherits may limit the alternatives at its command. For example, the 1992–93 recession aggravated the plight of Europe’s unemployed, making it more difficult to reduce the generosity of unemployment benefits. At the same time, by demonstrating the costs of labor market rigidities and the importance of competitiveness at the firm level, the experience of the 1992–93 recession may have actually encouraged structural adjustment and, ultimately, cohesion. If reforms are enduring, employment-generating growth will be stronger. This will brighten the prospects for completing the transition to EMU, and the Union’s commitment to stable exchange rates will gain credibility. This terminal condition, and the currency stability it delivers, will then feed back to a more employment-friendly economic environment, in a virtuous cycle.8 Conversely, if terminal conditions lack credibility, this may produce a ‘stopgo’ convergence process that hinders change. Temporary, unaccountable shifts in sentiment in financial markets thus may disrupt the convergence process permanently. A government can only protect itself from this threat by acquiring a reputation for subordinating other goals of economic policy to the pursuit of

256 IMPLICATIONS FOR EUROPE

convergence. Here again, multilateral surveillance can play a role in providing timely information on national economic policies in a way that enhances the reputations of deserving governments. THE CASE OF PORTUGAL In this section I first use the experience of Portugal to place the preceding points in historical perspective. Three convertibility and stability decisions—in 1854 by the parliament, as discussed in Chapter 6 by Jaime Reis, in 1931 by the government, as discussed in Chapter 7 by Fernando Teixeira dos Santos, and in 1992 by the central bank—are identified and interpreted. The 1989–92 regime change is described next. A comparison with Stage I of EMU in Greece, Ireland and Spain is then provided and evaluated. Currency experience: inflation and fiscal discipline Since Portugal first joined the gold standard, the currency has been stable relative to the dollar or the pound sterling four-fifths of the time. In a 1980 study I showed that episodes of exchange rate flexibility were associated with ‘troubled times like the aftermath of the 1891 crisis, the hyperinflation of the early 1920s and the present difficulties.’9 Threats to convertibility and stability originated in political and social instability. The prevalence of civil strife is evident in the fact that Portugal has experienced four successful revolutions in the twentieth century: in 1910, 1917, 1926 and 1974. As I show elsewhere, the budget deficit was only higher than the 3 per cent norm during troubled times like the 1914–24 period (when it reached 6 per cent) and after 1974 (8 per cent).10 The papers in Part III of this book are consistent with my earlier claim that currency convertibility and fiscal discipline were required for economic development. When convertibility was suspended in 1891, Portugal suffered a decline in its growth rate. This occurred even though suspension did not give rise to fiscal instability. In fact, fiscal policy was tightened in an attempt to buttress credibility and restore convertibility at an early date. Why then, if it was not associated with fiscal instability, did the suspension of convertibility hinder growth? For one thing, tighter fiscal policies created domestic adjustment problems. For another, the suspension of convertibility delayed the modernization of central bank operating procedures and hindered the development of a modern financial system. The experience of Portugal from the 1830s to the 1990s was marked by three attempts to join the convertibility club. On each occasion the timing of Portugal’s attempt differed from that of its neighbors. The 1854 establishment of convertibility preceded that of other European countries, aside from the United Kingdom. It should be understood in terms of the importance of British trade and finance to the Portuguese economy. In contrast, the establishment of convertibility in both 1931 and 1992 lagged that of other industrial countries.11

EUROPEAN CURRENCY CONVERGENCE 257

Table 9.2 Phases of Portuguese currency experience and gradual regime changes towards convertibility

Aside from the turbulent period before joining the gold standard in 1854, there were two reversions into instability, lasting from 1915 to 1924 and from 1974 to 1989. The three regime changes could be dated anytime between the beginning of the period of stability and the full operation of the new regime, that is between 1851 and 1855, between 1924 and 1932, and between 1989 and 1993. When the change in regime is gradual, however, policy credibility may not be established in as clear a way as otherwise. This reinforces the importance of political decisions in the face of historical contingencies.12 It also points to the effects of global disturbances on individual country decisions. Both in 1931 and in 1992, the international monetary system was about to enter into a turbulent phase when Portugal joined. In the latter case, the consequences of the 1992–93 recession are still being worked out. Even though the rationalization of the choice of stability is more difficult when the environment is unstable, the implication is the same: a regime change is a necessary but not sufficient condition for stability. It may be most appropriate to describe the 1854 decision as the result of a gradual regime change because the transition from bimetallism to the sterling standard was widely debated. In distinction, the link between the 1924 stabilization and the 1931 move was interrupted by a military coup. As for the 1992 decision, it was not perceived as part of a multi-term strategy, since a derogation had been granted until at least 1995. The phases reported in Table 9.2 identify the three gradual regime changes towards convertibility.13 Based on these dates, one can divide the century and a half into phases when the escudo was stable, unstable, convertible, and inconvertible. In Table 9.3, the results are expressed as a two-by-two matrix whose elements add up to one. Diagonal combinations were observed one quarter of the time each. The combination of inconvertibility and instability includes the phase before joining the gold standard, when data were particularly unreliable. Otherwise it would be observed significantly less frequently than the virtuous combination of the gold

258 IMPLICATIONS FOR EUROPE

Table 9.3 Relative duration of different currency regimes, 1834–1993

standard. The off-diagonal combinations are quite different. Convertibility and instability can be neglected, whereas the combination of inconvertibility and stability was recorded during half of the sample period. If the phases prior to 1854 are left out, the preference for stability rises to This suggests that the simple association between convertibility and stability is far from complete. In the Portuguese case, these complications may be attributable to politics. For example, convertibility may not have been a sufficiently safe option for the authoritarian political regimes that ruled Portugal for much of this period. Had the return to international borrowing which began in 1963 been followed by a gradual move to convertibility, as might have been possible given the country’s comfortable balance-of-payments position, then the figure in the upper-left cell of Table 9.3 would rise from one-fourth to nearly two-fifths, more in line with other European countries. However, Portugal’s financial system was heavily protected from the foreign competition which would have been associated with a convertible currency. Financiers’ political leverage was sufficient to prevent the country’s 1963 return to international capital markets from being accompanied by the liberalization of trade in financial services like that undertaken by other countries. In a more democratic regime, the influence of financial interests could have been counter-balanced by that of firms and consumers, and the outcome might have been different. Since the restoration of democracy led to high budget deficits, convertibility was only possible over one hundred years after the exit from the gold standard and over ten years during which the exchange rate accommodated inflation differentials with the Union. Again stability, rather than convertibility, was the most visible policy objective. The implications of this pattern of currency convertibility and exchange rate stability for Portugal’s monetary and fiscal performance was that Portugal’s fiscal performance consistently lagged behind that of the core of industrial countries, while comparing favorably to that of the Latin American periphery. In contrast, Portugal’s inflation performance is consistently superior, confirming the difficulty of definitively placing the country with either the periphery or the core. The 1989–92 regime change and its aftermath I now reconsider Portugal’s most recent attempt to restore currency convertibility in the light of this historical experience.

EUROPEAN CURRENCY CONVERGENCE 259

The most recent transition to stability and convertibility began with the lifting of the constitutional ban on privatization in the summer of 1989.14 Foreign capital began to flow into the Portuguese economy. Capital controls were then tightened to discourage excessive capital inflows, attracted by high rates of interest on public debt. Besides containing inflationary pressures, this reinforcement of financial protection was designed to give the banks time to prepare for the increased competition that was expected to result from the completion of the Single Market in financial services. With the brisk pace of economic growth and the country’s relatively low level of unemployment, however, there was strong upward pressure on wages, which the government aggravated by reforming the public sector wage determination scheme. Public expenditure also increased, and budget deficits rose. Inflation became a threat. The only remaining tool of inflation containment was the exchange rate. Portugal began by shadowing the Deutschmark, prior to the inauguration of Stage I. In October 1990 new statutes were adopted enhancing the independence and supervisory powers of the Bank of Portugal. Meanwhile, a program of budget consolidation, called Quantum, was put in place in 1990. The financial system remained protected, however, and inflation was still in double digits when the government won re-election in October 1991. In line with the procedure of multilateral surveillance adopted for Italy, a convergence program, dubbed Q2, was examined by ECOFIN at its December 1991 meeting. The measures listed in Q2 were then introduced in the budget approved in March 1992: the 1992 budget eliminated all money financing of deficits and broadened the tax base by eliminating the zero rate of VAT.15 Entry into the EMS was decided a few days after the budget was passed by Parliament. The rate was defined in terms of ECUs (1 ecu=180 escudos), slightly stronger than that which had prevailed previously. By buttressing credibility it was designed to allow for a lessening of capital controls and to deliver a reduction in interest rates. But the Bank of Portugal remained reluctant to dismantle controls. Hence, the new currency did not become convertible until after conditions in European financial markets had become unsettled. Differences between the government and the central bank were most apparent in banking supervision. The government wanted more effective monitoring of competition among banks to ensure the adoption of transparent policies concerning the cost of credit to small and medium-sized enterprises. The central bank, in contrast, was used to concentrating on prudential supervision narrowly defined. Echoing the gold-standard experiences described by de Cecco in Chapter 4 of this book, leaks to the financial press exaggerated the differences between the government and the central bank. In the unsettled environment of 1992–93, even minor disagreements threatened to excite the markets. Nevertheless, Portugal’s global bond issue in dollars, launched in September 1993 after successful yen and Deutschmark notes, was well received. It consolidated the upgrading of the state’s foreign debt from A+ to AA−, the first

260 IMPLICATIONS FOR EUROPE

such move since Ireland was upgraded in 1989. A global bond issue in ecus followed in early 1994. Learning from the wage hikes of 1991, Q2 stressed the leadership role of the public sector in wage negotiations. Yet moderation remained elusive. The unions, enjoying low unemployment, exploited their market power. The banks, not yet fully exposed to foreign competition, hesitated to reduce interest rates as desired by small firms and prospective home-owners. The failure to agree on an incomes policy was serious in light of the looming European recession. The damage was aggravated by changes in tax collection procedures due to the Single Market and by increases in the mobility of the tax base, which temporarily reduced receipts. Even though the nominal expenditure ceiling of Q2 was met, a supplementary budget was required as a result of the revenue shortfall. While the deficit was initially forecast to be as large as 9 per cent of GDP (and was still reported higher than 7 per cent in the 1995 budget), it turned out to be roughly the same as in 1991 and 1994. In any event, that Portugal credibly remained on its convergence path was evident in the fact that Brussels approved a Revised Convergence Program shortly before the 1994 budget was voted by parliament. Portuguese progress in comparative perspective How does Portugal’s experience compare with that of the other ‘cohesion states’: Spain, Ireland and Greece? Of the four, Spain stands out for its size: its population is twice that of the other three put together. While per capita income in Spain is higher, on average its variation by region is also more pronounced than in Greece or Portugal.16 The fact that the development challenges faced by Greece, Ireland and Portugal are national rather than regional only makes the interaction of convergence and cohesion more salient politically. In what follows I describe the performance of each of the cohesion states during the first stage of EMU (July 1, 1990 to December 31, 1993) and, using Commission forecasts for 1994–95, during the first two years of Stage II. The first column of Table 9.4 reports GDP per capita (adjusted for purchasing power parities) as a percentage of the EU average. On this basis real convergence looks stronger in Ireland and Portugal than in Greece and Spain. In 1994–95 Greek per capita incomes were only half the average EU level while those of Portugal were almost two-thirds. Irish and Spanish per capita GDP was nearly 80 per cent of the EU average. Similar distinctions emerge from the second column of Table 9.4, where unemployment rates are reported. Spanish and Irish unemployment is almost twice the EU average. Greek unemployment is at the average, while Portuguese unemployment is only half that of the EU as a whole.17 The fourth column reports output growth. Ireland is the fastest growing country, while Greek growth has been consistently below the EU average. Spain

EUROPEAN CURRENCY CONVERGENCE 261

Table 9.4 Convergence in the cohesion states

Source: European Commission, spring 1995 forecasts (GDP per head Autumn 1994). EU 15 includes ex-DDR in data beginning 1991 (1992 for labor productivity). Stage II: average of 1994–96. Stage I: average of seven semesters constructed from annual data. Note: Commission estimates need not follow the same methodology as national sources.

and Portugal grew at about the Union average during Stage I, and this will continue to be the case during Stage II if Commission forecasts are accurate.18 The fifth column focuses on external account, summarizing the link between real and nominal convergence insofar as external deficits are the ultimate cause of payments deficits, inflation or both. An excessive current account deficit may bring pressure to the foreign exchange market, particularly if the deficit is perceived as unsustainable. The negative signal will be especially strong if the public sector is spending too high a share of domestic resources. Two cases illustrating the polar extremes are Ireland, with large and consistent external surpluses, and Greece, with substantial deficits. Portugal and Spain display moderate external deficits, less than 1 per cent of GDP according to the forecasts for 1994–95. As for the public accounts (reported in Table 9.1 above) Portugal and Spain report budget deficits which are half of Greece’s but more than twice that of Ireland. Because of its lower level of public debt, however, Portugal’s interest burden is no higher than Ireland’s. Calculating wages and interest rates relative to the EU average and adjusting them for trends in labor productivity (as shown in the third column of Table 9.4 above) produces the ‘convergence diamond’ in Figure 9.1. All variables represented there are expressed as deviations from the EU 12 average. The figure suggests that the forecasts for Ireland for 1994–95 are uniformly inferior to the country’s performance during Stage I. That difference is especially pronounced with respect to productivity growth. For Spain, the forecast shows a slight improvement in nominal convergence and a slight deterioration in the rate of

262 IMPLICATIONS FOR EUROPE

productivity gain. The greatest improvement between Stage I and Stage II is forecast for Portugal. Policy evaluation The year in which the average rate of consumer price inflation first fell to single digits is one measure of when a country has acquired a reputation for financial stability. That date was 1983 in Ireland, 1985 in Spain, and 1992 in Portugal.19 There is little reason to question the credibility of the regime change undertaken by Ireland following its 1986 devaluation, and yet Irish reform was carried out in an environment of great uncertainty about the course of economic policy.20 It is worth recalling that the change in the fiscal regime toward lower taxes and improved spending control did not occur until 13 years after Ireland’s accession to the EC, and seven years after it joined the EMS. Spain’s regime change was signaled by its entry into the EMS in 1989. Although the 1992–93 recession created severe fiscal imbalances at the regional level, it never threatened the authorities’ commitment to convergence. Even the absence of a social consensus about wage moderation has not prevented a substantial fall in long-term interest rates (although the growing difficulties of the government in 1994 caused the inflation gap to widen again). In the cases of both Portugal and Spain, EMS entry was the signal that the regime had changed, in contrast to the budgetary consolidation that was decisive in Ireland. Meanwhile, all three states lifted their remaining restrictions on capital movements as part of the process of making their currencies fully convertible and gaining credibility with the markets. In Greece, nominal convergence has hardly begun. And any progress made on the inflation front has been offset by continued real divergence. The lack of a social consensus about wage moderation has been exacerbated by political instability. Even though the country has been subjected to strict multilateral supervision and surveillance since the first tranche of an EC loan was disbursed in 1991, few would claim that a convincing regime change is underway.21 With its low unemployment and sizable remittances, Portugal’s per capita GDP is already closer to Ireland’s than to Greece’s. Portugal has managed to attract foreign investment without experiencing a significant increase in unemployment like that in Ireland. But Portugal’s pattern of trade specialization is more dependent on low relative wages than is the case in Spain and Ireland. This is a mixed blessing. It makes the country’s industrial structure more vulnerable to adjustment costs but also more likely to benefit from the expansion of intra-European trade. CONCLUSION The ECU standard, like the gold standard before it, is a public good. Inevitably, adequate provision of the public good of systemic stability rests on the actions of

EUROPEAN CURRENCY CONVERGENCE 263

Figure 9.1 Convergence diamond

the largest national economy: Britain under the gold standard, the dollar under Bretton Woods, and Germany in Europe. Germany must provide not just the price stability demanded of the anchor country but a level of interest rates conducive to growth and adjustment throughout the Union. The incentive to free-ride in the provision of this public good is greatest for small countries, which have the least capacity to internalize the benefits of contributing it. In this sense, the benefits of establishing currency convertibility and maintaining exchange rate stability, which also contribute to the stability of

264 IMPLICATIONS FOR EUROPE

the convertible-currency system as a whole, are least apparent for small countries. From this follow the very different experiences of so-called core and peripheral countries that have long characterized the history of the international monetary system. The implications of this history for the prospects for EMU are that stability and convergence in Europe and a successful transition to monetary union require dedication on the part of the EU to providing the public good of monetary stability by maintaining low inflation and a level of interest rates consistent with growth and convergence. It requires a meeting of the minds between the recipients and donors on the magnitude and form of the transfers needed to help bring about convergence. It suggests that domestic political consolidation and multilateral convergence are both needed to ensure positive contributions by the smaller countries. Only with a mutually reinforcing interaction between the Unions richer and poorer and larger and smaller member states can the transition to monetary union be safely navigated. Any solution that is not based on the national cohesion of the member states willing and able to participate in an ECU standard would be unstable and would therefore lack credibility. The public good element of an ECU standard cannot be achieved against market sentiment, but history and geography both suggest that policy credibility can overcome hierarchy. Thus, through permanent negotiation among member states, national consensus on convergence brings about an enduring ECU standard. NOTES * Thanks are due to participants in the Arrábida Conference for comments and suggestions and to Nova for financial support of the research that went into this paper. Barry Eichengreen suggested many improvements in substance and style over the longer version appearing as Nova Economics Working Paper no. 226, September 1994. Nevertheless the author alone is responsible for any remaining shortcomings. 1 This insight, from Stiglitz and Weiss (1981), has been applied to Portuguese financial development after the nationalization of all banks, in 1975, in Macedo (1988). The hierarchy of financial markets has been stressed by Branson (1990) in connection with EMU. 2 Eichengreen (1994b) similarly argues that the existence of a well-defined escape clause was critical to the successful operation of the gold standard. 3 According to Stern (1991, p.180), Bismarck’s private banker warned him in 1874 that ‘the early introduction of an exclusive gold market would make Germany dependent on the British gold market, which the British defended by raising discount rates.’ I owe this reference to Stanford S.Warshawsky. 4 Abuses of the transfer mechanism have also been analyzed in the United States: Eichengreen (1992) discusses the scope for abuse and analyzes how safeguards are built into such programs (federal funding of state unemployment insurance programs, for example).

EUROPEAN CURRENCY CONVERGENCE 265

5 In the case of Portugal, the importance of geography was heightened by the fact that the announcement that the currency was to become convertible was made on August 13, only weeks before the outbreak of financial turmoil. 6 The procedure for issuing these guidelines was agreed to in Brussels late in 1993. For 1994 the guidelines instructed policy-makers ‘to allow market forces to display their full potential.’ 7 Portugal and Spain took steps in this direction immediately before and after the Treaty was signed, respectively. As mentioned earlier, independence does not involve banking supervision; this raises the issue of accountability for central banks with both competences. Eichengreen (1994a) addresses central bank accountability issues in connection with monetary policy. 8 This is labeled the ‘promised land syndrome’ in Williamson (1994), where policy reforms in the two Iberian countries are discussed. See also Torres and Giavazzi (1993). 9 The analysis in Macedo (1980, p.342) is extended using new estimates of nominal income in Macedo (1995a). 10 Macedo (1995a, 1995b) contains estimates for the various phases. 11 Relative to the industrializing countries of Asia and Latin America, however, the 1992 restoration of convertibility does not seem out of synch. Work by Marion and Klein (1994) on sixteen Latin American countries and Jamaica during 1957–91 shows the median duration of a dollar peg to be less than one year. 12 Government turnover is used in Macedo (1995a, 1995b) as a proxy for political stability, in spite of the changes in political regime and in voter participation. 13 The phases are justified in Macedo (1995a). See also Part III of this book; Eugénia Mata and Nuno Valérios chapter ignores the last attempt to join the gold standard, in the form of the aborted restoration of convertibility in 1931. Therefore they also neglect the 1924–31 gradual regime change, lumping it together with the 1930s. Conversely, they attach monetary relevance to World War II, and define an additional sub-period, 1940–45. The average budget deficit during World War II does not change relative to the next sub-period. External payments did turn from deficit to surplus and inflation rose to double digits, but the war effect did not last on either one of these variables. 14 Ironically, it coincided with a reversal of the strong deflation of 1986–88, so that the implications for price stability and public finances were not immediately apparent. 15 The stance of policy at this time is reviewed by Braz (1992). 16 Using data from the late 1980s, Panic (1992, p.140) stresses that regional disparities across EC member states are larger than those within them. 17 A comparison of the two time periods shows that unemployment is worsening everywhere, but that the deterioration is particularly severe in the two countries where its level was already high at the start of the period. 18 An error in the reported growth rate for Ireland in Table 1 of Macedo (1994), from which this column is updated, has been corrected. 19 The last previous instances of single digit inflation were 1972 in Greece and Spain, 1973 in Portugal, and 1978 in Ireland. 20 Walsh and Leddin (1992, p.133). I owe this reference to Frank Barry. 21 This remained true even after Greece removed all remaining capital controls in May 1994.

266 IMPLICATIONS FOR EUROPE

REFERENCES Bayoumi, Tamim and Barry Eichengreen (1994), ‘Restraining Yourself: Fiscal Rules and Stabilization’, Center for Economic Policy Research Discussion Paper no.1029, September. Branson, William (1990), ‘Financial Integration, Macroeconomic Policy and the EMS’, in Christopher Bliss and Jorge Braga de Macedo (eds) Unity with Diversity in the European Economy: The Community’s Southern Frontier, Cambridge: Cambridge University Press, pp. 104–30. Braz, José (1992), ‘Portugal from P1 to Q2: A Strategy of Sustained Regime Change’, in A Single Currency for Europe: Monetary and Real Aspects, London: CEPR and Bank of Portugal. Report of a Conference held in Estoril, whose proceedings are reported in Francisco Torres and Francesco Giavazzi (eds) (1933). Eichengreen, Barry (1992), ‘Is Europe an Optimum Currency Area?’ in Silvio Borner and Herbert Grubel (eds) The European Community After 1992: The View from Outside, London: Macmillan, pp. 138–53. Eichengreen, Barry (1993), ‘European Monetary Unification’, Journal of Economic Literature, vol.XXXI, September. Eichengreen, Barry (1994a), International Monetary Arrangements for the 21st Century, Washington, D.C.: The Brookings Institution. Eichengreen, Barry (1994b), ‘Deja Vu All Over Again: Lessons from the Gold Standard for European Monetary Unification’, in Tamim Bayoumi, Barry Eichengreen and Mark Taylor (eds) Economic Perspectives on the Classical Gold Standard, Cambridge: Cambridge University Press (forthcoming). European Commission (1990), ‘One Market, One Money’, European Economy no. 44, October. European Commission (1993), ‘Stable Money: Sound Finances—Community public finance in the perspective of EMU’, European Economy, no.53. Marion, Nancy and Michael Klein (1994), ‘Explaining the Duration of Exchange Rate Pegs’, paper presented at the 1994 NBER Summer Institute. Macedo, Jorge Braga (1980), ‘Portuguese Currency Experience: An Historical Perspective’, in Estudos em Homenagem ao Prof. Doutor J.J.Teixeira Ribeiro, vol.IV, Coimbra: Boletim da Faculdade de Direito, pp.311–52. Macedo, Jorge Braga (1988), ‘Comment on Ronald McKinnon, “Financial Liberalization in Retrospect: Policies in LDCs”’, in Gustav Ranis and T.Paul Schultz (eds) The State of Development Economics Progress and Perspectives, New York: Basil Blackwell, pp.411–15. Macedo, Jorge Braga (1994), ‘European Union and Cohesion’, in Europäisch Antagonismen, Zurich: Schweizerisches Institut für Auslandforschung, pp.33–57. Macedo, Jorge Braga (1995a), ‘Convertibility and Stability 1834–1994: Portuguese Currency Experience Revisited’, in Ensaios em Honra de Francisco Pereira de Moura, Lisbon School of Economics and Business. Macedo, Jorge Braga (1995b) ‘Portugal and European Monetary. Union: Selling Stability at Home, Earning Credibility Abroad’, in Francisco Torres (ed.) Monetary Reform in Europe, Lisbon: Catholic University of Portugal. Panic, M. (1992), European Monetary Union: Lessons from the Classical Gold Standard, London: St. Martin’s Press.

EUROPEAN CURRENCY CONVERGENCE 267

Stern, Fritz (1991), Gold and Iron: Bismarck, Bleichroder, and the Building of the German Empire, New York: Alfred Knopf. Stiglitz, Joseph and Andrew Weiss (1981), ‘Credit Rationing in Markets with Imperfect Information’, American Economic Review, 71, June. Torres, Francisco and Francesco Giavazzi (eds) (1993), Adjustment and Growth in the European Monetary Union, Cambridge: Cambridge University Press. Walsh, Brendan and A.Leddin (1992), The Macroeconomics of Ireland, 2nd edition, Dublin: Gill and Macmillan. Williamson, John (ed.) (1994), The Political Economy of Policy Reform, Washington: Institute for International Economics.

268

INDEX

Abramowitz, Moses 39 Aceña, Pablo Martin 31, 74 agriculture 208 Ambedkar, B.R. 135, 136 arbitrage/arbitrageurs 103, 104, 106, 107, 169, 170 Argentina 37–40, 41–4; see also Latin America asset prices see exchange rates Australia 25, 36, 112, 135, 244 Austria 93, 95, 104, 106, 112, 114, 121, 134 Austro-Hungarian Bank 135

supervision 191, 250–50, 262 Banking Act (UK, 1844) 25 Barbosa, Rui 129 Baring Brothers 40, 199, 209, 236 Barro, Robert J. and Gordon, David B. 12 Barry, Frank 265 Bastien, Carlos 177 Bayoumi, Tamin and Eichengreen, Barry 57, 63, 76, 147 beauty contest dynamics 107 Belgium 32, 88, 91, 92, 112, 120, 121, 127, 129, 166, 177, 253 belle epoque 101, 203, 205, 209–11, 239 Belmond Morgan syndicate 41 Benini, Rudolfo 108 Bento, Carlos 179 Bernanke, Benjamin 76; and James, Harold 76 bimetallism 11, 13, 14, 18, 19, 23, 27, 30, 32, 33, 74, 75, 85, 87, 89–98, 112, 114– 18, 117, 120, 121, 127, 129, 133, 136, 145, 146, 159, 164, 169, 170, 172, 174, 175, 176, 177, 205, 233, 256 Bismarck, Otto von 137, 264 Blanchard, Olivier and Quah, Danny 57 Bloomfield, Arthur I. 15, 101, 102, 110, 135 bonds 28, 33, 90, 92, 94, 104, 106, 146, 175, 188, 237, 259 Bonifácio, Maria de Fátima 179 Bordo, Michael D. 16, 34, 76, 147; and Capie, Forrest 176; and Jonung, Lars 76; and Kydland, Finn E. 10, 13, 19, 75, 144, 198, 201; and Santos, F.T. 198;

Baas, N.W.J. 96 backwardness hypothesis 123 balance of payments 4, 28, 33, 35, 151, 214, 215, 231, 257 Banca Nazionale 108 Bank of Belgium 127 Bank of England 16, 19, 25, 74, 95, 99, 102, 131, 134, 181, 183, 185, 194, 195– 6, 199, 231 Bank of France 17, 127, 135, 251 Bank of Italy 106, 108, 109 Bank of Japan 33–5 Bank of Lisbon 159, 160, 162–4, 167, 205 Bank of Portugal 162, 165, 177, 178–80, 179, 183, 185, 188, 191, 193, 195–7, 199, 201, 215, 216, 218, 224, 258, 259 Bank of Spain 31, 136, 251 banking 4; confidence 213; crises 152, 194, 257; panics 23, 25, 27;

269

270 INDEX

and Schwartz, Anna J. 76, 147, 177; and White Eugene N. 75 Branco, Carlos de Barros Soares 201 Branson, William 264 Brazil 26, 129, 152, 173, 199, 207, 208, 227, 231, 236 Bretton Woods 11–12, 17, 18, 19, 43, 44, 50, 56, 62, 63, 73, 75, 107, 143, 149, 153, 214; and adherence to/suspension of specie rules 18–37; and evaluation of adherence to rules 34– 8 Broadberry, S.N. and Crafts, N.F.R. 176 Brussels Conference (1920) 181, 200 budgetary discipline 170 budgets 123, 189, 199, 200, 207, 208, 224, 231, 257, 258–8, 262; deficits 56, 57, 132, 138, 188–90, 207, 208, 214, 216, 224, 237 Bundesbank 251; see also Rëichsbank Butlin, S.J. 26 Cabeçadas, J.M. 200 Cabral, Lourenço 179 Caetano, Marcelo 199 Caixa de Conversao 130 Caixa Geral de Depòsitos 199 Caja de Conversion 40 Calomiris, Charles W. 40, 41 Cameron, Rondo 176 Canada 25, 36, 75, 112, 135, 143 Canzoneri, Matthew and Henderson, Dale W. 15 capital 174, 231; access to 37; availability 213–14, 231, 253, 257; borrowed 138; controls 18, 35, 215, 251, 258, 265; hierarchy 244; inflows/outflows 3, 3, 36, 37–40, 41–4, 90, 184; long-term 152; short-term 99–11, 151, 152; and specie standard adherence 37–40, 41–4;

stabilization of 16 capital market 207 Castro, Alvaro de 187, 188, 200, 201, 237 Cecchetti, Stephen G. and Karras, Georgios 76 central banks 4, 15, 32, 85, 99, 104–7, 134, 151, 191, 248, 253, 254, 259; accountability of 250–50, 264; defined 101; foundation of 102; metallic reserves in 123–8, 145; and raids on gold 101 Cipolla, Carlo 102 Clarence-Smith, Gervase 179 closed economy 12 cohesion 245, 252, 254–4; cohesion fund 252; cohesion states 260 coinage 13, 29, 95, 96, 112, 125, 127–2, 131, 136–40, 145, 146, 159–1, 161–3, 165–7, 167, 168, 171, 172–4, 177, 178, 192, 201, 207, 209–10; 224, 230 Coinage Act (Japan, 1897) 33 Coinage Act (US, 1792) 23 colonial war 214 Comèrcio do Porto 201 commitment regime 12 Committee of Central Bank Governors 243 Conant, Charles A. 135, 136, 137, 138 constitutional rule 92, 141, 208 contingent rule 11, 73–6, 145, 198; adherence/suspension of 18, 19, 23–37; defined 12–18 convergence 43–5, 50, 56, 62–4, 246–6, 253, 262; diamond 260; programs 253, 259 convertibility 3, 10, 13, 97, 110–14, 123, 135–9, 145–8, 152, 231, 235, 239, 244, 255–5, 258; adoption of 3; defined 112, 114; effects of 120; of escudo 188–93, 197–8; of gold 17, 18; guaranteed 32;

INDEX 271

and inconvertibility 112, 114, 116, 133, 134, 143–6, 164, 195, 257; maintenance of 14, 103–6; restoration of 40, 191, 216; suspension of 6, 11, 19, 23, 25, 26–8, 31, 32, 39, 41, 114, 115, 195–6, 212, 230; threats to 41–4 Cooper, Richard 10 Cordes, Sinel de 189, 189, 200 core-periphery 5, 11, 114, 133, 151, 227, 244, 250; and adherence to/suspension of specie rules 18, 19, 23–37; economic performance of 43–5, 50, 56, 62–5, 73 Cortés-Condé, Roberto 39, 40 Costa, Afonso 181, 199 Cottrell, P.L. 134, 176 Council of Finance Ministers (ECOFIN) 8, 252, 253, 259 credibility 200, 245, 248 Crédit Mobiliet 92 Crisp, Olga 180 Cunliffe Committee 107 currency, blocs 5; collapse of 89; domestic 14; sustainable 4 Currency Banking School 74 d’Avila, Antonio José 178 Davis, Lance 74; and Huttenback, Robert A. 38, 76 de Cecco, Marcello 10, 85, 87, 117, 134, 180 debt 12, 13, 15, 29, 31, 75, 103, 106, 123, 144, 152, 162, 200, 210, 231, 237 deflation 31, 34, 117 DeKock, Gabriel and Grilli, Vittorio 13 Delors, Jacques 243 demand/supply 237; disturbances (shocks) 57–5, 73, 76, 153–5 depreciation 184, 186, 187, 195, 200, 203, 210–11, 215, 217, 231, 236, 238 Deutschmark 245, 258, 259

devaluation 15, 25, 32, 35, 36, 108, 207, 260 dictatorship 197 discretionary regime 12–13 divergence 262 ECOFIN see Council of Finance Ministers economic growth/development 11–12, 43– 5, 50, 56, 62–5, 73, 122, 123, 133, 135, 141, 144, 203–5, 208, 213–14, 217, 227, 238, 247; shocks to 154, 236, 239; and surveillance 253, 255 The Economist 164 ECU basket 259; standard 243, 243, 245, 262; deepening and widening 251–1; and financial markets 250–50; instability in stage II of EMU 247, 248– 9; prerequisites for convergence 246–6; state of convergence 247; see also European Monetary Union (EMU); monetary unification Edelstein, Michael 39, 76 Eichengreen, Barry J. 9, 14, 15, 16, 17, 25, 34, 40, 62, 74, 264; and McLean, Ian W. 178 Einzig, Paul 106 emigrants’ remittances 235–5 employment/unemployment 35, 255, 259, 260, 262, 265 Eschweiler, Bernhard and Bordo, Michael D. 75 escudo 181–4, 186–7, 188–93, 197–9, 200, 231, 248 Esteves, Paulo 218 euro 243, see also ECU European Commission, One Market, One Money 246 European Community 215 European Currency Unit (ECU) 243; see also ECU standard European Free Trade Association (EFTA) 215

272 INDEX

European Monetary System (EMS) 3, 8, 216, 243, 243, 248, 253, 259, 260–1 European Monetary Union (EMU) 98, 243, 243, 244, 245, 247, 248–9, 255, 262; see also ecu standard; monetary unification European Organization for Economic Cooperation 215 European Union (EU) 6, 93, 243, 244, 245, 247, 251–1, 260, 262 Excessive Deficits Procedure 7, 252–2 Exchange Rate Mechanism (ERM) 3, 243, 253 exchange rates 27, 28, 29–1, 75, 106–9, 108–11, 116, 132, 133, 135, 137, 181, 186, 197, 200, 203, 205, 214, 216, 224, 205, 227, 231, 233, 236, 238, 252; crisis in 248; fixed 8, 10, 15, 28, 35, 36, 90, 91, 141, 146, 245; flexible 248–9; floating 3, 7, 30; multiple 34–6, 37; policy 4, 7; restrictions 3; and surveillance 253; unified 35; (un)stable 33, 74, 85, 117, 252–2, 254, 255, 257 expectations 236 external accounts 207, 260; see balance of payments Federal Reserve Bank of New York 194 Fetter, F. 74 financial markets 248; see also capital markets financial press 259; see also media financial protection 258; see also capital controls Financial Times 192 fiscal defecit see budgetary discipline Fishlow, Albert 15, 135, 137 Flandreau, Marc 74, 85, 134, 135, 136, 137, 147 Flood, Robert P. and Isard, Peter 13

fluctuation margins 253; see also ERM Ford, A.G. 39, 134, 137 foreign exchange 3, 28, 30, 31, 34, 36, 104, 108, 135, 188, 191–2, 193; crises 248, 252 Fournier, Fiorillo 106 France 19, 23–6, 34–6, 73, 76, 88, 89, 91, 92, 93, 94, 112, 116, 120, 122, 127–30, 129, 137, 151, 159, 243, 253 Fratianni, Michele and Spinelli, Franco 30, 135 free trade 74 Frieden, Jeffrey A. 17, 147 Friedman, Milton 15, 74; and Schwartz, Anna J. 40, 41, 76, 199 Fritsch, Winston, and Franco, Gustavo H.B. 27 Gallarotti, Giulio M. 9, 15, 74, 147, 176 GDP 188–9, 203, 208, 216, 218–4, 227, 229–30, 247, 259, 260, 262 Genoa Conference (1922) 181, 188, 189 geographic fundamentals 248 geography of specie standard 246 Germany 19, 23–6, 35, 37, 73, 75, 88–2, 92, 93–8, 106, 110, 114, 120, 122, 127, 127, 129, 141, 143, 151, 179, 212, 245, 251, 253 Gibbs, H.H. and Greenfell, H.R. 90 Giovannini, Alberto 10, 15, 16, 25, 34, 75, 180 GNP 43, 56; see also GDP gold, bloc 201; price/supply of 26, 37, 85–8, 87–2, 97, 101–4, 167, 169–1, 174, 178 Gold Reserve Fund 125 gold standard 17, 19, 27, 28, 29–1, 32, 33, 38, 43, 44, 57, 75, 106, 134, 135, 188, 201, 203, 217, 236, 244, 245, 259; in (1854–91) 205–9; in (1868) 16–18, 112; in (1908) 117, 120; abandonment of 183; adoption of 5–6, 145, 159–60, 197–8, 230, 235;

INDEX 273

areas of 110–14; as asymmetric system 16; choice of 85–87, 89, 97–1; and cooperative monetary arrangements 127–2; domestic 12–14; effects/effectiveness 4; and foreign trade 86–87; history of 10–12, 132, 133; and imperial ties 129–3; international 14–17; and market confidence 4; and metallic base at the central bank 123–8; and national income 122–7; one way (reflecting barrier) 136; origins of 85–98; performance 10; and politics 85–9, 97; reasons for 146–9, 149–4; reputation of 15; return to 181–3, 192; and size of country 125–30; and stability 85, 87, 89; success of 13; transition to 120–35, 144, 145, 163–72; and war 130–5 Gold Standard Act (US, 1900) 41 gold-silver ratio 160, 161, 165, 166–8, 169– 2, 176, 230 Goodhart, C.A.E. 87 Granger causality tests 185, 199, 200 Great Depression 75, 153, 213, 231 Greece 29–1, 114, 131–5, 244, 247, 260– 60, 262, 265 Grilli, Vittorio 25, 41 Grillo, Giacomo 106, 108 Grimaldi, Marquês Camilo Pallavicino 178 Grossman, Herschel J. and Huyck, John B. Van 13 Guedes, José Guedes 180 Hawtrey, Ralph 101, 102 Hayashi, Fumio 15 Helfferich, K. 95 hierarchy 264;

see also capital markets Hirschman, Albert O. 101 Hohmann, Alexandre 74 Holtfrericch, Carl Ludwig 180 Hong Kong Bank 117 Huffman, Wallace E. and Lothian, James E. 177 hyperinflation 19, 75, 255 imperialism 129–3, 145 income, national 122–7, 135; redistribution of 246–6 India 125, 136, 137 inflation 7, 11, 28, 35, 36, 44, 56, 62, 106, 117, 191, 197, 203, 210, 211–12, 214, 215, 216, 217, 236, 239, 255–7, 258, 262, 265; post-war 183–8 Inflation Bill (US, 1874) 41 inflation tax see tax instability, political 208, 231, 239, 255 interest rates 15, 75, 99, 106, 262 Intergovernmental Conference (1996) 6, 8, 243, 252 International Monetary Conference (1881) 96 International Monetary Fund (IMF) 18, 19, 34, 35, 37, 214, 216 investment 208, 235, 239 Ireland 260–60, 262, 265 Italy 30, 90, 92–6, 104, 106, 108, 112, 114, 243, 245 Janssen, A.E. 95 Japan 32–5, 37, 44, 50, 73, 76, 129, 131, 135, 137 Jones, R. 146 Jonung, Lars 14, 74 Journal des Debats 164 Journal do Comercio 171 Justino, David 224 Kauch, P. 136, 177 Keating, John W. and Nye, John V. 76 Keynes, John Maynard (Keynesian) 10, 35, 63, 105, 107–10, 134, 137 Kindleberger, Charles P. 137

274 INDEX

Kiyotaki, N. and Wright, R. 146 Korea 129 Kredit Austalt 194, 232, 237 Krugman, Paul 8, 16 Kydland, Finn E. and Prescott, Edward 12 labor mobility 251 Lains, Pedro 218; and Reis, Jaime 218 Latin America 25–7, 36–8, 50, 56, 76, 108, 121, 129–3, 132, 137, 145, 152–4, 209, 244, 258, 265 Latin Monetary Convention 91, 92, 92–6, 96 Latin Monetary Union 29, 30, 31, 93, 95, 96, 98, 116, 122, 136 Latin Union Treaty Amendment (France, 1874) 127 Lavradio, Marquis of 179, 180 Law, John 19 Lazaretou, Sophie 29, 74, 135 League of Nations 189, 201 lemons premium 38 Lindert, Peter H. 16, 135 Lisbon Commercial Association 179 Llona-Rodriguez, Augustin 28, 74, 76 loans 15, 29, 34, 35, 38, 76, 103, 131, 132, 174–6, 180, 181, 185, 200, 207, 262 London 38, 89, 93, 99, 101, 102, 117, 130, 133, 165, 174–6 London Joint-Stock Bank 137 Lucas, Robert E. Jr. and Stokey, Nancy L. 13, 74 Luzzatti, Luigi 107, 136 Maastricht Treaty 6, 7–8, 243, 245, 246, 247, 252, 253, 254, 264 MacDonald, Ramsay 194–5 Macedo, Jorge Braga de 5, 6, 8, 9, 134, 199, 226, 264, 265 McKinnon, Ronald I. 10, 17 Madeira 218 Madrid European Council 243 managed float 43, 50, 56, 62 Mankiw, Gregory 13 Maria da Fonte uprising 161 Marion, Nancy and Klein, Michael 265

Marques, Carlos and Esteves, Paulo 218 Marshall, Alfred 107 Marshall Plan 215 Martin, David A. 134, 177, 178 Martins, Avila and Alves 179 Masayoshi, Matsukata 137 Mata, Eugénia 200, 208, 212 Matsuoka, K. 97 media 105–9 Meltzer, Allan H. and Robinson, Saranna Robinson 10 Mertens, J.E. 85, 89, 137 middle class 92 Miller, Marcus and Sutherland, Alan 75 Mitre, Bartolomé 39 modernization 86, 87, 94, 98 monetary systems 6; (1868) 112, 114–19; (1908) 117, 120; new 172–6; reform of 171, 254–4; research in 133; transitional dynamics 120–35 monetary unification 243–3, 262–3; and geography of the ECU standard 246–6, 248–51; implications from history 244–5; managing transition to 252–4; and Portugal 255–62; see also ECU standard; European Monetary Union (EMU) money, growth 50, 62–4; laundering 250; supply 209, 214, 224, 229 National Bank of Greece 29 Netherlands 32, 37, 90, 112, 120, 127–2, 135 Netherlands Bank 95, 135 Neves, Joáo César 224 Newton, Sir Isaac 19, 110 Nogaro, Bertrand 136 Nunes, Ana Bela et al. 218 Obstfeld, Maurice 14 Officer, Lawrence 15 oil shocks 215–16, 238

INDEX 275

Oporto Industrial Association 179 Oppers, S. 147 Ozkan, F. Gulcin and Sutherland, Alan 8 Paillard, G. 92 Pais, Sidonio 199 Panic, M. 265 Paollera, Gerardo della 74 paper currency 14, 19, 23, 28, 29, 30, 32, 33, 35–7, 75, 114–18, 116, 135, 136, 163, 164, 177, 179, 192, 201, 207, 233 Pareto, Vilfredo 107 Paris International Conference (1878) 97 peer pressure 252 pegging 3, 4, 7, 8, 13, 34, 194–7, 195, 212, 231, 237, 243 Pelaez, Carlos M. and Suzigan, Wilson 27 Pereira de Melo, Fontes 167, 175, 179 Perlman, Mark 136 permanent negotiation 264; see also peer pressure Peters, H.E. 153 Pimentel, Julio 177, 179 Pinheiro de Sousa, Magda 180 politics 4, 85–9, 97, 102, 245–5 Pope, David 74 Portugal 88, 110, 112, 117, 134, 141, 146, 244, 245, 247; and adherence to specie rules 30–2; adoption of gold standard 159–60, 175– 7; before gold standard 159–4; and ‘belle epoque’ (1891–1914) 209– 11; comments on 227–69; constitutional ban on privatization 258; coups and revolutions 209, 239, 255; economic trends 203–5, 216–25; and EMS 8, 216, 257; and the gold standard (1854–91) 205– 9; governments of 199; inflation and fiscal discipline 255–7; and new monetary system 172–6; parliament 8, 255, 259; and peg to sterling 194–7; policy evaluation 260–2;

post-war experience 183–8, 214–16; progress in comparative perspective 260–60; Regeneration movement in 168; regime change and aftermath 258–8; return to gold standard 181–3; rotativismo 235; stability in (1924–39) 212–13; and stability/convertibility of escudo 188–93, 197–8; transition to gold standard 163–72; war/post-war inflation 183–8, 211–12; and World War II (1939–45) 213–14 prices 44, 152, 185, 205, 208, 211, 213, 217, 227, 230, 254 productivity 260 prudential supervision 259 public accounts see budget Public Credit Act (US, 1869) 40 public good 262 Public Debt Office (Portuguese) 168 purchasing power parity 237 Q2 259 Quantum 258 Raffalovich, A. 105 railways 90, 138, 152, 175 recession 248, 254–4 Redish, Angela 74, 134, 146, 169, 179 regime change 254, 260 Reichsbank 75, 135 see also Bundesbank Reidl, R. 96 Reis, Jaime 30, 74, 88, 134, 198, 207, 209, 224, 230, 233 rental moderation 250, 262 Rente Thiers 131 reputation 247, 250, 254, 260–2 see also credibilty Resumption Act (US, 1875) 41, 75 Rich, Georg 74 Rodrigues, Daniel 188, 200, 201 Roll, Richard 40, 75 Root, L.Carroll 136 Rothschilds 130 rules of the game 15–16

276 INDEX

Russia 105–8, 107, 114, 116, 121, 135, 137–1, 180, 238, 239 Salant, Stephen and Henderson, Dale 137 Salazar, António d’Oliveira 183, 189, 192, 197, 200, 201, 202, 231, 237, 238 Santos, Fernando 5, 74, 213, 230, 236 Scandinavia 31–3, 37, 73, 88, 95, 112, 121, 127, 135, 181, 244, 247 Scandinavian Monetary Convention (1872) 95 Schwartz, Anna J. 16, 74 Seyd, Ernest 134 Sharkey, R.D. 40 Shearer, Ronald A. and Clark, Carolyn 26 Sherman Silver Act (US, 1893) 41 Sherman Silver Purchase Act (US, 1890) 41 Shimonoseki Peace Treaty (1895) 131 Shinjo, Hiroshi 32 shocks see demand/supply disturbances Short-term Finance Facilities 8 Silva, Rebello da 178 silver 33, 89, 94–8, 115, 122, 169; and international trade 87–1; price/supply of 85–8, 88, 90, 96, 125, 132, 141, 164, 169, 173, 178, 230 silver standard 11, 14, 18, 19, 28, 31, 90, 110, 117, 120, 121, 127, 128–2, 133, 146, 166, 179; support for 95–97 Simons, Henry C. 12 Smith, W.S. and Smith, R.T. 75 Snyder, Jack and Jervis, Robert 176 Société des Agriculteurs de France 96 Sousa, Rita 177 Spain 31, 88, 90, 116, 136, 166, 243, 244, 245, 247, 260–60, 262 Spalding, 136 specie standard, adherence/non-adherence to rule 10–11, 12–18, 19, 23–6, 39, 73–6, 145; and capital flows 37–40, 41–4; demand and supply disturbances 57–5, 73; during Bretton Woods 34–8; fixed price 10;

pre-Bretton Woods 19, 23–34; reputation of 14; stability and convergence 43–5, 50, 56, 62–4; success of rule 74 stability 44, 85, 87, 89, 198, 201, 203, 208, 210, 212–13, 216–17, 229, 233, 237, 238, 243, 244, 246, 257 Stackelberg strategic game 16 stagflation 11 sterling 183, 188, 194–7, 212, 230, 231, 237, 256 Stiglitz, Joseph and Weiss, Andrew 38, 264 stock market 193 supervision see banking supply see demand/supply surveillance, multilateral 243, 247, 253 Svennson, Lars E.O. 16 Switzerland 19, 32, 43, 90, 92, 112, 120, 121, 127, 129, 134 Taipa, Count of 171 tariffs 159, 179 taxes 12, 28, 29, 35, 75, 106, 132, 143, 146, 189, 239 Teixeira de Aragáo, Augusto Carlos 176, 177, 179 Telo, Antònio Josè 200 Thery, Edmond 108 Toniolo, Gianni and Acena, Pablo Martin 8 Topik, Steven 137 trade 108–11, 134, 181, 210, 231, 257 ultimatum, Britain 236 Unger, I. 40 Union Treaty see Maastricht Treaty United Kingdom 35, 74, 87, 88, 89, 91–5, 110–14, 117, 120, 134, 141, 146, 151, 172, 176, 181–3, 212, 227, 231, 243; and adherence to specie rules 23–6; as leader of gold standard 16–17 United States 19, 23–6, 34–7, 37–40, 75, 88, 106, 115, 144 unemployment see employment

INDEX 277

Valério, Nuno 205, 208, 224 VAR see vector autoregression variable geometry 5, 6, 244, 252 vector autoregression (VAR) 57–5, 76, 153 velocity of monetary circulation 211, 224, 230 Very-Short-Term Finance Facilities 9 Vieira, Antonio Lopes 180 Viner, Jacob 74, 99, 102 wage moderation 246, 259 Walsh, Brendan and Leddin, A. 265 wars 13, 19, 27, 28, 29, 30, 31, 34, 74–7, 93, 94, 114, 143, 161, 181, 184–5, 205, 213–14; role of 130–5 Warshawsky, Stanford S. 264 wealth 143, 144 Weber, Ernst Juerg 136 White, H.D. 107–10 Williamson, John 264 Willis, H. Parker 135 Withers, Hartley 102 Witte 97, 238 World Bank 214 Xavier, Alberto 200 Yeager, Leland B. 15 Young, John Parke 134