The Gold Standard; Theory and History

The Gold Standard in Theory and History

The revised edition of this acclaimed book provides a comprehensive analysis of the theory and history ofthe gold standard. It contains six new essays drawing on the latest research to provide a framework forcontemporary debates about the gold standard. New topics include:

Applications to the gold standard of the post-1990 literature on exchange-rate target zonesThe significance of the gold standard in the European Monetary Union debate

The Gold Standard in Theory and History will be an invaluable resource for students of macroeconomics,international economics and economic history at all levels.

Barry Eichengreen is John L.Simpson Professor of Economics and Political Science at the University ofCalifornia, Berkeley. He has written books on a wide variety of economics subjects. Marc Flandreau is aresearcher at the Centre National Recherche Scientifique in Paris, and an associate researcher at theObservatoire Français des Conjonctures Economiques, Paris. He is a specialist in the theory and history ofthe international monetary system.

THE GOLD STANDARD IN THEORYAND HISTORY

Second EditionEdited by

Barry Eichengreen andMarc Flandreau

London and New York

First published in 1985 byMethuen Inc

29 West 35th Street, New York, NY 10001

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 towww.eBookstore.tandf.co.uk.”

Second edition published 1997by Routledge

11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canadaby Routledge

29 West 35th Street, New York, NY 10001

© 1997 Barry Eichengreen and Marc Flandreau

All rights reserved. No part of this book may be reprinted or reproduced or utilized in anyform 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 DataA catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication DataA catalogue record for this book has been requested

ISBN 0-203-97887-0 Master e-book ISBN

ISBN 0-415-15060-4 (hbk)0-415-15061-2 (pbk)

For Michelle, pour Mathilde

Contents

Preface and Acknowledgements vii

1 Editors’ introductionBarry EichengreenMarc Flandreau

1

Part I The gold standard in theory 22

2 On the balance of tradeDavid Hume

24

3 The working of the prewar gold standardP.B.Whale

31

4 How the gold standard worked, 1880–1913Donald N.McCloskey and J.Richard Zecher

40

5 A portfolio-balance model of the gold standardTrevor J.O.Dick and John E.Floyd

54

6 The gold standard as a ruleMichael D.Bordo and Finn E.Kydland

70

Part II The gold standard in history 93

7 The significance of the gold pointsJohn Maynard Keynes

95

8 The myth and realities of the so-called gold standardRobert Triffin

101

9 Notes on the working of the gold standard before 1914A.G.Ford

116

10 The gold standard since Alec FordBarry Eichengreen

135

11 The dynamics of international monetary systems: international and domestic factors inthe rise, reign, and demise of the classical gold standardJeffry A.Frieden

150

Part III The interwar gold exchange standard 164

12 First interim reportCunliffe Committee on Currency and Foreign Exchange after the War

166

13 ReportMacnillan Committee on Finance and Industry

177

14 The gold exchange standardRagnar Nurkse

188

Part IV Bretton Woods and after 207

15 The gold-dollar system: conditions of equilibrium and the price of goldMilton Gilbert

209

16 The Bretton Woods system: paradise lost?Barry Eichengreen

224

Further reading 235

Index 239

vi

Preface and Acknowledgements

This is the second edition of The Gold Standard in Theory and History, revised after an interval of twelveyears. In addition to comprehensively revising the editors’ introduction, we have recast the table ofcontents, dropping three chapters from the first edition and adding six new ones.

The editors and publisher would like to thank the following for permission to reproduce copyrightmaterial.

P.B.Whale and Tieto Limited, chapter 3; Donald N.McCloskey and J. Richard Zecher, chapter 4, whichfirst appeared in Jacob A.Frenkel and Harry G.Johnson (eds), The Monetary Approach to the Balance ofPayments (George Allen & Unwin); Trevor Dick, John Floyd and Academic Press, chapter 5, which firstappeared in Explorations in Economic History, April 1991; Michael D.Bordo, Finn Kydland and AcademicPress, chapter 6, which first appeared in Explorations in Economic History, October 1995; The RoyalEconomic Society, chapter 7, which first appeared in The Treatise on Money (Macmillan) and wasrepublished as volume 6 of The Collected Writings of John Maynard Keynes (Cambridge University Press);Robert Triffin, chapter 8, from Our International Monetary System: Yesterday, Today and Tomorrow byRobert Triffin, copyright 1968 by Random House, Inc., reprinted by permission of Random House, Inc.; A.G.Ford, chapter 9, originally printed in Oxford Economic Papers, February 1960, reprinted by permissionof Oxford University Press; Barry Eichengreen and Cambridge University Press, chapter 10, whichoriginally appeared in Steven Broadberry and N.F.R.Crafts, Britain in the International Economy, 1870–1939 (Cambridge University Press); Jeffry A. Frieden and Westview Press, chapter 11, which originallyappeared in Jack Snyder and Robert Jervis, Coping with Complexity in the International System (WestviewPress); chapter 15 first appeared in Milton Gilbert, The Gold-Dollar System: Conditions of Equilibrium andthe Price of Gold’, Essays in International Finance, No. 70, November 1968, copyright 1968, reprinted bypermission of the International Finance Section, Department of Economics, Princeton University; BarryEichengreen and the Association d’économie financière, chapter 16, which originally appeared (in French)in Bretton Woods: Mélanges pour un Cinquantenaire (Association d’économie financière).

1Editors’ introduction

Barry Eichengreen and Marc Flandreau

The financial globalization that followed the collapse of the Bretton Woods System opened a new chapter inthe history of international monetary relations. The founding fathers of Bretton Woods—the eminent Britisheconomist John Maynard Keynes and US Treasury official Harry Dexter White prominent among them—sought to create a more perfect international monetary order conducive to financial stability and economicgrowth. Their system was crafted to contain destabilizing flows of ‘hot money’ and to allow governments toadjust policy to domestic conditions. For a few years it seemed to work. But by the 1970s internationalfinancial markets had regained the upper hand. International capital flows seemingly unprecedented inscope first undermined fixed currency pegs. Following the transition to floating exchange rates, financialmarkets threatened officials seeking to avail themselves of their newfound freedom with capital flight, thecollapse of the currency, and inflation. When US President Bill Clinton’s advisor James Carville famouslyremarked that if reincarnated he wanted to come back as the bond market, he was only articulating manyofficials’ sense of helplessness when confronted by a world of global finance.

Critics of the current system contend that the generalized float now prevailing provides neither thestability needed for effective international specialization nor the flexibility required for independent action.Turbulence and volatility in international financial markets disrupt firms’ production and investmentdecisions. Misaligned currencies confer arbitrary competitive advantages, evoking complaints from thosewho produce in competition with imports. The ‘temporary’ trade restraints adopted in response are noteasily removed once the exchange-rate fluctuation is reversed.

Dissatisfaction with current arrangements naturally prompts proposals for reform. Virtually all suchproposals seek to limit exchange-rate variability by establishing a system of multilateral currency bands (ortheir variants: crawling pegs and target zones).1 The member states of the European Union, among whomeconomic integration is particularly close, have established the European Monetary System (EMS), amultilateral exchange-rate grid, to limit the variability of their currencies. Countries as diverse as Finland,Sweden, Mexico, Argentina, and Estonia have pegged their exchange rates to the currencies of majortrading partners for various periods of time.

But the decision to peg the exchange rate is easier to issue than to enforce. In 1992–3, the EuropeanMonetary System erupted in a crisis that expelled two major participants, Italy and the United Kingdom,from its Exchange Rate Mechanism and forced the remaining countries to widen their fluctuation bandsfrom to 15 per cent. The Mexican policy of pegging the peso to the dollar came asunder in 1994, whenan attempt to realign led to a full-blown panic and financial collapse.

The implication is that pegging the exchange rate has become exceedingly difficult—some would sayimpossible for any length of time—in today’s world of liquid financial markets and quicksilver capitalflows. The resources commanded by currency traders far exceed the reserves of central banks. By forcing

governments seeking to defend the currency to raise interest rates to politically insupportable levels, themarkets have the capacity to demolish the peg at any time.

Things may have been different under the Bretton Woods System of pegged-but-adjustable exchangerates, but this was because controls limited international capital movements, sheltering currencies frommarket pressures. In today’s globalized capital markets, where capital controls have become exceedinglyexpensive to enforce, countries will inevitably gravitate toward greater exchange-rate flexibility.

The gold standard is the obvious challenge to this conventional wisdom. When it prevailed, currencieswere successfully pegged against one another despite the presence of open capital markets.2 Internationalfinancial transactions were unrestricted, and foreign lending and borrowing, as a share of global GNP,reached even higher levels than today.3 Yet exchange rates were stabilized within 1 per cent bands forextended periods, a record of stability that remains unparalleled even today.

Popular images of the gold standard reflect present problems as much as past realities. When the firstedition of this book appeared a dozen years ago, the appeal of the gold standard was its association withprice stability. Countries were just emerging from the high inflation of the 1970s. The fact that prices hadbeen little different in 1870 and 1913 recommended the system to those who valued price stability. In thecourse of the last ten years, inflation was brought under better control, and its threat was superseded in theeyes of many by the problem of exchange-rate instability. Correspondingly, the appeal of the gold standardtoday is as a mechanism for stabilizing exchange rates and smoothing the balance-of-payments adjustmentprocess.

In the textbook story, the gold standard worked smoothly because it was automatic. Each country’smoney supply was linked to its gold reserves, and balance-of-payments adjustment was accomplished byinternational shipments of precious metal. Each country being subject to the same gold standard discipline,the system brought about a de facto harmonization of policies and an admirable degree of exchange-ratestability.

Unfortunately, this vision of the gold standard, like the unicorn in James Thurber’s garden, is a mythicalbeast. Far from the normal state of affairs prior to the twentieth century, the gold standard prevailed on aglobal scale for barely a third of a century. The experience of the industrial economies was moresatisfactory than that of countries specializing in the production of primary products; international creditorshad happier experiences than debtors. The gold standard did not prevent the international transmission offinancial crises, nor did it preclude suspensions of convertibility. Discretionary actions by nationalauthorities featured prominently in the gold standard’s operation in both normal periods and times of crisis.

If we reject the gold standard myth, we must then come to terms with the reality. This is an enterprise towhich scholars from a number of different disciplines can contribute. In the work of economists we findmodels of the gold standard as a self-equilibrating system of markets. One class of models focuses on themechanisms by which balance-of-payments equilibrium and exchange-rate stability were maintained.Another analyzes how deflation stimulated mining activity and augmented the supply of monetary gold,stabilizing the price level. Such models are useful for checking the internal consistency of accounts of theoperation of the monetary system, but by their stylized nature they abstract from important aspects of itsstructure.

In the work of historians we find detailed studies of particular aspects of that structure. Some consider thegold standard’s impact on individual countries or on relations between them. Others focus on the role of centralbanks and even of particular central bankers. These studies have little in common with the gold standard myth.But while providing a wealth of institutional and historical detail on particular episodes and markets, theydiscourage efforts to generalize by virtue of the detailed and idiosyncratic nature of their accounts.

2 THE GOLD STANDARD IN THEORY AND HISTORY

In work by political scientists, finally, we find attempts to explain the emergence and operation of thegold standard in terms of interest-group pressures, institutions, and international political interactions.4

These are factors that make no appearance in the models of economists and play at best a subsidiary role inmost historians’ accounts of international monetary relations: for example, the extent of the electoralfranchise, conflicts between landed and industrial interests, and management of the gold standard system bythe leading commercial and financial powers.

The existence of three distinct literatures is frequently taken to indicate that our knowledge of the goldstandard is incomplete. In fact, most of the elements needed to paint a complete picture are at hand.Completing it requires only that we blend the contributions of economic theorists, historians, and politicalscientists. Like the blind man with the elephant, students of the gold standard have gained their impressionsfrom an awareness confined to parts of the beast. Theorists have sought to model the gold standard’soperation, but only occasionally have they drawn guidance from the work of historical scholars. Historianshave concentrated on particular instances of the gold standard’s operation. Political scientists havedownplayed the role of markets and historical happenstance in order to highlight political dynamics.

Interaction between these three sets of scholars and integration of these literatures are needed to providefresh insights. This volume therefore brings together selections on the international gold standard from theliteratures of economics, history, and political science. It is directed at students of all three disciplines in thehope that their understanding of the gold standard will be enriched and that they will acquire a new appreciationof the gains from intellectual trade.

The emergence of the international gold standard

The gold standard is frequently portrayed as the normal state of affairs prior to 1913. But in fact, it prevailedon a global scale only for a third of a century, from 1880 to 1914. Prior to that, currencies were generally basedon silver instead of gold or on a combination of the two metals. Britain was the principal exception, havingbeen on a full legal gold standard from 1821 and on a de facto gold standard from 1717, when Sir IsaacNewton, then Master of the Mint, set too high a silver price for the gold guinea. With the Mint price of silverlower than its international market price, Britain’s newly reminted full-bodied silver coins were quicklydriven from circulation.

During the Napoleonic Wars, the imperatives of war finance led to inflation and suspension ofconvertibility. Parliament brought this episode to a close by passing a law in 1819 that required the Bank ofEngland to make its notes redeemable in gold at the market price prevailing in 1821. This placed the Bank ofEngland in a peculiar position, since it was still a private institution, albeit one with special privileges andobligations, including the obligation to redeem its notes for specified quantities of gold. Its position wasfurther complicated by the Bank Charter Act of 1844, which divided the Bank into an Issue Departmentresponsible for backing the note circulation with gold, and a Banking Department to undertake commercialactivities and market intervention.

All the while, the United States and France operated bimetallic standards. Their Mints stood ready totransform specified quantities of gold or silver into coins of comparable value. In effect, these countries rantwo simultaneous commodity price stabilization schemes, pegging their currencies to those of countries onboth gold and silver standards, and thus implicitly pegging the gold-silver price ratio. Although the US hadto adjust Mint prices of the two metals from time to time to keep both metals in circulation, this arrangementwas remarkably successful at stabilizing the gold-silver price ratio at to 1 for more than three-quartersof a century (Flandreau 1995).

EDITORS’ INTRODUCTION 3

In the United States, the Coinage Act of 1792 established a Mint ratio of 15 ounces of silver to an ounceof gold, approximating the market price. But increases in Mexican and South American silver productionsoon caused the price of silver to fall to to 1. With gold undervalued at the Mint, silver was broughtthere to be coined, and gold was shipped abroad where its price was higher (the opposite of the effect ofIsaac Newton’s undervaluation of silver—cf. p. 4 above). As a result, through the first half of the 1830s theUnited States was effectively on a silver standard. The Coinage Act of 1834 raised the Mint ratio to 16 to 1in an attempt to restore gold coins to circulation. Gold discoveries in California and Australia and the steadyoutput of Russia’s mines then depressed the price of gold, causing silver to be exported and gold to becoined.5 The US was effectively placed on the gold standard until convertibility was suspended with theoutbreak of the Civil War.

While France’s experience was similar in many respects, its larger internal circulation of both gold andsilver insulated it from disturbances to the availability of the two metals. The Mint ratio of to 1 wasbelow the market ratio for many years after 1803, and gold was exported (in exchange for silver) until the1840s. Once mid-century gold discoveries augmented supplies of the yellow metal, gold came to dominatethe French monetary circulation. But, even at the height of the California gold rush, both gold and silvercoin continued to circulate (Flandreau, 1995).

While Britain was first to adopt the gold standard, her example was not followed until the second half ofthe nineteenth century. Fears of inflation due to gold discoveries deterred other nations (Sayers, 1933),6 anduntil the 1870s there did not exist a critical mass of gold standard countries to attract others to the system.Indeed, at mid-century the dominant direction of movement was away from the gold standard, not toward it.The possibility of inflation due to newly mined gold flowing into the coffers of central banks ledgovernments to suspend gold coinage. This was the response of Belgium, Switzerland, and the Netherlands,although larger countries like France and England hesitated to take this step for fear of further destabilizingthe international system.7

In the end, the anticipated inflation never materialized, due in part to the operation of the bimetallicsystem itself. As gold flowed toward the bimetallic countries, their silver flowed toward countries on silverstandards. Thus the impact of the gold rushes in California and elsewhere on the money stocks of thebimetallic countries was minimized by the operation of this ‘parachute effect’ .8

By 1860 it had become clear that the gold-silver exchange rate would not be significantly displaced bydiscoveries in California and elsewhere; in response, Belgium and Switzerland resumed coining gold. Theirresumption of bimetallism turned out to be the first step toward the creation of a global gold standard.

The expansion of Europe’s trade in the 1860s heightened the attractions of exchange-rate stability.Although France’s bimetallic system helped to stabilize the exchange rates linking the gold and silver blocs,a common basis for Europe’s currencies would have been even better for weaving together the continent’strade. In 1867 an international conference was held in Paris for the purpose of establishing a commonstandard. With gold now comprising the majority of the French monetary circulation and England on thegold standard, the yellow metal was the natural focus of negotiations. The bimetallic lobby remainedpowerful, however, and the 1867 conference failed to overcome its opposition and agree on internationalaction.

The outcome hoped for by many delegates, namely the establishment of an international gold standard,eventually resulted anyway from the unilateral actions of individual governments. Germany initiated theprocess in 1871. Although the country had previously derived some advantage from its silver standard intrade with Eastern Europe, by 1870 most of that region had suspended convertibility. The indemnityreceived in 1871–3 as victor in the Franco-Prussian War provided the resources needed to carry out a


currency reform. Germany established a gold-based currency unit, the mark, and used her indemnity topurchase about half of the gold needed for circulation.

The Germans sought to complete the process by selling their silver on world markets, taking advantage ofbimetallic France’s commitment to purchase it (and sell gold). Concerned not to aid its German rival, Parisresponded by limiting silver coinage. Without France to peg the gold-silver exchange rate, the solidity ofthe bimetallic bloc was shattered. The members of the Latin Monetary Union (Belgium, Switzerland, Italy,and Greece, along with France), as well as smaller silver standard countries (the Netherlands, Denmark,Norway, and Sweden), followed by limiting or suspending silver coinage and shifting to gold (Flandreau,1996b).

The more countries adopted the gold standard, the more attractive it became for the others. Gold soonbecame the monetary standard in virtually every European country.9 The international gold standard reachedacross the Atlantic in 1879, when the United States, putting civil war and reconstruction behind it, restoredgold convertibility, and into Asia when Japan followed. By the early 1900s most of the non-European worldhad gone on to gold; the principal exceptions were China, which remained on silver, portions of LatinAmerica with silver and bimetallic standards, and bimetallic Persia (Eichengreen and Flandreau, 1996).

Forms of the gold standard

There were important differences in how different countries operated their gold standards. Gold coincirculated only in France, Germany, the United States, Australia, South Africa, and Egypt. Other countriesissued token coin and paper currency convertible into gold. Some countries were on full gold standards,where gold convertibility was automatic (Britain, Germany, and the United States), while others operated‘limping’ gold standards, where gold convertibility was at the option of the authorities, who reserved theright to make use of the silver clause of their still officially bimetallic monetary statutes (France andBelgium).

One might also differentiate gold standards according to the ‘cover system’, which linked currency andcoin in circulation to the gold reserve. Countries with a ‘fiduciary system’ allowed the authorities to issue acertain quantity of unbacked currency (the fiduciary issue), while requiring remaining currency to be fullycollateralized with gold. This was the arrangement in Britain, Finland, Japan, Norway, and Russia. Countrieswith ‘proportional systems’, on the other hand, treated all currency alike but permitted the central bank tomaintain a ratio of gold reserves to currency issue of less than 100 per cent.10 This was the system usedby Belgium, Holland, and Switzerland. Germany, Austria-Hungary, Italy, and Sweden maintained hybridarrangements combining features of the two systems. In some countries additional flexibility was lent bypermitting the note issue to exceed the legal limit upon payment of a tax by the central bank or bypermitting reserves to fall below their legal minimum on the Finance Minister’s authorization.

Finally, the statutes and practices governing the composition of international reserves differed acrosscountries. In India, the Philippines, and much of Latin America, reserves took the form of claims oncountries whose currencies were convertible into gold. In Russia, Japan, Austria-Hungary, Holland,Scandinavia, and the British Dominions, some but not all international reserves were held in this form.Typically, such countries maintained a portion of their reserves in British treasury bills or bank deposits inLondon. If their liabilities were presented for conversion, the central bank or government could convert anequivalent quantity of sterling into gold at the Bank of England. Japan, Russia, and India were the largestcountries to engage in this practice; together they held nearly two-thirds of all foreign exchange reserves.


Price performance under the gold standard

John Stuart Mill (1865) described the mechanism by which the gold standard worked to preserve pricestability. He emphasized the link between money supplies and gold stocks and the tendency for the flow ofnewly mined gold to respond to changes in the price level. For example, as the world economy expandedand demands for money grew, downward pressure was placed on prices. With governments pegging theprice of gold in terms of domestic currency, the relative prices of other commodities fell. The increase in thereal price of gold then elicited additional supplies of the metal.11 The larger monetary gold stock thatresulted meant a larger money supply, limiting the downward pressure on prices (Barro, 1979).

Even Mill admitted that time was required for the mining response to occur: for instance, it took morethan twenty years of deflation, from 1873 to 1896, before gold discoveries in the Klondike and elsewherereversed the trend and inaugurated a period of inflation. Moreover, a number of economists emphasized therandomness of gold finds, arguing that chance discoveries dominated relative prices as a determinant ofgold production. Marshall (1923) for one was dubious that long-run price stability would be provided by asystem dependent on the ‘hazards of mining’.

Ultimately, this dispute over price performance can only be settled by comparisons between the goldstandard and alternative monetary regimes. Bordo (1981) compared the rate of inflation, unemployment,and income growth under the gold standard and after World War II for Britain and the United States. Hefound that the rate of change of wholesale prices was more moderate under the gold standard.12 Even therelatively low inflation rates under the gold standard can be cast in a less favorable light, however, sincethey result from averaging together two decades of deflation prior to 1896 with two subsequent decades ofinflation.13

Some would argue that it was not inflation or deflation but their variability that complicated decisionmaking.14 Bordo (1981) therefore computed the variability of inflation under the gold standard, BrettonWoods and the post-Bretton Woods float; he concluded that it is difficult to conclude in favor of any ofthese monetary regimes.15 In his 1993 article he extended these comparisons to seven industrial countries:again, the finding was that the standard deviation of inflation was no lower under the gold standard than inthe heyday of Bretton Woods (1959–71) or in the subsequent period of floating rates.16

It might be argued that only that proportion of the variation in prices that was unpredictable is relevantfor economic welfare. Had individuals been able to predict prices, they could have adapted accordingly andavoided misallocating resources. The historian’s problem then becomes to estimate predictability. Cooper(1982) attempts to infer price predictions from ex post real interest rates. He finds that under the gold standardex post real interest rates rose during deflationary periods and a fell during inflation as if individuals failedto anticipate price changes.17 Callahan (1994) estimates forecasting equations for prices and compares themean squared error of the forecast across monetary regimes. An advantage of this approach is that itconsiders prices directly rather than attempting to infer price forecasts from interest rates. Its weakness isthat it is based on an ‘as if’ assumption; it assumes that agents behaved ‘as if’ they had knowledge of theforecasting equation. Moreover, there are innumerable plausible specifications of the forecasting equation.For all these reasons, rankings of the gold standard and alternative monetary regimes in terms of pricepredictability remain inconclusive.

Exchange rates under the gold standard

In the absence of any clear-cut superiority in terms of price performance, the outstanding feature of the goldstandard would seem to be the extent of exchange-rate stability. But even this generalization should betaken with a grain of salt. A non-negligible number of countries abandoned the gold standard and allowed


their currencies to depreciate between 1880 and 1914. They included Italy, Portugal, Spain, Greece,Bulgaria, Argentina, Chile, and Mexico. Exchange-rate stability did not extend to China, El Salvador, orHonduras, which remained on silver throughout the period and whose currencies fluctuated in terms ofgold.

Exchange-rate adjustments were often the result of balance-of-payments crises. A country’s export pricescollapsed, or its capital imports dried up. As its balance of payments moved into deficit, the imminentexhaustion of its gold reserves forced the central bank to suspend gold convertibility (to abandon the policyof pegging the domestic currency price of gold). Often balance-of-payments crises were accompanied byfinancial crises, in which banks collapsed and stock and bond markets crashed. These panics spreadcontagiously across borders, transmitted at the speed of the international telegraph, provoking globalfinancial crises in 1884, 1890, 1893, and 1907.

And yet what is striking about the gold standard is that none of the major powers was forced to departfrom it for any extended period of time. This statement holds for North America, Western Europe, andBritain’s overseas Dominions and Commonwealth, as well as for the Austro-Hungarian and RussianEmpires after 1890. Throughout this area, exchange rates were impressively stable. Although there was noshortage of shocks to commodity and financial markets, disturbances to the balance of payments weredispatched without destabilizing exchange rates or otherwise undermining the solidity of the gold standardsystem.

This ease of adjustment should perhaps not be credited to the gold standard itself. It could have been nothingmore than a reflection of favorable underlying conditions. In this view causality ran not from the stability ofexchange rates to the relatively rapid growth of international trade and incomes; rather, the autonomousfactor was industrialization, which stimulated incomes and trade, and currency stability and smooth balance-of-payments adjustment were the consequences. This could have been the case insofar as favorableunderlying conditions lent countries the capacity to accommodate whatever disturbances came down thepike. In particular, the decline in spending needed to curtail the demand for imports and the fall in domesticprices needed to stimulate the demand for exports would have been relatively easy to bear when incomeswere growing.

Generations of scholars have insisted that there was more to it than this. How the adjustment processworked has thus become ‘the ritual question’ of students of the gold standard (to quote de Cecco, 1974).Three main approaches to it can be distinguished. One emphasizes the efficiency and automaticity of theadjustment mechanism. A second portrays the gold standard as a check on inflationary government financeand ascribes exchange-rate and balance-of-payments stability to the soundness of the public finances. Athird finds the key in the unique combination of policy credibility and flexibility that the gold standardallowed.

The efficiency view

The traditional explanation for ease of balance-of-payments adjustment (and, by implication, for thestability of exchange rates) is the efficiency and automaticity of the adjustment mechanism. The inevitablestarting point for any discussion of that mechanism is the price-specie flow model of David Hume (1752)—chapter 2 of this volume. The simplest way to understand Hume’s model is in terms of a stylized model of aworld economy with two commodities: a consumer good and gold. Consider the effect of a disturbance tothis economy, say a one-time addition to the home country’s stock of gold. At initial prices, there will be anexcess supply of gold and an excess demand for goods. Prices must adjust to restore equilibrium in bothmarkets. The assumption of the price-specie flow model is that transactions occur initially among domestic


residents. As they simultaneously attempt to sell gold and purchase consumer goods, the price of goodsrises (equivalently, the real price of gold falls). The price of consumer goods in terms of gold is now higherat home than abroad. Domestic residents have an incentive to obtain consumer goods from abroad wherethey are relatively cheap. Foreigners, for their part, have an incentive to obtain gold from the home country,where its price is low (in terms of consumer goods). A surplus of consumer goods is shipped to the homecountry, and a surplus of gold is shipped abroad. In the absence of adjustments on the production side(which we consider below), residents of the foreign country must reduce their spending on consumer goodsto make a surplus available for export. Residents of the home country, meanwhile, increase their spending,absorbing additional imports. In terms of the external accounts, the home country runs a balance-of-paymentsdeficit whose corollary is a loss of gold.

This description of the price-specie flow mechanism illustrates a point that is frequently overlooked ormisunderstood. It demonstrates that disputes over whether balance-of-payments equilibrium is restoredthrough adjustments in commodity markets or adjustments in gold markets are fundamentally misplaced.Adjustments in these two markets are two sides of the same coin. In a two-commodity world there existsonly one relative price, so a rise in one commodity price is the same as a fall in the other. When one marketclears, so must the other, as a result of Walras’ law. It is not meaningful to ask whether adjustment takesplace in one market or the other, or to debate whether the gold market or other markets were ‘moreimportant’ to the adjustment mechanism.

Hume intended his model as an analytical device rather than a description of reality. He was aware, forexample, that transactions do not occur first within and only then between nations. Arbitrage in both goldand consumer goods creates a powerful incentive for the maintenance of the international law of one price,as emphasized by McCloskey and Zecher (1982)—chapter 4 of this volume. Except for tariff barriers andtransport costs, there is no reason for prices to differ by wider margins across than within countries.

McCloskey and Zecher’s objection has some basis in fact, as students of the gold standard haveunderstood since the beginning of the twentieth century. The Harvard professor Frank Taussig and hisstudents, in a series of studies published early in the century, established that relative price of home andforeign goods (the international terms of trade) rarely behaved in the manner predicted by the price-specieflow model.18 Whale (1937) similarly observed that commodity-price equilibrium was maintained socontinuously that few if any relative price movements could be observed during the adjustment process.

This tendency to downplay the influence of the terms of trade led commentators to stress the role ofinterest-rate differentials and international capital flows in the elimination of payments imbalances.Britain’s Cunliffe Committee (1919), chapter 12 of this volume, in recommending that the gold standardshould be restored as soon as possible after World War I, described its operation in terms of a price-specieflow model extended to admit a role for interest-rate differentials and international capital flows. Tounderstand the implications of these extensions, it is useful to repeat our previous ‘thought experiment’, aone-time increase in the domestic gold stock under the assumption that transactions take place first amongresidents of the home country.19 The initial excess supply of gold now has as its counterpart an excessdemand for both consumer goods and financial assets. The domestic prices of goods and securities bothrise; the domestic gold price falls. Residents of the home country have an incentive to obtain their consumergoods and securities from abroad, where both are now cheap; foreigners have an analogous incentive toobtain their gold from the home country. The value of the home country’s net exports of gold equals thevalue of its net imports of consumer goods and securities. The trade deficit no longer equals the goldoutflow; now, the balance-of-payments deficit equals the international transfer of gold.20 It matters littlewhether we describe this process in terms of price or interest-rate effects, since saying that security pricesare higher at home than abroad following the gold discovery is the same as saying that interest rates are


lower at home than abroad. Capital will flow from the country where interest rates are low to where they arehigh, until security prices and interest rates are equalized internationally.

Early discussions of the role of capital movements in gold standard adjustment posited that this processtook time. Capital flowed gradually from low- to high-interest-rate countries until the interest differentialdisappeared. In parallel with McCloskey and Zecher’s critique of the assumption that price differentials incommodity markets were eliminated gradually, subsequent authors asked why capital should not flow withsufficient speed to eliminate the interest differential all but immediately. Dick and Floyd (1992) andCalomiris and Hubbard (1996) take this point to an extreme, arguing that interest differentials were negligibleand that the quick response of capital flows played a dominant role in adjustment.

The point of this excursion through alternative models of the balance-of-payments adjustment process isthat interpretations emphasizing the role of commodity prices, spending, interest rates, and capital flows arein fact compatible with one another. They simply focus on different nodes of the same interconnectedmarkets. Some emphasize international divergences in the prices of traded goods. Others stress theparallelism in commodity prices, focusing instead on changes in spending and interest rates (Whale, 1937;Scammell, 1965; Triffin, 1968). Still others argue that arbitrage was equally prevalent in commodity andasset markets, precluding persistent divergences across countries in either prices or interest rates; theyemphasize instead the importance of changes in spending (McCloskey and Zecher, 1976) or equilibratingcapital flows (Dick and Floyd, 1992).

Yet these authors should be seen not as presenting incompatible models but simply as attaching differentvalues to the same parameters. Their answers boil down to different answers to a common set of questions.How quickly did arbitrage equalize the prices of traded goods across countries? How efficient was arbitragein financial markets? Were securities denominated in different currencies better substitutes for gold or oneanother? The essential point is that these are all questions of how much weight to attach to different aspectsof the same adjustment process.

The discipline view

Economists have long argued that the obligation to maintain convertibility served as a check on inflationaryfinance. The key to price and exchange-rate stability, according to this view, was that sound public financeswere an integral part of the gold standard ideology. Governments ran balanced budgets, which insulatedcentral banks from pressure to purchase government debt and inject currency and credit into the economy.Indeed, adherence to the gold standard mandated this behavior, since persistent budget deficits which led tomonetization and inflation would have exhausted the central bank’s reserves and forced the abandonment ofgold convertibility.

In consequence, the gold standard became a signal of financial probity—a Good Housekeeping Seal ofApproval. Bordo and Rockoff (1996) show that governments were charged lower interest rates by themarkets when they adhered to the gold standard. This is consistent with the notion that investors were in thebusiness of evaluating the financial integrity of governments and took gold convertibility as an importantsignal. But it falls short of proving that this in turn led governments to avoid deficits. The very fact that somecountries (Portugal, Argentina) moved off gold suggests that the seal-of-approval mechanism worked lessthan perfectly.

Other authors have focused on monetary rather than fiscal discipline. In their view, whatever the behaviorof the fiscal authorities, the central bank saw itself as responsible for doing what was necessary to defend itsgold reserves. Central banks therefore adjusted the interest rates at which they made credit available with aneye toward the maintenance of external balance. If the exchange rate fell to the gold export point and


reserves began to decline, the central bank raised the discount rate (the rate at which it advanced money todiscount houses and other financial intermediaries).21 This increase in the cost of credit put upward pressureon interest rates and reduced the availability of finance economywide. It discouraged spending in interest-rate sensitive sectors, depressing inventory accumulation and fixed investment. As spending fell, the tradebalance strengthened. And, as financial conditions tightened, capital flowed in from abroad, attracted byhigher returns. The higher central bank discount rate thus helped the exchange rate to recover. Keynescalled this practice playing by the gold standard ‘rules of the game’.22

An appeal of this account was that it could explain the great anomaly of the price-specie flow model: thatgold flows on the predicted scale were rarely observed. A country experiencing a payments deficit had to adjustby cutting its spending; in the price—specie flow model, the mechanism prompting this was a gold outflow,which reduced the money supply and tightened financial conditions. But a central bank playing by the rulescould produce the same effect without waiting for significant gold movements. At the first sign of goldlosses, it would raise the discount rate, producing the requisite financial tightening without the need forsignificant gold flows.

The problem, it turned out, was that evidence of ‘rules-of-the-game’ behavior was weak. In a classicstudy, Bloomfield (1959) found that pre-World War I central banks violated those rules in the majority ofyears and countries he considered. Rather than draining liquidity from the market when their reservesdeclined (and augmenting it when they rose), they frequently did the opposite.

The modern synthesis: the gold standard as a credible target zone

Bloomfield’s important article underscored the discretion possessed by nineteenth-century central banks. Indoing so it identified a paradox: if central banks were committed to defending the gold parity, how couldthey at the same time behave in this way, violating the rules of the game?

The answer, modern analysts have emphasized, is that the gold standard implied not pegged exchangerates but rather a narrow band within which the exchange rate could move; and even this narrow bandendowed the authorities with significant monetary autonomy. In modern terminology, the gold standard wasnot a system of currency pegs; it was a set of target zones.

Explaining this requires us to describe two of the gold standard’s technical features: the gold points andgold devices. The nineteenth-century gold market was highly developed, but it operated subject totransaction costs. These costs meant that exchange rates could fluctuate within narrow margins, known asthe gold export and import points, without offering gold traders significant arbitrage profits. The easiest wayto see this, paradoxically, is to imagine a world free of transaction costs. The Bank of England stood readyto convert a pound sterling into an ounce of (11/12 fine) gold on demand. The US Treasury, for its part,committed to paying out an ounce of gold of equal purity for $4.86. The exchange rate between sterling andthe dollar was then locked at 4.86 to one. If it fell even slightly (toward 4.85), there would be an incentive tobuy gold from the Bank of England for £1, ship it to the US, sell it to the American Treasury for $4.86, andconvert those dollars into sterling on the foreign exchange market, where they now commanded more than£1 (£1.0021 to be precise). This would increase the demand for sterling and reduce the demand for dollarsuntil equilibrium was restored. Indeed, in the absence of transactions costs in international gold markets, theresponse would be instantaneous, and there would be no leeway for the exchange rate to move at all.

In practice, arbritragers had to pay to ship gold across the Atlantic, to insure it while in transit, and toborrow the funds needed for its purchase (or forego interest earnings on bank deposits in the meantime).Due to technological progress in shipping and other markets, these expenses declined from about percent of the value of the gold bought and sold in the 1840s to perhaps 0.5 per cent on the eve of World War


I.23 Thus, the dollar/sterling exchange rate could rise or fall by at most this amount before there was anincentive to engage in arbitrage and prevent the exchange rate from moving further. These upper and lowerbounds were the gold import and export points.

While the computation of transactions costs in trade in bullion is as old as international banking, goldpoint estimates made their official debut in September 1877, when The Economist began reporting them forthe leading financial centers. Its figures were based on the work of Ernest Seyd, a bullion dealer who hadauthored a study of international financial markets (Seyd, 1868). Seyd’s estimates were the basis forMorgenstern’s pioneering statistical study (Morgenstern, 1959), in which he claimed that the exchange ratesbetween the major financial centers had repeatedly strayed outside the bands defined by the gold points.This led some authors (viz. Clark, 1984) to conclude that the gold market was inefficient—that arbitragehad failed to prevent violations of the gold points. Officer (1986, 1989, 1996) explained away theseanomalies by challenging Morgenstern’s assumption that central banks always sold gold at the Mint price.The Bank of England, although obligated to buy gold at £3 17s 9d an ounce and to redeem its notes withgold coins whose mint price was £3 17s l0d, could also dispense worn coins that were worth significantlyless than their mint price. By paying out such coin, the Bank could manipulate the gold points.24 Exchangerates adjusted for these variations, according to Officer, never violated the band given by the gold points.25

The practice of paying out worn coin was an example of the so-called gold devices. In addition to payingout coin whose metallic content was eroded by wear, tear, and clipping, the central banks of countries likeFrance and Belgium on ‘limping gold standards’ could also redeem their notes in depreciated silver. Thus,while exchange-rate fluctuations between the currencies of countries on full gold standards were limited bythe gold import and export points, those vis-à-vis France and Belgium, whose central banks might pay outsilver instead, were potentially not constrained to equally narrow bands.26

Let us now return to the question of how discretionary monetary policy was compatible with goldconvertibility. The key, we now see, lay in the gold points and gold devices, which gave central banks roomfor maneuver. Governments could pursue, monetary policies which caused the exchange rate to depreciateso long as the latter did not violate the gold points and threaten the central bank with reserve losses.Interestingly, monetary autonomy might be quite significant, as Keynes (1930, Chapter 7 in this volume)was first to show. He provided an analysis of the tradeoff between the width of the band and the extent ofmonetary autonomy which uncannily anticipated Krugman’s (1991) model of exchange-rate target zonesand Svensson’s (1994) use of it to analyze the tradeoff between exchange-rate stability and monetaryindependence. The Keynes-Svensson analysis showed that even a narrow band endowed the authorities withsignificant capacity to buffer shocks. This suggests that the gold points and gold devices, by providingmacroeconomic room for maneuver, contributed importantly to the gold standard’s success.

An increase in domestic credit which caused the currency to depreciate would produce the lower interestrates desired by central banks only if, of course, the exchange rate was expected to recover subsequently.Imagine that this was so. Say the exchange rate fell toward the gold export point (where domestic currencywas sufficiently cheap that it was profitable to convert currency into gold, export the latter, and use it topurchase foreign exchange). As soon as the central bank began losing reserves, funds would flow in fromabroad in anticipation of the profits that would accrue to investors in domestic assets once the central banktook steps to support the exchange rate. Because there was no question about the authorities’ commitment tothe parity, capital flowed in quickly and in significant quantities.27 Interest rates tended to fall. Thedepreciation of the exchange rate was limited, leaving it well above the gold export point.

Thus, central banks could deviate from the ‘rules’ because their commitment to the maintenance of goldconvertibility was credible. Although it was possible to find repeated ‘violations of the rules’ over periodsas short as a year, over longer intervals central banks’ domestic and foreign assets moved together. Central


banks possessed the capacity to breach the rules of the game in the short run, in other words, because therewas no question about their obeying them in the long run.28 Knowing that the authorities would ultimatelytake whatever steps were needed to defend convertibility, investors shifted capital toward weak-currencycountries, financing their deficits even when their central banks temporarily violated the rules of the game.

The capstone of this new literature is Bordo and Kydland’s (1995) synthesis of the discipline anddiscretion views (Chapter 6 in this volume), which characterizes the gold standard as a contingent rule. Inthe target zone interpretation, capital flows tended to react in stabilizing ways, keeping the exchange ratewithin the band given by the gold points, so long as the markets were convinced of the authorities’commitment to the defense of gold convertibility. Bordo and Kydland argue that the same could even betrue when governments were forced to suspend convertibility temporarily and allow the exchange rate tofall out of that band, so long as the markets were convinced that they remained committed to resumingconvertibility at the previous rate once the exceptional disturbance motivating the temporary suspension hadpassed. So long as the markets were convinced of the government’s commitment to gold convertibility, theywould bid up the currency in anticipation of the eventual restoration of convertibility at the traditional parity,stabilizing the exchange rate. This mechanism only operated, of course, provided that the contingency inresponse to which suspension took place was exceptional and independently verifiable—a war or a naturaldisaster, for example—and so long as the authorities remained committed to restoring convertibility as soonas that contingency passed. But if these conditions were met, the gold standard worked as a contingent rule,to be obeyed in normal times but which could be disregarded in crises without damaging the authorities’credibility.

The roots of credibility

Everything depended, then, on the credibility of the authorities’ commitment to the maintenance (orrestoration) of convertibility. The key to understanding how the gold standard worked, therefore, is toidentify the roots of that credibility.

In the late nineteenth century, the credibility of governments’ commitment to the maintenance of goldconvertibility rested on foundations grounded in economics, diplomacy, politics, and ideology. That thedecades leading up to 1913 comprised a period of international peace, compared to what came before andafter, minimized shocks to government budgets and international payments.29 The absence of a clearlyarticulated theory linking central bank policy to the state of the economy limited the pressure on themonetary authorities to pursue output and employment goals. The same was true of the absence of aKeynesian theory linking budget deficits to aggregate demand and the level of employment. Thus, theideology of hard money extended to the maintenance of budget balance, limiting the scope for recklessfiscal policy to destabilize the external accounts. In the countries at the core of the system, rulingparliamentary elites held a substantial fraction of their wealth in government bonds and thus favored pricestability. Those who valued other policy targets—workers concerned with unemployment, for example—rarely possessed the vote.30 For all these reasons, political support for the gold standard was pervasive.Political scientists have thus come to characterize it as a ‘diffuse regime’ that did not rely on a formalsystem of rules and restraints but rather on broad-based support of the society and polity.31

Admittedly, there were always skeptics of the gold standard’s merits.32 Nor could ruling elites alwaysafford to ignore potential political threats from populist and socialist forces. In France, for example, the1878 Plan Freycinet providing for massive infrastructure investments aimed at increasing the supply of jobsin order to keep more radical French socialists at bay. It led to a deterioration in the public accounts, raisingthe government debt as a share of GDP from 80 to almost 100 per cent. In Germany, civil servants pressed


for increased public spending on retirement funds, and neither the government budget nor the monetarysystem was entirely insulated from the effects. In the United States, farmers strapped by falling agriculturalprices lobbied for the restoration of free silver coinage. Their populist movement came close to electing thepresidential candidate William Jennings Bryan and pushing the country off the gold standard in 1896(Frieden, 1996, Chapter 11 in this volume).

Still, in none of these countries did anti-gold standard agitation succeed. What pressure existed for thepursuit of other policy goals was exceptional and, ultimately, limited. At the gold standard’s European andNorth American core, its political, ideological, and economic underpinnings sufficed to sustain the system.When push came to shove and the authorities in these countries had to choose between interest-rateincreases to keep the gold standard from collapsing and interest-rate reductions to stimulate production, theynever hesitated to opt for the former (Sayers, 1936).

The situation was different elsewhere. In Latin America, for example, credibility was less complete.Inflationist agitation among farmers, à la the US, combined with unstable political systems, shaky finances,and external shocks to drive governments on to depreciated paper. But the very contrast with Europehighlights what was unique and distinctive about the gold standard system. At its core—in the industrialnations of Northern Europe and their overseas dependencies—the commitment to gold convertibilityremained paramount.

But, as time would show, this constellation of circumstances was unique. Neither it nor the commitmentto gold would long survive the coming of the twentieth century.

From the prewar gold standard to the interwar gold exchange standard

World War I, like the Napoleonic Wars a century before, threw a wrench in the works of this carefullybalanced international monetary mechanism. The imperatives of war led to unbalanced budgets, exportcontrols, and inflation incompatible with a fixed domestic currency price of gold. Virtually every countrybut Britain and the United States suspended the convertibility of currency into gold and placed barriers inthe way of international gold shipments.33 Budget deficits persisted beyond the conclusion of hostilities,although it was the financial needs of reconstruction rather than the war itself that now drove publicspending. With the return to peacetime conditions, it was no longer feasible to control prices and foreignexchange transactions. Rates of inflation and currency depreciation accelerated, forcing even the UK tosuspend gold convertibility.

Not until 1925 did Britain stabilize financial conditions sufficiently to restore convertibility at the prewarrate. The members of her Commonwealth and most of the smaller countries of Continental Europe followedsuit. But France, Belgium, Italy, and Germany found it impossible to turn back the clock. Given the lengthof time that inflation had run out of control, restoring the old gold parity would have meant reducing pricesby more than half (often significantly more), which was impossible economically and politically. Francesettled for six months of deflation before stabilizing at a devalued parity at the end of 1926. In Germany,inflation so debased the currency as to require monetary reform; the Weimar Republic established a newcurrency unit before restoring gold convertibility.

Superficially, normalcy had been restored by the end of 1926, with the major countries back on gold.Except for the withdrawal of gold coin from circulation (in most countries, residents had been required to turnit over to the authorities to help finance the war, and the latter were not inclined to turn it back), thesystem’s structure had not obviously changed. But subsequent experience bore little resemblance to thatunder the gold standard of prewar years. Balance-of-payments adjustment was anything but smooth;discount-rate increases now threatened recession and failed to attract gold inflows.


The gold standard as an international system survived for barely five years. Starting in 1931 one countryafter another, beginning with the UK itself, was forced to abandon convertibility and allow its currency tofloat downward. By 1936 the transition to floating was complete, and the international gold standard was nomore.

There is no shortage of explanations for this unsatisfactory experience. (A catalog of popularinterpretations and a critique of the interwar gold standard’s operation is the 1931 Report of Britain’sMacmillan Committee—chapter 13 of this volume.) One strand of thought emphasizes that the interwarsystem was more fragile and susceptible to shocks than its prewar predecessor. Central banks now held asubstantial share of their reserves in the form of interest-bearing foreign exchange (for example, the bondsof governments whose currencies were themselves convertible, mainly sterling and the dollar). Anyquestion about the stability of sterling or the dollar could lead to a massive liquidation of foreign reserves,as central banks scrambled to replace them with gold before suffering capital losses due to foreigndevaluation. By their actions, they might cause the reserve base of the system to implode and deflationarypressure to ramify internationally.

In addition, the prewar pattern of trade relations had been shattered by the creation of new nationalborders in Central and Eastern Europe (due mainly to the breakup of the Austro-Hungarian Empire) and bythe proliferation of tariffs, on which governments new and old relied to raise revenues. International flows offinancial and physical capital did not fit together as neatly as before 1913. Before World War I, Britain, themajor exporter of financial capital, was also a major exporter of capital goods. This stabilized the paymentsof the country at the center of the international system. When London lent abroad, the countries on thereceiving end used the funds to purchase machinery, equipment, and ships from the industries of Lancashireand Clydeside, automatically balancing Britain’s external accounts.

After World War I, however, Britain was no longer the preeminent lender, that role having been assumedby the United States. The borrowers no longer relied on one country for their imports of machinery andcapital equipment, as members of her Commonwealth had relied on Britain before 1913. Thus, merchandiseexports no longer financed foreign lending to the same extent.

While all these points are relevant, there is a tendency to exaggerate their novelty. None was really aradical departure from the prewar status quo. The practice of holding of foreign exchange reserves waswidespread under the prewar gold standard (Lindert, 1969); paper currency and bank deposits alreadyaccounted for 90 per cent of the money supply before 1913, and gold only 10 per cent (Triffin, 1968).Countries other than Britain had relied heavily on import tariffs for revenue and other purposes. WhileLondon accounted for perhaps half of all overseas lending in the four decades ending in 1913, Paris, Berlin,Amsterdam, and, toward the end of the period, New York had also played significant roles. And the linkagefrom exports of financial capital to exports of physical capital, however important for Britain, was hardlyevident in these other countries.

A second set of explanations emphasizes the decline of the international leader, or ‘hegemon’, Britain,and reluctant acceptance of this mantle by the United States. Keynes, that famous phrase-maker, coined thename for this phenomenon too, referring to Britain, and the Bank of England in particular, as the ‘conductorof the international orchestra’. Before 1913, when credit conditions were overly lax and the Bank ofEngland raised its rate, the impact on international financial markets was so profound that other centralbanks had no choice but to follow suit. This follow-the-leader behavior brought about a de factoharmonization of policies worldwide. In effect, the Bank used its discount rate as the conductor’s baton.With British trade and finance accounting for such large shares of the respective world totals, what wasgood for Britannia was good for the international system.


The decline in Britain’s commercial and financial leverage over the course of World War I and throughthe 1920s, coupled with the reluctance of the United States to accept the maestro’s baton, allowed otherplayers’ eyes to wander from the podium. Central banks and governments no longer played harmoniously.The climax of this cacophony was the collapse of the gold standard in 1931.

Charles Kindleberger (1973) generalized this interpretation of the destabilizing effects of the absence of ahegemon. Before World War I, he argued, Britain had provided an open market for the distress goods ofother countries. Her lending fluctuated countercyclically, rising in periods of global slowdown, falling inbooms. She acted as international lender of last resort, providing capital to foreign central banks andgovernments when the stability of the gold standard was threatened. Between the wars, in contrast, USlending was procyclical, and her market was blockaded. Moreover, the US failed to acknowledge the needfor an international lender of last resort. When financial crises engulfed the gold standard in 1931, thefailure of the United States to provide international-lender-of-last-resort facilities removed the only fireextinguisher capable of dousing the flames.

It is no surprise that this interpretation was advanced by one of the leading historians of internationalmonetary affairs after World War II, a period when the dominance of US policy was overwhelming.34 But,as a matter of historical fact, it is unclear that the stabilizing influence of Britain had been so dominant priorto 1913. Often it was Britain herself that was subject to international financial strains. In 1890 she was theinternational borrower of last resort (obtaining support from the Bank of France and the Russian StateBank, among others) not the international lender. Often the Bank of England had to alter her discount rate toconform to international financial conditions rather than leading them.35

As these examples illustrate, international cooperation had also contributed importantly to the prewar oldstandard’s success. Scammell (1965) speaks of the growth of an international financial fraternity which wasmanifested in cooperation among the central banks of the leading countries. That cooperation, whileepisodic, was critical in times of crisis.36 But it took place only under very specific circumstances, asemphasized by Flandreau (1997). Its extension was motivated by selfish interests at least as much as by anyrecognition of the existence of common goals. While apparent between 1890 and 1910 (years known as ‘theheyday of the international gold standard’), cooperation was much less extensive before and after.

During the interwar period, in any case, the obstacles to international cooperation were formidable. And,given the difficult financial situation bequeathed by World War I, the need for it was greater than before.Disputes over war debts and reparations, and more generally the fact that the shadow of war was never farremoved, spoiled the climate for cooperation. The Bank for International Settlements, the logical vehicle forlast-resort lending, was disabled by having been created to manage the transfer of German reparations. Theideological underpinnings of the prewar gold standard no longer carried the same force. Proto-Keynesianideas surfaced in a growing number of places. Policy makers in different countries interpreted the economicmalaise in different ways and prescribed different policy responses, rendering concerted international actionall but impos sible. Increasingly influential special interest groups emphasized conflicts between thepolicies needed to advance cooperation with other countries and those required to address domestic problems.

As this last point reveals, perhaps the single most important difference between the prewar and interwarstandards was the credibility of the commitment to gold. Before the war, there was no question thatconvertibility was the preeminent goal of policy. The 1920s saw a growing consciousness of the problem ofunemployment: governments and trade unions gathered and published statistics on its prevalence, many forthe first time. Keynes and others articulated theories of the connections between central bank policy and theeconomy. Interest groups lobbied for outputand employment—friendly policies. Parliamentary labor partiesgave voice to those concerned with unemployment. The extension of the franchise made it costly forgovernments to neglect their views.


For all these reasons, when the exchange rate fell to the gold export point and a rise in interest rates wasrequired, it was no longer clear that the central bank had political backing. This might render it reluctant toraise rates when employment was growing slowly. Capital, rather than flowing in, in anticipation of theexchange rate’s subsequent recovery, might flow out, in anticipation of devaluation. The conflict betweeninternal and external balance was heightened, and with it the dilemma facing the monetary authorities.Between 1925 and 1935, they had more occasion to demonstrate their resolve than in the four decadespreceding 1913. But, in the newly politicized conditions of the interwar period, they failed the test. Thatfailure led ultimately to the collapse of their laboriously reconstructed gold standard.

The Bretton Woods gold-dollar standard and after

Thus, a variety of economic and political changes, foremost among them the declining credibility of thecommitment to convertibility, had by the second half of the 1930s rendered the gold standard a thing of thepast. For similar reasons, the politicization of monetary policy making and pressure for the pursuit of policygoals other than exchange-rate stability make it unlikely that anything resembling a global gold standardwill be reestablished in the foreseeable future. Countries may have reformed their central bank statutes tobuttress the independence of monetary policy makers, but even the most independent central banks are stillaccountable to a national parliament or legislature, which can pass laws limiting their autonomy and overridingtheir decisions. Monetary policy today is embedded in a political environment, inevitably limiting thecredibility of officials’ commitment to a particular target.

Despite these changes, the gold standard continues to cast a long shadow. The Bretton Woods System thatgoverned international monetary relations for a quarter of a century after World War II is commonlyreferred to as a gold-dollar standard: while countries other than the US pegged their currencies to the dollarwithin bands roughly twice as wide as those demarcated by the old gold points, the US as the reserve-currency country continued to peg the dollar to gold. This system differed from the gold standard by theprevalence of controls on capital movements and the option countries possessed to realign, but theinspiration drawn from the gold standard was clear.37 While the monetary role of gold receded with thecollapse of Bretton Woods in 1971–3, as recently as 1994 the chairman of the Federal Reserve Board stillcited the price of gold as one of the indicators of inflationary pressure he considered when framingmonetary policy.

As noted above, a growing number of developing and semi-industrialized countries have moved towardgreater exchange-rate flexibility. Only the member states of the European Union have been moving in theother direction, toward monetary union. While a monetary union which eliminates all scope for exchange-ratechanges might be thought to resemble the gold standard in important respects, it is in fact very different.Monetary union eliminates not just exchange-rate variability, a la the gold standard, but the exchange rateitself and therefore the very issue of exchange-rate credibility.38 It not only limits monetary autonomy to thatconsistent with keeping the exchange rate within a narrow band (like that given by the gold points); rather,in a monetary union the monetary autonomy of the constituent states is eliminated entirely.

A more germane analogy is with the currency boards of countries like Estonia, Hong Kong, Lithuania,and Argentina, under which their central banks are required, by statute or constitution, to convert thedomestic currency into fixed amounts of foreign exchange, thereby pegging the exchange rate irrevocably.But countries adopting currency—board arrangements are, without exception, in extreme and exceptionalcircumstances. A demonstrable inability to control monetary policy, as in Argentina, or a highly unstabledomestic and foreign economic environment, as in Estonia at the time of collapse of central planning, arerequired to justify the adoption of such extreme policies. The exceptional nature of these circumstances


suggests that ‘normal’ countries are generally unwilling to do the same and that countries currentlyoperating currency boards will forsake them for more discretionary systems sometime in the future. Intoday’s democratic societies, in other words, there is little prospect of the reestablishment of the gold standardor a gold-standard-like system, for better or for worse.

Notes

1 Examples include the recommendation of the Bretton Woods Commission (1994) for closer policy coordinationleading to the stabilization of exchange rates, and Williamson’s (1985) blueprint for a system of exchange-ratetarget zones among the leading industrial countries, echoed recently in Bergsten and Henning (1996).

2 The basic features of a gold standard are interconvertibility between money and gold at a fixed official price andthe freedom for individuals to import and export gold. An international gold standard exists when a number ofcountries adhere to these principles.

3 See Feis (1930), Fishlow (1985), and Bayoumi (1990).4 The locus classicus of political-economy analysis of the gold standard is Polanyi (1944). For recent

contributions, see in addition to the chapter by Frieden the important book by Gallarotti (1995). The resurgence ofinterest among political scientists in the operation of the gold standard is another of the new developments in theintervening period between the first and second editions of this book. In part, it can be understood as a product ofthe early 1980s literature on international regimes (Krasner, 1983).

5 By 1851, a silver dollar was worth 104 cents on the open market, and silver was rarely used in transactions.6 One less-developed economy with pervasive public finance problems, Portugal, took advantage of the surge in

gold output to reestablish gold convertibility. But given the underdeveloped finances of the Portuguese nation, itis not surprising that this country, ‘first to join the gold standard’ in 1854 (Reis, 1996), would also be among thefirst to leave it, in the 1890s.

7 In addition, reform was effectively opposed by bankers who derived handsome profits from gold-silver arbitrage.8 The terminology is from Chevalier (1859).9 The principal exceptions were Spain, Russia, and Greece. Spain, however, effectively suspended silver coinage.

Russia and Greece, while formally remaining on the bimetallic standard for some time, in practice hadinconvertible currencies. Austria-Hungary and Italy did not legally adopt gold convertibility but from the turn ofthe century pegged their currencies in terms of gold.

10 Although the central banks of Finland, Germany, Italy, Japan, Sweden, and even Great Britain did not have amonopoly of note issue in the final decades of the nineteenth century, in all of these countries the circulation ofother bank notes was small and declining.

11 To put the point another way, the falling price level implied lower costs of production for the gold-miningindustry. The value of its output remained unchanged, of course, since governments pegged the domestic-currency price of gold. In response, additional resources were allocated to that activity. In addition, to the extentthat deflation caused the price of jewelry to fall relative to that of coin, bracelets and rings were presented at theMint to be coined, increasing the quantity of money in circulation and moderating the downward pressure onprices.

12 In contrast, comparisons of per capita income growth do not suggest any obvious superiority over recentmonetary arrangements; if anything, the opposite is true.

13 Bordo (1993) provides more comprehensive comparisons for seven countries which point to broadly similarconclusions.

14 Similarly, it might be argued that the operation of the monetary standard has relatively little impact on theaverage rate of income growth but important implications for its stability.

15 For Britain, Bordo finds that the standard deviation of prices was lowest under the gold standard, but for the USit was not. In both countries, rates of growth of real income were more variable under the gold standard.

16 The same is true of the standard deviation of output.


17 A limitation of this approach is that no attempt is made to correct for other factors which may have caused shiftsin ex ante real interest rates. Barsky and DeLong (1991) estimate more sophisticated interest-rate models: theyconclude that short-run price fluctuations under the gold standard were largely unpredictable, but that investorslearned with time to anticipate price level trends.

18 See for example Taussig (1927).19 As for the restrictive assumption, nothing essential depends on what is essentially an expository device.20 The payments deficit is the sum of the trade deficit and the capital outflow, capital outflow being another name

for domestic purchases of foreign securities.21 In addition, they could alter the terms of discounting (broadening or limiting the eligibility of different classes of

bills) or announce the rationing of discounts. Increasingly, central banks intervened in the market, using ‘openmarket operations’, to back up the effects of discount-rate changes (to make the rate ‘effective’), sellingsecurities and draining liquidity from the market at the same time they raised the discount rate. Often securitysales were undertaken in conjunction with repurchase agreements. Some central banks like the Bank of Francealso intervened on the foreign exchange market. But the discount rate was the principal instrument of the Bank ofEngland throughout the gold standard years.

22 His first use of the phrase may have been in ‘The Economic Consequences of Mr Churchill’ (1925).23 Figures are from Officer (1996, p. 182). The total spread between gold points is equal to two times the figures

reported here.24 The degree of wear and tear had to be smaller than that required for the coins to be legal tender, which placed an

upper limit on the price at which the Bank could sell gold.25 Some of Seyd’s formulae had acknowledged that the Bank of England was not committed to sell gold at the Mint

price but only to pay out Sovereigns. But his 1868 book was published toward the end of a 20-year period duringwhich the Bank had consistently sold gold at the Mint price, leading him to emphasize that practice.

26 Some central banks made only limited use of their wide fluctuation bands. For example, it was thought that theBank of France should not more than double the size of its fluctuation band relative to that of countries on full goldstandards.

27 Econometric evidence documenting these relationships is supplied by Olivier Jeanne (1995).28 This is John Pippinger’s (1984) characterization of Bank of England discount policy in this period, for which he

provides econometric evidence. 29 Still, World War I loomed increasingly ominously on the horizon after the turn of the century, signalling clearly

that this period was drawing to a close.30 The exceptions, such as the United States and France, where male suffrage was universal, were also places where

the operation of the gold standard was disputed and where its structure was subject to exceptional provisions.31 See for example Gallarotti (1995).32 The English economist Ralph Hawtrey, for example, emphasized the impact of central bank policy on output and

investment. See Hawtrey (1913).33 Even the British authorities used red tape and moral suasion to discourage gold exports.34 To be precise, the US is credited with having stabilized the international system at the outset, notably through the

extension of the Marshall Plan, before destabilizing it subsequently, through the pursuit of inadequatelydisciplined monetary and fiscal policies.

35 This was in order to render her discount rate ‘effective’-that is, to keep it in sufficient touch with market rates soas to sustain a reasonable volume of business. The Bank’s continual battles to render its rate effective are thesubject of Sayers (1957).

36 As emphasized in Eichengreen (1992).37 Problems with the composition and volume of reserves under the post-World War II gold-dollar standard are the

subject of the chapter by Gilbert. The concluding chapter by Eichengreen emphasizes the role of capital controls.38 Issues of the credibility of the European Central Bank’s commitment to the pursuit of price stability will remain,

of course, and with it questions about the stability of the euro vis-à-vis the currencies of other regions andcountries.


References

Barro, Robert J. (1979), ‘Money and the Price Level Under the Gold Standard,’ Economic Journal, 89, 13–33.Barsky, Robert and J.Bradford DeLong (1991), ‘Forecasting Pre-World War I Inflation: The Fisher Effect and the Gold

Standard’, Quarterly Journal of Economics, 106, 815–36.Bayoumi, Tamim (1990), ‘Saving-Investment Correlations: Immobile Capital, Government Policy, or Endogenous

Behavior?’ IMF Staff Papers, 37, 360–87.Bergsten, C.Fred and C.Randall Henning (1996), Global Economic Leadership and the Group of Seven, Washington,

DC, Institute for International Economics.Bloomfield, Arthur (1959), Monetary Policy Under the International Gold Standard, 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.——(1993), ‘The Bretton Woods International Monetary System: A Historical Overview’, in Michael D.Bordo and

Barry Eichengreen (eds), A Retrospective on the Bretton Woods System: Lessons for International MonetaryReform, Chicago, University of Chicago Press, pp. 3–108.

Bordo, Michael D. and Finn Kydland (1995), ‘The Gold Standard as a Contingent Rule’, Explorations in EconomicHistory, 32, 423–64.

Bordo, Michael D. and Hugh Rockoff (1996), ‘The Gold Standard as a Good Housekeeping Seal of Approval’, Journalof Economic History, 56(2), 389–428.

Bretton Woods Commission (1994), Bretton Woods: Looking to the Future, Washington, DC, Bretton WoodsCommission.

Callahan, Colleen M. (1994), ‘The Nineteenth-Century Silver Movement and Aggregate Price Uncertainty’, in ThomasWeiss and Donald Schaeffer (eds), American Economic Development in Historical Perspective, Stanford, StanfordUniversity Press, pp. 223–40.

Calomiris, Charles and Glenn Hubbard (1996), ‘International Adjustment under the Classical Gold Standard: Evidencefor the US and Britain, 1879–1914’, in Tamin Bayoumi, Barry Eichengreen, and Mark Taylor (eds), ModernPerspectives on the Gold Standard, Cambridge, Cambridge University Press, pp. 189–217.

Chevalier, Michel (1859), On the Probable Fall of Gold: The Commercial and Social Consequences which may Ensue,and the Measures which it Invites, With a Foreword by Richard Cobden, New York, D.Appleton and Company.

Clark, T.A. (1984), ‘Violations of the Gold Points, 1890–1908’, Journal of Political Economy, 92, 791–823.Clarke, Stephen V.O. (1967), Central Bank Cooperation, 1923–1931, New York, Federal Reserve Bank of New York.Cooper, Richard (1982), ‘The Gold Standard: Historical Facts and Future Prospects’, Brookings Papers on Economic

Activity, I, 1–45.De Cecco, Marcello (1974), Money and Empire: The International Gold Standard, London, Blackwell.Dick, Trevor J.O. and John Floyd (1992), Canada and the Gold Standard: Balance of Payments Adjustment, 1871–

1913, Cambridge, Cambridge University Press.Dutton, John (1984), ‘The Bank of England and the Rules of the Game Under the International Gold Standard: New

Evidence’, in Michael D.Bordo and Anna J. Schwartz (eds), A Retrospective on the Classical Gold Standard,Chicago, University of Chicago Press, pp. 173–202.

Eichengreen, Barry (1992), Golden Fetters: The Gold Standard and the Great Depression, 1919–1939, New York,Oxford University Press.

Eichengreen, Barry and Marc Flandreau (1996), ‘The Geography of the Gold Standard’, in Jorge Braga de Macedo,Barry Eichengreen, and Jaime Reis (eds), Currency Convertibility: The Gold Standard and Beyond, London,Routledge, pp. 113–43.

Einzig, Paul (1962), The History of Foreign Exchange, London, Macmillan.Feis, Herbert (1930), Europe: The World’s Banker, New Haven, Yale University Press.Fishlow, Albert (1985), ‘Lessons from the Past: Capital Markets During the Nineteenth Century and the Inter-War

Period’, International Organization, 39, 383–439.


Flandreau, Marc (1995), L’or du monde: la France et la stabilité du système monétaire international, 1848–1873, Paris,L’Harmattan.

——(1996a), ‘Adjusting to the Gold Rush: Endogenous Bullion Points and the French Balance of Payments: 1846–1870’, Explorations in Economic History 33, 417–39.

——(1996b), ‘The French Crime of 1873: An Essay on the Emergence of the International Gold Standard 1870–1880’,Journal of Economic History 56, 862– 97.

——(1997), ‘Central Bank Co-operation in Historical Perspective 1829–1914: A Skeptical View’, Review of EconomicHistory (forthcoming).

Flood, R.P. and P.Garber (1984), ‘Gold Monetization and Gold Discipline’, Journal of Political Economy, 92(1),90–107.

Ford, A.G. (1962), The Gold Standard, 1880–1913: Britain and Argentina, London, Oxford University Press.Frieden, Jeffry (1996), ‘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 withComplexity in the International System, San Francisco, Westview Press, pp. 137–62.

Gallarotti, Giulio (1995), The Anatomy of an International Monetary Regime: The Classical Gold Standard 1880–1914,New York, Oxford University Press.

Gregory, T.E. (1932), The Gold Standard and its Future, New York, Dutton.Hawtrey, Ralph (1913), Good and Bad Trade, London, Constable.Jeanne, Olivier (1995), ‘Monetary Policy in England 1893–1914: A Structural VAR Analysis’, Explorations in

Economic History, 32, 302–26.Keynes, John Maynard (1925), The Economic Consequences of Mr Churchill, London, Macmillan.——(1930), A Treatise on Money, Vol. II: The Applied Theory of Money, London, Macmillan.Kindleberger, Charles (1973), The World in Depression, 1929–1939, Berkeley, University of California Press.——(1985), Keynesianism Versus Monetarism and Other Essays in Financial History, Boston, Allen & Unwin.Krasner, Stephen (1983), International Regimes, Ithaca, Cornell University Press.Krugman, Paul (1991), ‘Target Zones and Exchange Rate Dynamics’, Quarterly Journal of Economics, 106, 669–82.Lindert, Peter B. (1969), ‘Key Currencies and Gold, 1900–1913’, Princeton Studies in International Finance, No. 24,

International Finance Section, Department of Economics, Princeton University.McCloskey, Donald and J.Richard Zecher (1976), ‘How the Gold Standard Worked 1880–1913’, in Jacob Frenkel and

Harry G.Johnson (eds), The Monetary Approach to the Balance of Payments, London, Allen & Unwin,pp. 184–208.

——(1984), ‘The Success of Purchasing Power Parity: Historical Evidence and Its Implications for Macroeconomics’,in Michael D.Bordo and Anna J.Schwartz (eds), A Retrospective on the Classical Gold Standard, Chicago,University of Chicago Press, pp. 121–72.

Machlup, Fritz (1964), International Payments, Debts and Gold, New York, Charles Scribners Sons.Marshall, Alfred (1923), Money, Credit and Commerce, London, Macmillan.Mill, J.S. (1865), Principles of Political Economy, reprinted A.M.Kelly, New York, 1961.Morgenstern, Oskar (1959), International Financial Transactions and Business Cycles, Princeton, Princeton University

Press.Officer, Lawrence H. (1986), ‘The Efficiency of the Dollar-Sterling Gold Standard, 1890–1908’, Journal of Political

Economy 94, 1038–73.——(1989), ‘The Remarkable Efficiency of the Dollar-Sterling Gold Standard, 1890–1906’, Journal of Economic

History, 49, 1–41.——(1996), Between the Dollar-Sterling Gold Points, Cambridge, Cambridge University Press.Pippinger, John (1984), ‘Bank of England Operations, 1893–1913’, in Michael D. Bordo and Anna Schwartz (eds), A

Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press, pp. 203–33.Polanyi, Karl (1944), The Great Transformation, New York, Rinehart & Company.Reis, Jaime (1996), ‘Portugal: First to Join the Gold Standard, 1854’, in Jorge Braga de Macedo, Barry Eichengreen,

and Jaime Reis (eds), Currency Convertibility: The Gold Standard and Beyond, London, Routledge, pp. 159–81.


Sayers, Richard S. (1933), ‘The Question of the Standard in the 1850s’, Economic History, Economic JournalSupplement, 2, 575–601.

——(1936), Bank of England Operations 1890–1914, London, P.S.King.——(1957), Central Banking After Bagehot, Oxford, Clarendon Press.Scammell, W.M. (1965), ‘The Working of the Gold Standard’, Yorkshire Bulletin of Economic and Social Research,

17, 32–45.Seyd, E. (1868), Bullion and Foreign Exchanges, Theoretically and Practically Considered, Followed by a Defence of

the Double Valuation, London, Effingham Wilson.Svensson, Lars E.O. (1994), ‘Why Exchange Rate Bands? Monetary Independence in Spite of Fixed Exchange Rates’,

Journal of Monetary Economics, 33, 157–99.Taussig, Frank (1927), International Trade, New York, Macmillan.Triffin, Robert (1968), Our International Monetary System: Yesterday, Today and Tomorrow, New York, Random

House.Viner, Jacob (1937), Studies in the Theory of International Trade, New York, Harper and Row.Whale, P.B. (1937), ‘The Working of the Prewar Gold Standard’, Economica, 4, 18–32.Williamson, John (1985), The Exchange Rate System, Policy Analyses in International Economics No. 5, Washington,

DC, Institute for International Economics, revised edn (1st edn 1983).Yeager, Leland B. (1969), ‘Fluctuating Exchange Rates in the Nineteenth Century: The Experiences of Russia and

Austria’, in Robert Mundell and Alexander Swoboda (eds), Monetary Problems of the International Economy,Chicago, University of Chicago Press, pp. 61–89.

——(1976), International Monetary Relations: Theory, History and Policy, New York, Harper and Row.


Part I

The gold standard in theory

Introduction

One of the most durable questions in the literature on the gold standard is how the adjustment processmaintained payments balance and currency stability. The most influential answer is the price-specie flowmodel of David Hume, described in Chapter 2. Hume’s model highlights the role of relative prices inbringing imports and exports into equality. The model advanced by P.B.Whale shifts the focus to interestrates and international capital flows, minimizing the importance of gold flows, the factor emphasized byHume. Donald McCloskey and J.Richard Zecher shift the focus again, this time to money supply and moneydemand, criticizing the Humeian view that adjustment relied on relative price movements and insteadstressing changes in wealth and money balances. Much in the way that McCloskey and Zecher minimize therole of relative commodity prices, Trevor Dick and John Floyd minimize interest-rate differentials, arguingthat capital flows responded quickly to international differences in interest rates, which in turn allowed thebalance of payments to accommodate disturbances to the economy. Finally in Part I, Michael D.Bordo andFinn Kydland analyze the gold standard as a contingent rule, highlighting the authorities’ ability to suspendgold convertibility in response to exceptional shocks without damaging their credibility.

2On the balance of trade

David Hume*

It is very usual, in nations ignorant of the nature of commerce, to prohibit the exportation of commodities,and to preserve among themselves whatever they think valuable and useful. They do not consider, that, in thisprohibition, they act directly contrary to their intention; and that the more is exported of any commodity themore will be raised at home of which they themselves will always have the first offer.

It is well known to the learned, that the ancient laws of Athens rendered the exportation of figs criminal;that being supposed a species of fruit so excellent in Attica, that the Athenians deemed it too delicious forthe palate of any foreigner. And in this ridiculous prohibition they were so much in earnest, that informerswere thence called sycophants among them, from two Greek words, which signify figs and discoverer.There are proofs in many old acts of parliament of the same ignorance in the nature of commerce,particularly in the reign of Edward III. And to this day, in France, the exportation of corn is almost alwaysprohibited; in order, as they say, to prevent famines; though it is evident, that nothing contributes more tothe frequent famines, which so much distress that fertile country.

The same jealous fear, with regard to money, has also prevailed among several nations; and it required bothreason and experience to convince any people, that these prohibitions serve no other purpose than to raisethe exchange against them, and produce a still greater exportation.

These errors, one may say, are gross and palpable: But there still pre vails, even in nations wellacquainted with commerce, a strong jealousy with regard to the balance of trade, and a fear that all theirgold and silver may be leaving them. This seems to me, almost in every case, a groundless apprehension;and I should as soon dread, that all our springs and rivers should be exhausted, as that money shouldabandon a kingdom where there are people and industry. Let us carefully preserve these later advantages;and we need never be apprehensive of losing the former.

It is easy to observe, that all calculations concerning the balance of trade are founded on very uncertainfacts and suppositions. The custom-house books are allowed to be an insufficient ground of reasoning; noris the rate of exchange much better; unless we consider it with all nations, and know also the proportions ofthe several sums remitted; which one may safely pronounce impossible. Every man, who has ever reasonedon this subject, has always proved his theory, whatever it was, by facts and calculations, and by anenumeration of all the commodities sent to all foreign kingdoms.

The writings of Mr Gee struck the nation with an universal panic, when they saw it plainly demonstrated,by a detail of particulars, that the balance was against them for so considerable a sum as must leave themwithout a single shilling in five or six years. But luckily, twenty years have since elapsed, with an expensive

* From Essays, Moral, Political and Literary, vol. 1, London, Longmans, Green, 1898, pp. 330–41, 343–5, abridged.This essay was first published in 1752.

foreign war; yet is it commonly supposed, that money is still more plentiful among us than in any formerperiod.

Nothing can be more entertaining on this head than Dr Swift, an author so quick in discerning themistakes and absurdities of others. He says, in his Short View of the State of Ireland, that the whole cash ofthat kingdom formerly amounted but to 500,000l.; that out of this the Irish remitted every year a neatmillion to England, and had scarcely any other source from which they could compensate themselves, andlittle other foreign trade than the importation of French wines, for which they paid ready money. Theconsequence of this situation, which must be owned to be disadvantageous, was, that, in a course of threeyears, the current money of Ireland, from 500,000l. was reduced to less than two. And at present, I suppose,in a course of thirty years it is absolutely nothing. Yet I know not how, that opinion of the advance of richesin Ireland, which gave the Doctor so much indignation, seems still to continue, and gain ground with everybody.

In short, this apprehension of the wrong balance of trade, appears of such a nature, that it discovers itself,wherever one is out of humour with the ministry, or is in low spirits; and as it can never be refuted by aparticular detail of all the exports, which counterbalance the imports, it may here be proper to form a generalargument, that they may prove the impossibility of this event, as long as we preserve our people and ourindustry.

Suppose four-fifths of all the money in Great Britain to be annihilated in one night, and the nationreduced to the same condition, with regard to specie, as in the reigns of the Harrys and Edwards, whatwould be the consequence? Must not the price of all labour and commodities sink in proportion, andeverything be sold as cheap as they were in those ages? What nation could then dispute with us in anyforeign market, or pretend to navigate or to sell manufactures at the same price, which to us would affordsufficient profit? In how little time, therefore, must this bring back the money which we had lost, and raiseus to the level of all the neighbouring nations? Where, after we have arrived, we immediately lose theadvantage of the cheapness of labour and commodities; and the farther flowing in of money is stopped byour fulness and repletion.

Again, suppose, that all the money of Great Britain were multiplied fivefold in a night, must not thecontrary effect follow? Must not all labour and commodities rise to such an exorbitant height, that noneighbouring nations could afford to buy from us; while their commodities, on the other hand, becamecomparatively so cheap, that, in spite of all the laws which could be formed, they would be run in upon us,and our money flow out; till we fall to a level with foreigners, and lose that great superiority of riches,which had laid us under such disadvantages?

Now, it is evident, that the same causes, which would correct these exorbitant inequalities, were they tohappen miraculously, must prevent their happening in the common course of nature, and must for ever, inall neighbouring nations, preserve money nearly proportionable to the art and industry of each nation. Allwater, wherever it communicates, remains always at a level. Ask naturalists the reason; they tell you, that,were it to be raised in any one place, the superior gravity of that part not being balanced, must depress it, tillit meet a counterpoise; and that the same cause, which redresses the inequality when it happens, must forever prevent it, without some violent external operation.1

Can one imagine, that it had ever been possible, by any laws, or even by any art or industry, to have keptall the money in Spain, which the galleons have brought from the Indies? Or that all commodities could besold in France for a tenth of the price which they would yield on the other side of the Pyrenees, without findingtheir way thither, and draining from that immense treasure? What other reason indeed is there, why allnations, at present, gain in their trade with Spain and Portugal; but because it is impossible to heap up

ON THE BALANCE OF TRADE 25

money, more than any fluid, beyond its proper level? The sovereigns of these countries have shown, thatthey wanted not inclination to keep their gold and silver to themselves, had it been in any degree practicable.

But as any body of water may be raised above the level of the surrounding element, if the former has nocommunication with the latter; so in money, if the communication be cut off, by any material or physicalimpediment (for all laws alone are ineffectual), there may, in such a case, be a very great inequality ofmoney. Thus the immense distance of China, together with the monopolies of our India companies,obstructing the communication, preserve in Europe the gold and silver, especially the latter, in much greaterplenty than they are found in that kingdom. But, notwithstanding this great obstruction, the force of thecauses abovementioned is still evident. The skill and ingenuity of Europe in general surpasses perhaps thatof China, with regard to manual arts and manufactures; yet are we never able to trade thither without greatdisadvantage. And were it not for the continual recruits, which we receive from America, money wouldsoon sink in Europe, and rise in China, till it came nearly to a level in both places. Nor can any reasonableman doubt, but that industrious nations, were they as near us as Poland or Barbary, would drain us of theoverplus of our specie, and draw to themselves a larger share of the West Indian treasures. We need nothave recourse to a physical attraction, in order to explain the necessity of this operation. There is a moralattraction, arising from the interests and passions of men, which is full as potent and infallible.

How is the balance kept in the provinces of every kingdom among themselves, but by the force of thisprinciple, which makes it impossible for money to lose its level, and either to rise or sink beyond theproportion of the labour and commodities which are in each province? Did not long experience make peopleeasy on this head, what a fund of gloomy reflections might calculations afford to a melancholyYorkshireman, while he computed and magnified the sums drawn to London by taxes, absentees,commodities, and found on comparison the opposite articles so much inferior? And no doubt, had theHeptarchy subsisted in England, the legislature of each state had been continually alarmed by the fear of awrong balance; and as it is probable that the mutual hatred of these states would have been extremely violenton account of their close neighbourhood, they would have lorded and oppressed all commerce, by a jealousand superfluous caution. Since the union has removed the barriers between Scotland and England, which ofthese nations gains from the other by this free commerce? Or if the former kingdom has received anyincrease of riches, can it reasonably be accounted for by any thing but the increase of its art and industry? Itwas a common apprehension in England, before the union, as we learn from L’abbe du Bos,2 that Scotlandwould soon drain them of their treasure, were an open trade allowed; and on the other side the Tweed acontrary apprehension prevailed: With what justice in both, time has shown.

What happens in small portions of mankind, must take place in greater. The provinces of the Romanempire, no doubt, kept their balance with each other, and with Italy, independent of the legislature; as muchas the several counties of Great Britain, or the several parishes of each county. And any man who travelsover Europe at this day, may see, by the prices of commodities, that money, in spite of the absurd jealousyof princes and states, has brought itself nearly to a level; and that the difference between one kingdom andanother is not greater in this respect, than it is often between different provinces of the same kingdom. Mennaturally flock to capital cities, sea-ports, and navigable rivers. There we find more men, more industry,more commodities, and consequently more money; but still the latter difference holds proportion with theformer, and the level is preserved.3

Our jealousy and our hatred of France are without bounds; and the former sentiment, at least, must beacknowledged reasonable and well-grounded. These passions have occasioned innumerable barriers andobstructions upon commerce, where we are accused of being commonly the aggressors. But what have wegained by the bargain? We lost the French market for our woollen manufactures, and transferred the commerceof wine to Spain and Portugal, where we buy worse liquor at a higher price. There are few Englishmen who


would not think their country absolutely ruined, were French wines sold in England so cheap and in suchabundance as to supplant, in some measure, all ale, and home-brewed liquors: But would we lay asideprejudice, it would not be difficult to prove, that nothing could be more innocent, perhaps advantageous.Each new acre of vineyard planted in France, in order to supply England with wine, would make it requisitefor the French to take the produce of an English acre, sown in wheat or barley, in order to subsistthemselves; and it is evident, that we should thereby get command of the better commodity.

There are many edicts of the French king, prohibiting the planting of new vineyards, and ordering allthose which are lately planted to be grubbed up: So sensible are they, in that country, of the superior value ofcorn, above every other product.

Mareschal Vauban complains often, and with reason, of the absurd duties which load the entry of thosewines of Languedoc, Guienne, and other southern provinces, that are imported into Brittany and Normandy.He entertained no doubt but these latter provinces could preserve their balance, notwithstanding the opencommerce which he recommends. And it is evident, that a few leagues more navigation to England wouldmake no difference; or if it did, that it must operate alike on the commodities of both kingdoms.

There is indeed one expedient by which it is possible to sink, and another by which we may raise moneybeyond its natural level in any kingdom; but these cases, when examined, will be found to resolve into ourgeneral theory, and to bring additional authority to it.

I scarcely know any method of sinking money below its level, but those institutions of banks, funds, andpaper-credit, which are so much practised in this kingdom. These render paper equivalent to money,circulate it throughout the whole state, make it supply the place of gold and silver, raise proportionably theprice of labour and commodities, and by that means either banish a great part of those precious metals, orprevent their farther increase. What can be more short-sighted than our reasonings on this head? We fancy,because an individual would be much richer, were his stock of money doubled, that the same good effectwould follow were the money of every one increased; not considering, that this would raise as much theprice of every commodity, and reduce every man, in time, to the same condition as before. It is only in ourpublic negotiations and transactions with foreigners, that a greater stock of money is advantageous; and asour paper is there absolutely insignificant, we feel, by its means, all the ill effects arising from a greatabundance of money, without reaping any of the advantages.4

Suppose that there are 12 millions of paper, which circulate in the kingdom as money (for we are not toimagine, that all our enormous funds are employed in that shape) and suppose the real cash of the kingdomto be 18 millions: Here is a state which is found by experience to be able to hold a stock of 30 millions. Isay, if it be able to hold it, it must of necessity have acquired it in gold and silver, had we not obstructed theentrance of these metals by this new invention of paper. Whence would it have acquired that sum? From allthe kingdoms of the world. But why? Because, if you remove these 12 millions, money in this state is belowits level, compared with our neighbours; and we must immediately draw from all of them, till we be full andsaturate, so to speak, and can hold no more. By our present politics, we are as careful to stuff the nation withthis fine commodity of bank-bills and chequer-notes, as if we were afraid of being over-burthened with theprecious metals.

It is not to be doubted, but the great plenty of bullion in France is, in a great measure, owing to the wantof paper-credit. The French have no banks: Merchants bills do not there circulate as with us: Usury orlending on interest is not directly permitted; so that many have large sums in their coffers: Great quantitiesof plate are used in private houses; and all the churches are full of it. By this means, provisions and labourstill remain cheaper among them, than in nations that are not half so rich in gold and silver. The advantagesof this situation, in point of trade as well as in great public emergencies, are too evident to be disputed.


The same fashion a few years ago prevailed in Genoa, which still has place in England and Holland, ofusing services of China-ware instead of plate; but the senate, foreseeing the consequence, prohibited the useof that brittle commodity beyond a certain extent; while the use of silverplate was left unlimited. And Isuppose, in their late distress, they felt the good effect of this ordinance. Our tax on plate is, perhaps, in thisview, somewhat impolitic.

Before the introduction of paper-money into our colonies, they had gold and silver sufficient for theircirculation. Since the introduction of that commodity, the least inconveniency that has followed is the totalbanishment of the precious metals. And after the abolition of paper, can it be doubted but money willreturn, while these colonies possess manufactures and commodities, the only thing valuable in commerce,and for whose sake alone all men desire money.

What pity Lycurgus did not think of paper-credit, when he wanted to banish gold and silver from Sparta!It would have served his purpose better than the lumps of iron he made use of as money; and would alsohave prevented more effectually all commerce with strangers, as being of so much less real and intrinsicvalue.

It must, however, be confessed, that, as all these questions of trade and money are extremelycomplicated, there are certain lights, in which this subject may be placed, so as to represent the advantagesof paper-credit and banks to be superior to their disadvantages. That they banish specie and bullion from astate is undoubtedly true; and whoever looks no farther than this circumstance does well to condemn them;but specie and bullion are not of so great consequence as not to admit of a compensation, and even anoverbalance from the increase of industry and of credit, which may be promoted by the right use of paper-money. It is well known of what advantage it is to a merchant to be able to discount his bills upon occasion;and every thing that facilitates this species of traffic is favourable to the general commerce of a state. Butprivate bankers are enabled to give such credit by the credit they receive from the depositing of money intheir shops; and the bank of England in the same manner, from the liberty it has to issue its notes in allpayments. There was an invention of this kind, which was fallen upon some years ago by the banks ofEdinburgh; and which, as it is one of the most ingenious ideas that has been executed in commerce, has alsobeen thought advantageous to Scotland. It is there called a Bank-credit; and is of this nature. A man goes tothe bank and finds surety to the amount, we shall suppose, of a thousand pounds. This money, or any part ofit, he has the liberty of drawing out whenever he pleases, and he pays only the ordinary interest for it, whileit is in his hands. He may, when he pleases, repay any sum so small as twenty pounds, and the interest isdiscounted from the very day of the repayment. The advantages, resulting from this contrivance, aremanifold. As a man may find surety nearly to the amount of his substance, and his bank-credit is equivalentto ready money, a merchant does hereby in a manner coin his houses, his household furniture, the goods inhis warehouse, the foreign debts due to him, his ships at sea; and can, upon occasion, employ them in allpayments, as if they were the current money of the country. If a man borrow a thousand pounds from aprivate hand, besides that it is not always to be found when required, he pays interest for it, whether he beusing it or not: His bank-credit costs him nothing except during the very moment in which it is of service tohim: And this circumstance is of equal advantage as if he had borrowed money at much lower interest.Merchants, likewise from this invention, acquire a great facility in supporting each other’s credit, which is aconsiderable security against bankruptcies. A man, when his own bank-credit is exhausted, goes to any ofhis neighbours who is not in the same condition; and he gets the money, which he replaces at hisconvenience.

After this practice had taken place during some years at Edinburgh, several companies of merchants atGlasgow carried the matter farther. They associated themselves into different banks, and issued notes so lowas ten shillings, which they used in all payments for goods, manufactures, tradesmen’s labour of all kinds;


and these notes, from the established credit of the companies, passed as money in all payments throughoutthe country. By this means, a stock of five thousand pounds was able to perform the same operations as if itwere six or seven; and merchants were thereby enabled to trade to a greater extent, and to require less profitin all their transactions. But whatever other advantages result from these inventions, it must still be allowedthat, besides giving too great facility to credit, which is dangerous, they banish the precious metals: andnothing can be a more evident proof of it, than a comparison of the past and present condition of Scotland inthat particular. It was found, upon the recoinage made after the union, that there was near a million of speciein that country: But notwithstanding the great increase of riches, commerce, and manufactures of all kinds,it is thought, that, even where there is no extraordinary drain made by England, the current specie will notnow amount to a third of that sum.

But as our projects of paper-credit are almost the only expedient, by which we can sink money below itslevel; so, in my opinion, the only expedient by which we can raise money above it, is a practice which weshould all exclaim against as destructive, namely, the gathering of large sums into a public treasure, lockingthem up, and absolutely preventing their circulation. The fluid, not communicating with the neighbouringelement, may, by such an artifice, be raised to what height we please. To prove this, we need only return toour first supposition, of annihilating the half or any part of our cash; where we found, that the immediateconsequence of such an event would be the attraction of an equal sum from all the neighbouring kingdoms.Nor does there seem to be any necessary bounds set, by the nature of things, to this practice of hoarding. Asmall city, like Geneva, continuing this policy for ages, might engross nine tenths of the money of Europe.There seems, indeed, in the nature of man, an invincible obstacle to that immense growth of riches. A weakstate, with an enormous treasure, will soon become a prey to some of its poorer, but more powerfulneighbours. A great state would dissipate its wealth in dangerous and ill-concerted projects; and probablydestroy, with it, what is much more valuable, the industry, morals, and numbers of its people. The fluid inthis case, raised to too great a height, bursts and destroys the vessel that contains it; and mixing itself withthe surrounding element, soon falls to its proper level.[…]

From these principles we may learn what judgment we ought to form of those numberless bars,obstructions, and imposts which all nations of Europe, and none more than England, have put upon trade;from an exorbitant desire of amassing money, which never will heap up beyond its level, while it circulates;or from an ill-grounded apprehension of losing their specie, which never will sink below it. Could any thingscatter our riches, it would be such impolitic contrivances. But this general ill effect, however, results fromthem, that they deprive neighbouring nations of that free communication and exchange which the Author ofthe world has intended, by giving them soils, climates, and geniuses, so different from each other.

Our modern politics embrace the only method of banishing money, the using of paper-credit; they rejectthe only method of amassing it, the practice of hoarding; and they adopt a hundred contrivances, whichserve to no purpose but to check industry, and rob ourselves and our neighbours of the common benefits ofart and nature.

All taxes, however, upon foreign commodities, are not to be regarded as prejudicial or useless, but thoseonly which are founded on the jealousy above-mentioned. A tax on German linen encourages homemanufactures, and thereby multiplies our people and industry. A tax on brandy increases the sale of rum,and supports our southern colonies. And as it is necessary, that imposts should be levied, for the support ofgovernment, it may be thought more convenient to lay them on foreign commodities, which can easily beintercepted at the port, and subjected to the impost. We ought, however, always to remember the maxim ofDr Swift, That, in the arithmetic of the customs, two and two make not four, but often make only one. It canscarcely be doubted, but if the duties on wine were lowered to a third, they would yield much more to thegovernment than at present: Our people might thereby afford to drink commonly a better and more


wholesome liquor; and no prejudice would ensue to the balance of trade, of which we are so jealous. Themanufacture of ale beyond the agriculture is but inconsiderable, and gives employment to few hands. Thetransport of wine and corn would not be much inferior.

But are there not frequent instances, you will say, of states and kingdoms, which were formerly rich andopulent, and are now poor and beggarly? Has not the money left them, with which they formerly abounded?I answer, If they lose their trade, industry, and people, they cannot expect to keep their gold and silver: Forthese precious metals will hold proportion to the former advantages. When Lisbon and Amsterdam got theEast-India trade from Venice and Genoa, they also got the profits and money which arose from it. Wherethe seat of government is transferred, where expensive armies are maintained at a distance, where greatfunds are possessed by foreigners; there naturally follows from these causes a diminution of the specie. Butthese, we may observe, are violent and forcible methods of carrying away money, and are in timecommonly attended with the transport of people and industry. But where these remain, and the drain is notcontinued, the money always finds its way back again, by a hundred canals, of which we have no notion orsuspicion. What immense treasures have been spent, by so many nations, in Flanders, since the revolution,in the course of three long wars! More money perhaps than the half of what is at present in Europe. But whathas now become of it? Is it in the narrow compass of the Austrian provinces? No, surely: It has most of itreturned to the several countries whence it came, and has followed that art and industry, by which at first itwas acquired. For above a thousand years, the money of Europe has been flowing to Rome, by an open andsensible current; but it has been emptied by many secret and insensible canals: And the want of industry andcommerce renders at present the papal dominions the poorest territory in all Italy.

In short, a government has great reason to preserve with care its people and its manufactures. Its money,it may safely trust to the course of human affairs, without fear or jealousy. Or if it ever give attention to thislater circumstance, it ought only to be so far as it affects the former.

Notes

1 There is another cause, though more limited in its operation, which checks the wrong balance of trade, to everyparticular nation to which the kingdom trades. When we import more goods than we export, the exchange turnsagainst us, and this becomes a new encouragement to export; as much as the charge of carriage and insurance ofthe money which becomes due would amount to. For the exchange can never rise but a little higher than thatsum.

2 Les Intérêts d’Angleterre malentendus.3 It must clearly be remarked, that throughout this discourse, wherever I speak of the level of money, I mean always

its proportional level to the commodities, labour, industry, and skill, which is in the several states. And I assert,that where these advantages are double, triple, quadruple, to what they are in the neighbouring states, the moneyinfallibly will also be double, triple, quadruple. The only circumstance that can obstruct the exactness of theseproportions, is the expense of transporting the commodities from one place to another; and this expense issometimes unequal. Thus the corn, cattle, cheese, butter, of Derbyshire, cannot draw the money of London, somuch as the manufactures of London draw the money of Derbyshire. But this objection is only a seeming one:For so far as the transport of commodities is expensive, so far is the communication between the placesobstructed and imperfect.

4 We observed in Essay III [not included here] that money, when increasing, gives encouragement to industry,during the interval between the increase of money and rise of the prices. A good effect of this nature may followtoo from paper-credit; but it is dangerous to precipitate matters, at the risk of losing all by the failing of thatcredit, as must happen upon any violent shock in public affairs.


3The working of the prewar gold standard

P.B.Whale*

In discussions concerning the postwar gold standard and the possibility of restoring an efficientinternational monetary system, it is natural to take the operation of the gold standard in prewar times as akind of standard. It is therefore of more than historical interest to know exactly how the prewar goldstandard did work.

The most generally accepted view on this matter—the classical explanation, as it may be convenientlytermed—may be summarized as follows. A condition of balance of payments disequilibrium led, if morethan temporary, to an international gold movement: the gold movement induced changes in the volume ofmonetary circulation in the countries concerned, a contraction here and an expansion there: the changes incirculation brought about the changes in incomes and prices required to adjust the balance of trade. Inestablishing a connection between the gold movement and the change in the volume of circulation, it isnecessary to take account of the nature of the currency system involved. As credit means of payment, bankliabilities in the form of notes and deposits, have long been the most important forms of money in the moreadvanced countries, the effect of a gold movement there depended on the reaction of the banking system toa change in its reserves; and in so far as a given change in reserves called for a multiple change in the bankliabilities, the process normally involved a change in the terms on which the banks lent, or the discount rate.

We shall not be concerned here to discuss the part played by price and income changes in maintaininginternational equilibrium. Our concern is with the process by which these price and income changes werebrought about. As there appears to be some difference of opinion as to which points in the foregoingexplanation are really important, it may be well to state that in this discussion the two following will beregarded as crucial:

1 The view that general changes in incomes and prices only occur as a result of changes in the volume ofcirculation. This may be termed the Quantity Theory assumption. (It will be noticed that if the QuantityTheory means this, it is not a truism.)

2 The view that in a modern economy these changes in circulation are brought about by changes indiscount or interest rates. This means of course that the maintenance of the gold standard requires from timeto time a disturbance in the processes of saving and investing in particular countries: if true, it provides animportant argument against the gold standard system.1

It may be noticed that I do not regard the part played by actual gold movements as crucial. Already,before the [First World] War, foreign balances were coming to be used as reserves in place of gold, and somewriters make a great deal of this change. It may certainly have had the effect of making adjustments more‘one-sided’, but apart from this, it does not appear to me to necessitate any important modification of the

* Economica, February 1937, pp. 18–32.

theory, at any rate so long as the exchange reserves are treated in the same way as gold reserves. I shall howeverhave something to say about the practice of basing money on foreign balances in one particular connection(see below).

*It is noteworthy that Taussig, although in a way he has been the leading exponent of the classical

explanation, has often warned his readers that the evidence gives the theory very doubtful support. Further,I think it may be said that on the whole the attempts at historical verification undertaken by the youngermembers of the Harvard School have increased rather than diminished the doubts. (This is least true perhapsof Professor Viner’s book, but Viner does not share my views as to which points are ‘crucial’.)

In fairness, certain circumstances may be mentioned which would tend in any case to make verificationdifficult.

1 The distribution of newly mined gold obviated to a certain extent the necessity for gold movementsbetween non-producing countries. A country could suffer a relative loss of gold through failing to acquireits normal share in the new supplies.

2 Similarly, in the adjustment of price relations between two countries, it would not be necessary for theirdiscount rates actually to move in opposite directions, if there were at the same time a general upward ordownward trend in rates.

But when account has been taken of these, certain difficulties remain.1 The point which has troubled Taussig most is that in many cases the adjustments appeared to have occurred

more immediately and with less friction than his theory would lead one to expect.2 The dependence of international price adjustments on national discount policies raises the question

whether discrepancies in interest rates might not lead to inconvenient capital movements. This has oftenbeen discussed as a theoretical issue. The evidence collected by Professors Andrew and Beach, to whichfurther reference will be made presently, appears to show that in fact gold movements to and from both theUnited States and Britain before the War are more easily connected with relations of interest rates than withrelations of price levels.

3 The prewar policy of certain leading central banks hardly seems to have been comformable to therequirements of a gold standard operating in the manner envisaged by the classical explanation. It isnotorious that both the Bank of France and the National Bank of Belgium consistently pursued the policy ofkeeping their discount rates low and steady.2 It may be said that the Bank of France was able to avoidvarying its discount rates because the country was only on a limping standard. But this point cannot besubstantiated unless it can be shown that the exchange value of the franc fell below the gold export point forappreciable periods. Other special circumstances might be adduced in the case of France (e.g. the limitedimportance of credit money, in later years the size of the metallic reserve). But it is doubtful if they wouldapply to Belgium also, and the fact that both these countries were able to maintain their exchange paritiesthrough changing circumstances, while their discount rates only followed the movements of rates elsewherein a diminished degree, is certainly remarkable. The case of the Bank of England is different, for thatinstitution certainly did have to take active measures, including the variation of its discount rate, for theprotection of its gold reserve. But the object of discount policy may be to control capital movements, andthe recent studies of Mr Sayers and Professor Beach show that the Bank often preferred other devices to theuse of its discount rate in order to avoid disturbing internal credit conditions. In fact the Bank seems to havebeen reluctant to raise the discount rate unless internal trade could stand it or actually needed a curb (i.e.from the point of view of internal stability). In short, then, none of these central banks really observed ‘therules of the game’; and yet the system worked.

32 THE PREWAR GOLD STANDARD

4 It is sometimes asked why a different process of adjustment should be necessary in the relationsbetween different countries from that which may be supposed to operate in the relations between differentparts of the same country. The fairly obvious answer is that within any country, at any rate if it has a unitarybanking system, the common money in the hands of the public can be shifted from one part to another,whereas between countries direct monetary transfers mean movements of the reserves on which a multiplesuperstructure of credit is based. But on this view Scotland with its separate banking system should berelated to England as a foreign country rather than as a different part of Britain. Admittedly the reserves ofthe Scottish banks take the form of London balances rather than gold holdings. But is there evidence thatprice equilibrium between England and Scotland is maintained by the Scottish banks varying their interestrates in response to changes in their reserves?

These are my main grounds for doubts, but two other cases may be mentioned which, although they donot belong to the prewar working of the gold standard, none the less serve to test the theory underlying theclassical view on that matter.

5 I am told, by Mr Ashton, that serious difficulties were often experienced in Lancashire in the lateeighteenth century owing to the shortage of local means of payment, and that various expedients had to beresorted to in consequence. Now according to orthodox theory the local price level should have adjusteditself to the actual supply of means of payment; and if this meant that the price level was unduly low inrelation to other parts of the country, the favourable balance of trade should have brought in supplies ofmoney to raise the price level. But this did not happen apparently; the local price level was determined moredirectly by the national system of prices. Might not the national price level be similarly determined by theworld system of prices?

6 The fall of prices in France in 1920, when the franc was still inconvertible, presents an interesting case.In the spring of that year certain measures of a deflationary tendency were adopted. The discount rate of theBank of France was raised; taxation was increased to improve the budgetary position; and provision wasmade for annual reductions of the indebtedness of the government to the central bank. I do not think it canbe denied that there was a connection between these measures and the ensuing fall in prices. It is possible toargue that the fall was simply the result of internal deflation. On the other hand, having regard to themagnitude of the effect and the comparative inefficacy of similar deflationary measures at other times, itmay be more plausible to connect the fall in French prices with the fall in world gold prices which tookplace at the same time. On this view the significance of the French measures was that by restoringconfidence in the currency for a time they prevented the exchange from depreciating; with the exchangescomparatively steady, the French price system became subject to world influences. This at any rate is theinterpretation favoured by most French economists.

*The crucial points in the classical explanation were challenged long ago by the writers of the Banking

School. But the members of this school were unable to offer a satisfactory alternative explanation; and it ismainly the absence of this, I think, which accounts for the continued predominance of the classical view.The purpose of what follows is to provide an alternative hypothesis.

We may begin with a line of thought first developed, I believe, by Mr Harrod in his InternationalEconomics.3 Mr Harrod points out that a decline in the effective demand for a country’s exportsautomatically curtails the incomes of the exporters, and through these, the incomes of those who sell to theexporters, and so on. Further it is shown that, in the absence of a counteracting influence, the contraction ofincomes must proceed just far enough to restore the balance of trade; for the contracting tendency is onlyexhausted by people being induced to economize on foreign trade goods (imports or home-produced goodswhich can find an export market). A counteracting influence is present, however, in the fact that individuals

THE GOLD STANDARD IN THEORY AND HISTORY 33

will reduce their money balances, either because they wish for a time to maintain their expenditure despitethe reduction of their incomes, or because with reduced incomes they no longer consider it necessary tokeep so large a balance. This ‘setting free’ of money tends to restore incomes; and incomes not beingcontracted sufficiently, there is a deficit in the balance of trade and gold flows out. A recent German writer,Dr Maier, treats gold flows somewhat similarly as a result of an excess of money balances at the equilibriumlevel of incomes.4

Whether one regards gold flows as the result of ‘setting free’ balances rather than as the directconsequence of the falling off of exports, also whether or not one agrees with Mr Harrod that the firstprocess of income contraction is instantaneous, both seem to depend on a matter of conven tion. If oneadopts the convention that incomes are earned and spent in the same period, these views appear to becorrect; if one adopts the convention that incomes are earned in one period (Mr Robertson’s ‘day’) and spentin the next, they are not. But there is no need to pursue this point. In relation to our practical issues, thisanalysis seems at first sight to make very little difference. So far as the final adjustment of incomes isbrought about by gold flows (or for that matter by changes in exchange reserves), the process may be thatoutlined in the classical theory; gold flows affect reserves and the credit system has to adapt itself tochanges in reserves by altering the terms of lending.

But there is another possibility which gives this analysis great practical significance. This is that as eachindustry, from the export industry onwards, experiences a decline in the value of its output, there will be areduction of bank loans and a cancellation of credit money. The balances which would otherwise bereleased to restore incomes would thus be extinguished and the ‘counteracting influence’ noticed aboveremoved. Gold movements and changes in reserves will occur, but they may on this view be accompanimentsand not causes of the change in the volume of credit. The argument does not of course show that thechanges in the volume of credit and in reserves will naturally tend to be proportional to one another. Thismay not matter if there is some flexibility in reserve ratios; or the reserves may be adjusted to the creditvolume, as suggested in the next section. In any case the extent to which the reserves control the volume ofcredit is on this view greatly reduced.

The significant point of difference between this and the classical view is that according to this the changein the volume of credit comes about, as it were, spontaneously, without being induced by a change ininterest rates. Some might seek to minimize this point by arguing that in these circumstances the ‘naturalrate’ of interest falls, so that the maintenance of the same ‘money’ rate is equivalent to a rise. Certainly, to usethe old terminology, investment falls short of savings. But I think there is a difference between a change inthe demand for loans due to a change in the aggregate demand for products, and one due to a change in therelationship between the market rate of interest and the marginal productivity of capital schedule. Take thecase of an increase in the marginal productivity of capital, interest rates remaining the same (or a fall ininterest rates, marginal productivity remaining the same); and assume, as is most appropriate in applying thenatural rate of interest concept, that resources are fully employed. The outcome will be an increase ofborrowing for the purpose of making production more capitalistic, and, assuming no change in voluntarysaving, an immediate distortion, in the Austrian terminology, of the structure of production. On the otherhand, if borrowing is increased as a result of an increase in aggregate effective demand from outside, it isthe direction of the new demands which will determine the re-allocation of resources, and the bidding up ofthe prices of factors and materials by the borrowers will effect primarily a horizontal change in industry. Itis true that this is not the whole story. Since factories and instruments of production are often notconvertible from one purpose to another, the expansion of any industry, even if accompanied by the declineof another, is likely to call for new investment, to raise, in short, the schedule of marginal productivity ofcapital. Hence if interest rates remain the same, there will be some distortion of the structure of production,


not so much because the capital developments in connection with the expanding industry are inappropriateas because there will be insufficient curtailment of capital development in other directions. When thisconcession is made, however, I think there remains a substantial difference between the two types of case.And apparently it is not necessary to make a corresponding concession in the case of a decline in aggregatedemand. It is true that in this case the production of capital goods is likely to fall off more than that ofconsumption goods during the period in which total output is declining, because all changes in total outputtend to be reflected disproportionately in the production of capital goods. But this is analytically distinctfrom the kind of distortion envisaged in the Austrian theory. Once costs have been adjusted and fullemployment regained, the change in the direction of employment which is likely to be involved will have thesame effect as in the opposite case (an increase in foreign demand) in raising the demand for capital. Henceit cannot be said that in this case the failure to lower interest rates introduces ‘structural distortion’.

In any case, leaving aside these abstruse questions of capital theory, adjustments in the volume of creditthrough changes in demand at the same rate of interest are likely to be more immediate than those inducedby changes in the rate, and the complications resulting from discrepancies of interest rates betweencountries are avoided.

*Our theory must also take account of the possibility that gold movements, instead of being the

determinants of the supply of money, may themselves be determined by monetary requirements. Accordingto the strict quantity theory the demand for money is always adjusted to the supply through changes in itsvalue, and monetary requirements must be the same as the actual supply. I do not need to discuss whether thisis true for an isolated community; the suggestion is that in a regime of fixed exchange rates the monetaryrequirements of a particular country may be altered by changes in prices or trade activity independently ofany prior change in the supply of money. If then the national supply of money is inelastic, owing, forexample, to reserve requirements, an increase in monetary requirements will tend to raise interest rates, bycurtailing the offer of funds for investment or increasing the demand for loans if the need is for larger bankbalances, by curtailing bank reserves if the need is for cash in the narrower sense. Higher interest rates maypromote a gold inflow through a movement of short-term capital. (The same relationship between price levels,interest rates and monetary requirements can be seen, I think, in seasonal movements. Seasonal variations inmonetary requirements appear to affect interest rates more than price levels.)

Evidence of this kind of relationship between gold movements and the monetary system is provided bythe work of Professor A. Piatt Andrew and Professor Beach on the cyclical gold movements of the UnitedStates and the United Kingdom respectively. In both cases it has been found that gold flowed in duringperiods of active trade and rising prices, and flowed out in the opposite phase of the cycle. These facts areexplicable on the orthodox theory only on the assumption that the countries concerned lagged behind therest of the world in cyclical movements. It is highly improbable that this was the case with both the UnitedStates and the United Kingdom. The more reasonable interpretation is that these gold movements dependedin the first instance on the relation of interest rates. Further evidence of concomitant movements of gold intoand out of internal circulation confirms the view that it was the monetary requirements determined by agiven price level which provided the underlying cause of the international gold movements. Presumably thecountries which acted as the counter-parties to Britain and America in these movements were in a positionto allow a considerable fluctuation in their gold reserves.

In the foregoing argument it is not intended to deny that the rate of interest exercised a controllinginfluence over the movement of prices and the general course of the trade cycle. High interest rates weredoubtless of crucial importance in bringing booms to an end. What is contended is that in these particularcases the raising of interest rates did not have the effect of producing a relative reduction of prices in certain


countries. High rates in London led rather to a world fall in prices, partly because of the sympatheticmovement of rates elsewhere, partly because of the effect on British entrepôt trade and British long-termforeign investment.

The restriction of the argument to certain cases is also important. Professor Beach’s study of British goldmovements relates to the period from 1880 to 1914. It is possible that he would have obtained differentresults if he had taken an earlier period. The data have not, I think, been systematically studied, but one hasthe impression that in several of the earlier nineteenth-century booms the expansion of credit in Englandoutpaced any similar movement in other countries, with the result that there occurred an adverse balance ofpayments and a foreign drain. (I gather from his review in the Statistical Journal that Mr Hawtrey considersthat this may sometimes have happened in the course of the period studied by Professor Beach.) This is allin accordance with the classical line of thought, but it is also reconcilable with our alternative hypothesis.According to the argument of the preceding section, a change in the direction of demand which tends to turnthe balance of trade against a country exercises at the same time a corrective deflationary influence. In thecase of an inappropriate increase in home demand due to internal credit expansion, the effect on the balanceof trade—the ‘leakage’ of demand abroad—should curtail the tendency for demand to expand cumulatively.But if the original impulse to credit expansion continues to operate, the trade balance may remainunfavourable; and gold may continue to flow out until the loss of reserves compels action to be taken. If themarkets of the country are very closely linked with foreign markets, the loss of reserves by the ordinarybanks should lead to an almost immediate correction of the expansive impulse, except in so far as asympathetic movement is generated in the other countries; for example there would appear to be littledanger of a persistent tendency for credit to expand faster in Canada than in the United States. On the otherhand, if the independent home market is relatively large, the original expansion may go on for some timeand give rise to secondary expansions, before the full effect on the balance of payments is felt. In this casethen there is need for anticipatory action and central bank intervention may be called for.

Our argument does not dispose, therefore, of the need for a national regulation of credit in order toprevent or correct an undue expansion in comparison with other countries. This regulation involves, directlyor indirectly, the adjustment of interest rates; but in this case it can be said, I think, that the money rates aremerely being kept in line with the natural rate,5 and any movement of capital which results can be ascribedto an increase in the demand for capital.

*There remains to be considered the process of adjustment in connection with capital movements,

assuming that the importing country requires additional real capital and not merely additional supplies ofmeans of payment, as in the case considered at the beginning of the previous section.6

A transfer of capital involves in itself a redistribution of spending power, and this is likely to beaccompanied by a change in the direction of demand. This is the basis for the view that the ‘transfer’process involves a change in the terms of trade. So far as this happens, there will have to be a change in therelationship between the price and income structures of the countries concerned; the factors of production willhave become more valuable in some countries, less valuable in others. But this change will be of the samekind as that required by any other change in the direction of demand, and all that has been said above aboutthe process of adjustment applies equally here.

The redistribution of spending power itself, however, apart from any change in the direction of demandand the terms of trade, may require a redistribution of money or means of payment. This is a point with respectto which there are certain differences of opinion. Some writers (Iversen, White, for example) hold that theremust first be a redistribution of money, although it may be only a temporary one, in order that theredistribution of spending power shall take place at all. Others (including the present writer) think that the


making of a foreign loan or investment itself shifts spending power pro tanto, but that this shift requires tobe accompanied by a monetary redistribution if the process of transfer is to be smooth.7 Nurkse wouldapparently reject both these views and see no necessity for monetary changes apart from changes in theterms of trade.

Assuming that the primary shift in spending power does call for a monetary transfer, the question arisesas to how this may be brought about. Most obviously it may be affected by a movement of gold; and if goldreserves are ample in the lending country, the process may be a simple and natural one. The lendingcommunity makes its surrender of wealth to the borrowers partly in the form of money, transferring this bymeans of a movement of gold. But if the gold movement necessitates a multiple contraction of credit inorder to preserve the minimum reserve ratio in the lending country (accompanied perhaps by a multipleexpansion of credit in the borrowing country), several complications arise. The change in money supplieswill be excessive: interest rates will be distorted to bring about the credit changes: and the changes ininterest rates will impede the international movement of capital. In some cases, of course, interest rates maybe raised in the lending country with the deliberate object of checking the movement of capital and goldabroad. It is when interest rates are raised in these ways and not as simple consequence of the transference ofreal capital, that it may be said that the demand for capital abroad has the effect of raising rates above the naturallevel.

There is, however, another way in which these monetary transfers may be brought about: that is, bycreating additional money in the borrowing country on the basis of money held in the lending country.There is evidence, I think, that this was commonly done by the banks in many of the countries in whichBritain invested—the Dominions and British colonies, the South American countries and Japan. Thelending country is in this way spared the loss of gold, yet the borrowing country gets increased supplies ofmoney. If the banks in the borrowing country treat their foreign (or overseas) balances as being exactlyequivalent to gold reserves, and proceed to make them the basis for a multiple expansion of their notes ordeposits, the complications mentioned above will not, it is true, be entirely avoided. But they may be contentto allow their foreign balances to increase relatively to their deposits; and I believe that the evidence againshows that the banks in the countries mentioned did commonly act in this way, restrained presumably bysome instinctive recognition that any greater expansion of credit would soon have to be reversed.

The character of the process is also affected by whether the banks of the borrowing country continue tohold their external funds in liquid form or convert them into interest-earning assets held in the lendingcentre. The ideal arrangement would appear to be for these banks to create local deposits of an amountexactly equal to their increased external balances and to hold these balances completely idle. This wouldgive the closest possible approximation to the interregional shifts of circulation which take place within anyone country. But the possibility that the external balances, instead of being ‘sterilized’, will be invested,must be reckoned with. What this means is that spending power will not be directly reduced in the lendingcountry by the full amount of the foreign (overseas) loans; that part of the surrender of wealth which wouldotherwise have taken the form of a surrender of money will not take place at all. If we can assume that allother countries (not merely the borrowing country) are prepared to base their money on that of the lendingcountry, this failure to contract spending power in the latter can be compensated by a greater expansion ofcredit in the borrowing country, leading to a certain expansion in ‘neutral’ countries also. That is to say,relative values will be the same as in the ‘ideal’ process but money prices will be on a higher level. In thisinflationary form the process of adaptation will probably be easier at the time of an increase of foreignlending, but correspondingly greater difficulties will be experienced in the debtor countries when the flowof one-sided payments has to be reversed, that is, when debt repayments and interest charges come toexceed new loans. So far as the condition mentioned above is not fulfilled and there are neutral or third-


party countries (possibly also some borrowers) which are not willing to base their money on that of thelending country, it seems inevitable that the latter should lose gold.

*It must be emphasized that the suggestions made in the preceding discussion—particularly the suggestion

of a more direct adaptation of the volume of credit to what is required by international price relations than iscontemplated in the classical theory—are no more than tentative. They provide a hypothetical view of theworking of the gold standard which is more in accordance with my own impressions with regard to thehistorical facts than the accepted view; but it is admitted that my view requires a more exhaustive testing,with regard both to the cases in which the gold standard has worked successfully and to those in which ithas broken down, before it can claim acceptance.

In the meantime it may be useful to indicate certain consequences which would follow if my view werecorrect.

1 Since gold movements (or, more generally, changes in reserves) and discount-rate adjustments aredisplaced from their central position in the process of international price adjustment, the question of ‘observingthe rules of the game’, as this is ordinarily understood, loses much of its importance. Indeed in some cases itwill appear desirable that gold movements (changes in reserves) should not be accompanied by a multiplechange in the volume of credit money;8 and if the ordinary banks maintain a constant ratio between theirliabilities and their balances at the central bank, this may require some offsetting of gold movements by thelatter. With respect to discount-rate variations, it is possible that the abnormal size of international short-term capital movements in recent years has been due in part to the excessive use of this method ofregulating the international position.

2 Whilst central bank policy is still important in certain respects, a new importance is given to the policyof the ordinary banks. If it is essential that the volume of credit should speedily adapt itself to changes intrade conditions, a new justification is found for the otherwise rather discredited theory that notes and depositsshould be covered by ‘self-liquidating’ loans. (Perhaps the instincts of the Banking School were right onthis point, too.) But the restriction of loans to the provision of working capital is not enough. For, as thereader has probably been objecting, it is quite possible that manufacturers faced with a declining demandwill seek to increase their working capital in the form of stocks of unsold goods, at least for a time; and sofar as the banks allow them to borrow for this purpose, the volume of credit will change in the wrongdirection. (It is similarly possible that an improvement in trade will lead at first to a repayment of bankloans, but this is not very likely unless stocks have been abnormally large.) The behaviour of the banks inface of changing trade prospects is therefore of critical importance in my view, and what matters is not so muchthe rates of interest which they charge as their willingness to lend at all in certain cases. It would seem moreprobable that the banks would behave in the right way (from this point of view) if they are faced withgradual changes than if they are confronted suddenly with a situation seriously out of adjustment, e.g. bystabilization of the currency at too high a value.

3 It has not been found necessary to refer in the preceding discussion to the question of income and costrigidities. This only becomes relevant if we are considering not merely the rectification of the balance ofpayments but also the maintenance of full employment. One of the factors which is often supposed to beresponsible for wage rigidity—the payment of unemployment allowances from public funds—may, however,have a bearing on the former issue. So far as the maintenance of the expenditure of the workers in depressedindustries is not mere transfer of expenditure, it must tend both to affect the volume of imports directly andto impair the cumulative process of income contraction.


Notes

1 cf. C.H.Walker, ‘The working of the pre-war gold standard’, Review of Economic Studies, vol. I, no. 3.2 Except in the case of the former for a period in the 1850s and 1860s.3 Roy Harrod, International Economics, London, Nisbet, 1933. Mr Harrod has developed his ideas further in The

Trade Cycle, Oxford, Clarendon Press, 1936. The present article was completed before I had opportunity to studyhis later work—hence the absence of any reference to it in the text.

4 Karl Freidrich Maier, Goldwanderungen, Jena, G.Fischer, 1935.5 I admit that in the case in which there is general unemployment of productive factors, this does not mean any

more than the rate which equalizes saving and investment at the existing level of incomes. But at least themaintenance of this rate is different from fixing a rate at which there is (or would be) an excess of saving with theexisting level of incomes.

6 The transition which has already been made to the present tense will warn the reader of the hypothetical characterof the argument.

7 See my article in Economica of February 1936.8 This apart from the obvious case of temporary movements of gold.


4How the gold standard worked, 1880–1913*

Donald N.McCloskey and J.Richard Zecher†

The monetary theory and its implications for the gold standard

Each intellectual generation since the mercantilists has revised or refined the understanding of how thebalance of payments is kept in equilibrium under a system of fixed exchange rates, and all theseunderstandings find a place in the historical literature on the gold standard of the late nineteenth century. Itis difficult, therefore, to locate the orthodox view on how the gold standard worked, for it is many views. Ifone can find historical and economic writings describing the gold standard (and other systems of fixedexchange rates) in the manner of Hume, as a price-specie-flow mechanism, involving changes in the levelof prices, one can also find writings describing it in the manner of Marshall, involving changes in theinterest rate, or of Taussig, involving changes in the relative price of exportables and importables, or ofOhlin, involving changes in income. The theoretical jumble is made still more confusing by a number of factualanomalies uncovered lately.1 Among other difficulties with the orthodox views, it has been found that thegold standard, even in its heyday, was a standard involving the major currencies as well as gold itself, andthat few, if any, central banks followed the putative ‘rules of the game’.

This essay reinterprets the gold standard by applying the monetary theory of the balance of payments tothe experience of the two most important countries on it, America and Britain. Before explaining, testingand using the theory in detail, it will be useful to indicate a few of the ways in which accepting it willchange the interpretation of the gold standard of the late nineteenth century. The most direct implication isthat central bankers did not have control over the variables over which they and their historians havebelieved they had control. The theory assumes that interest rates and prices are determined on worldmarkets, and therefore that the central bank of a small country has little influence over them and the centralbank of a large country has influence over them only by way of its influence over the world as a whole.

A case in point is the Bank of England. It is often asserted, as Keynes put it, that

* An earlier and longer version of this essay (available on request) was presented to the Workshop in Economic Historyat the University of Chicago and to the Cliometrics Conference at the University of Wisconsin. We wish to thank theparticipants in these meetings for their comments. The friendly scepticism of Moses Abramovitz, C.K.Harley, HughRockoff, Jeffrey Willliamson and our colleagues at the University of Chicago, among them Stanley Fischer, Robert J.Gordon, A.C.Harberger, Harry G.Johnson, Arthur Laffer and H.Gregg Lewis, contributed to a sharpening of theargument.

†From Jacob A.Frenkel and Harry G.Johnson (eds.), The Monetary Approach to the Balance of Payments, London,Allen & Unwin, 1976, pp. 184–6, 192–208, abridged.

During the latter half of the nineteenth century the influence of London on credit conditionsthroughout the world was so predominant that the Bank of England could almost have claimed to bethe conductor of the international orchestra. By modifying the terms on which she was prepared tolend, aided by her own readiness to vary the volume of her gold reserves and the unreadiness of othercentral banks to vary the volumes of theirs, she could to a large extent determine the credit conditionsprevailing elsewhere.2

When this musical metaphor is examined in the light of the monetary theory it loses much of its charm. If itis supposed, as in the monetary theory, that the world’s economy was unified by arbitrage, and if it issupposed further that the level of prices in the world market was determined, other things equal, by theamount of money existing in the world, it follows that the Bank’s potential influence on prices (and perhapsthrough prices on interest rates) depended simply on its power to accumulate or disburse gold and otherreserves available to support the world’s supply of money. By raising the interest rate (the Bank Rate) at whichit would lend to brokers of commercial bills, the Bank could induce the brokers, or whoever else in theBritish capital market was caught short of funds, to seek loans abroad, bringing gold into the country andeventually into the vaults of the Bank. If it merely issued bank notes to pay for the gold the reservesavailable to support the supply of money would be unchanged, for Bank of England notes were used both athome and abroad as reserves. Only by decreasing the securities and increasing the gold it held—anautomatic result when it discouraged brokers from selling more bills to the Bank and allowed the bills italready held to come to maturity—could the Bank exert a net effect on the world’s reserves. In other words,a rise in the Bank Rate was effective only to the extent that it was accompanied by an open marketoperation, that is, by a shift in the assets of the Bank of England out of securities and into gold. Theamounts of these two assets held by the Bank, then, provide extreme limits on the influence of the Bank onthe world’s money supply. Had the Bank in 1913 sold off all the securities held in its banking department itwould have decreased world reserves by only 0.6 per cent; had it sold off all the gold in its issuedepartment, it would have increased world reserves by only 0.5 per cent.3 Apparently the Bank was no morethan the second violinist, not to say the triangle player, in the world’s orchestra. The result hinges on theassumption of the monetary theory that the world’s economy was unified, much as each nation’s economyis assumed to be in any theory of the gold standard. If the assumption is correct the historical inference isthat the Bank of England had no more independent influence over the prices and interest rates it faced than,say, the First National Bank of Chicago has over the prices and interest rates it faces, and for the samereason.

A related inference from the monetary theory is that the United Kingdom, the United States, and othercountries on the gold standard had little influence over their money supplies. Since money, like othercommodities, could be imported and exported, the supply of money in a country could adjust to its demandand the demand would depend on the country’s income and on prices and interest rates determined in theworld market. The creation of money in a little country would have little influence on these determinants ofdemand and in consequence little influence over the amount actually supplied. How ‘little’ America andBritain were depends on how large they were relative to the world market, and in a world of fullemployment and well-functioning markets the relevant magnitude is simply the share of the nation’s supplyof money in the world’s supply. One must depend on an assumption that the money owned by citizens of acountry was in rough proportion to its income, for the historical study of the world’s money supply is still inits infancy.4 In 1913 America and Britain together earned about 40 per cent of the world’s income, Americaalone 27 per cent.5 A rise in the American money supply of 10 per cent, then, would raise the world’smoney supply on the order of 2.7 per cent; the comparable British figure is half the American. Clearly, in

HOW THE GOLD STANDARD WORKED, 1880–1913 41

the jargon of international economics, America and Britain were not literally ‘small countries’. Yet 2.7 percent is far from the 10 per cent implied by the usual model, that of a closed monetary system, and theBritish figure is far enough from it to make it unnecessary for most purposes in dealing with the Britishexperience to look closely into the worldwide impact of British policy.

Finally, the monetary theory implies that it matters little whether or not central banks under the goldstandard played conscientiously the ‘rules of the game’, that is, the rule that a deficit in the balance ofpayments should be accompanied by domestic policies to deflate the economy. The theory argues thatneither gold flows nor domestic deflation have effects on prevailing prices, interest rates and incomes. Theinconsequentiality of the rules of the game may perhaps explain why they were ignored by most centralbankers in the period of the gold standard, in deed if not in words, with no dire effects on the stability of thesystem.[…]

Did international markets work well?

If arbitrage—or, more precisely, a close correlation among national price levels brought about by theordinary working of markets—can be shown to characterize the international economy of the late nineteenthcentury, many of the conclusions of the monetary theory will follow directly and the rest will gain inplausibility. In the monetary theory, the international market short-circuits the effects of domestic policy onAmerican prices, and the expansion of the domestic supply of money spills directly into a deficit in thebalance of payments.

It is essential, therefore, to examine the evidence for this short-circuiting. As a criterion of itseffectiveness, we use the size of the contemporaneous correlations among changes in the prices of the samecommodities in different countries. We have chosen a sample of the voluminous information on prices forexamination here.6 The statistical power of the tests is not as high as one might wish, for even if two nationsshared no markets they could none the less exhibit common movements in prices if they shared similarexperiences of climate, technological change, income growth or any of the other determinants of prices. Inthe long run, indeed, the other theories of the balance of payments imply some degree of correlation amongnational prices. For this reason we have resisted the temptation to improve the correlations by elaborateexperimentation with lags and have concentrated on contemporaneous correlations, that is, on correlationsamong prices in the same year. If international markets worked as sluggishly as the other theories assume,there would be little reason to expect contemporaneous correlations to be high.

The simplest way to think about arbitrage is in terms of a single market. Given fixed exchange rates andthe vigorous pursuit of profit through arbitrage, the correlation between price changes for a homogeneouscommodity in two countries, say America and Britain, separated by transportation costs and tariffs, wouldbe zero within the limits of the export and import points and unity at those points. A regression of British onAmerican prices would test simultaneously for the lowness of the commodity’s cost of transportation,including tariffs, relative to its price and the vigour with which prices were arbitraged. The good would notactually have to be traded between the two countries for the correlation to be high: the mere threat ofarbitrage, or a common source of supply or demand, would be sufficient for goods with low transport costs.For goods actually flowing in trade in a uniform direction over the period 1880 to 1913, such as wheat fromAmerica to Britain, one would expect the correlation to be perfect and the slope of the correspondingregression to be unity, no matter what the cost of transport or the level of tariffs, so long as these did notchange. They both did change, of course, as exemplified by the failure of the German price of wheat to fallas far as the British or American during the 1880s, as the Germans imposed protective duties on wheatimports.7 None the less, the average correlation among the changes in American, British and German prices


of wheat is high, about 0.78. A regression of the annual change in British prices on the change in Americanprices (Britain had no tariffs on wheat, but the cost of ocean transport was falling sharply in the period)yields the following result (all the variables here and elsewhere in this section are measured as annualabsolute changes; the figures below the coefficients in parentheses are standard errors; the levels of thevariables have been converted to an index in which the average levels are equal to one).8

BWT=0.0076+0.646 AWT R2=0.58(0.0012) (0.102) D.−W.=2.02

One would expect errors in the independent variable to affect this and the later regressions, biasing the slopetowards zero (there were changes in the source of the American wheat price, for example, and after 1890 itis a New York price alone). The value of 0.646 would be a lower bound on the true slope and the valueimplied by a regression of the American on the British price (1.124) an upper bound. The two boundsbracket reasonably closely the value to be expected theoretically, namely, 1.0, and the constants in bothregressions (which represent the trend in the dependent price over time) are insignificantly different from zero.Not surprisingly, in short, wheat appears to have had a unified world market in the late nineteenth century; afortiori, so did gold, silver, copper, diamonds, racehorses and fine art.

This conclusion can be reinforced from another direction. For wheat the reinforcement is unnecessary,for few would doubt the international character of the wheat market, but it is useful to develop here the lineof argument. Because of transport costs, information costs and other impediments to a perfect correlationamong changes in national prices, any use of the notion of a perfectly unified market must be anapproximation, within one country as well as between two countries. For purposes of explaining the balanceof payments economists have been willing to accept the approximation that within each country there is oneprice for each product, setting aside as a second-order matter the indisputable lack of perfect correlationbetween price changes in California and Massachusetts or between price changes in Cornwall andMidlothian. It is reasonable, therefore, to use the level of the contemporaneous correlation between theprices of a good in different regions within a country as a standard against which to judge the unity of themarket for that good between different countries. If the correlations between the prices of wheat inAmerica, Britain and Germany were no lower than those between the prices of wheat in, say, different partsof Germany, there would be no grounds for distinguishing between the degree of unity in the nationalGerman market and in the international market for wheat. This was in fact the case. The average correlationbetween changes in the prices of wheat in pairs of German cities (Berlin, Breslau, Frankfurt, Königsberg,Leipzig, Lindau and Mannheim) from 1881 to 1912 was 0.85, quite close to the average correlation for thethree countries over the same period of 0.78.

One could proceed in this fashion through all individual prices, but a shorter route to the same objectiveis to examine correlations across countries between pairs of aggregate price indexes. Contrary to theintuition embodied in this thought, however, there is no guarantee, at any rate none that we have been ableto discover, that the correlation of the indexes is an unbiased estimator of the average degree of correlationamong the individual prices or, for that matter, that it is biased in any particular direction.9 In other words,barriers to trade could be high or low in each individual market without the aggregate correlationnecessarily registering these truths. None the less, putting these doubts to one side, we will trust henceforthto the intuition.

The pioneers of the method of index numbers, Laspeyres, Jevons and others writing in the middle of thenineteenth century, produced indexes of wholesale prices—believable indexes of retail prices began to beproduced only in the 1890s and implicit GNP deflators, of course, much later—and in consequence wholesale


price indexes dominated empirical work on the balance of payments in the formative years of the theory.The contempora neous correlation between annual changes in British and American wholesale prices 1880–1913 is 0.66, high enough in view of the differences in weights in the indexes and in view of the lowcorrelation of annual changes implied by the lags operating in the orthodox theories to lend support to thepostulate of a unified world market.

It is at this point, however, that supporters of the orthodox theory begin to quarrel with the argument, as didTaussig with those bold enough to suggest that world markets in more than merely traded goods wereintegrated in the late nineteenth century, or as did the many doubters of the theory of purchasing powerparity with those who used wholesale prices to indicate the appropriate rates of exchange after World War I.The standard objection has been that wholesale price indexes are biased samples from the distribution ofcorrelations because they consist largely of easily traded goods, ignoring nontraded services andunderrepresenting nontraded goods. A large lower tail of the distribution, it is said, is left off, leading to a falseimpression that national price levels are closely correlated.

A point that must be made at once, however, is that traded goods, in the sense of goods actually tradedand goods identical to those actually traded, were not a small proportion of national income. Historians andeconomists have usually thought of the openness of economies in terms of the ratio of actual exports orimports to national income, and have inferred that the United States, with a ratio of exports to nationalincome of about 0.07 in the late nineteenth century, was relatively isolated from the influence ofinternational prices and that the United Kingdom, with a ratio of 0.28, was relatively open to it. Yet in bothcountries consumption of tradable goods, defined as all goods that figured in the import and export lists,was on the order of half of national income.10 If any substantial part of the national consumption orproduction of wheat, coal or cloth entered international markets in which the country in question was asmall supplier or demander, the prices of these items at home would be determined exogenously by pricesabroad. Wholesale indexes, if they do indeed consist chiefly of traded goods, are not so unrepresentative ofall of national income as might be supposed.

But what of the other, nontradable half of national income? Surely, as James Angell wrote in 1926, ‘fornon-traded articles there is of course no direct equalisation [of price] at all’.11 The operative word in thisassertion is ‘direct’, for without it the assertion is incorrect. The price of a good in one country isconstrained not only by the direct limits of transport costs to and from world markets but by the indirectconstraints arising from the good’s substitutability for other goods in consumption or production. This wasclear to Bertil Ohlin. who asked.

To what extent are interregional discrepancies in home market prices kept within narrow limits notonly through the potential trade in these goods that would come into existence if interregional pricedifferences exceeded the costs of transfer, but also through the actual trade in other goods?12

It is not surprising to find Ohlin asking such a question, for the analytical issue is identical to the one thatgave birth to that errant child of the Heckscher-Ohlin theory, factor-price equalization. The price of the milkused as much as the wage of the labour used is affected by the international price of butter and cheese. Arise in the price of a traded good will cause substitutions in production and consumption that will raise theprices of nontraded goods. To put the point more extremely than is necessary for present purposes, in ageneral equilibrium of prices the fixing of any one price by trade determines all the rest. The adjustment tothe real equilibrium of relative prices, which must be achieved eventually, can be slow or quick. Themonetary theory assumes that it is quick.


If it were in fact slow, one would expect the contemporaneous correlation between prices for countries onthe gold standard to fall sharply as more comprehensive price indexes, embodying nontraded goods, arecompared. This is not the case. The correlation between the annual changes in the GNP deflators 1880-1912for America and Britain is 0.60, to be compared with the correlation for wholesale prices alone of 0.66. Theregressions of the annual changes of American on British deflators and British on American were (standarderrors in parentheses; levels of the price variables converted to indexes with their averages as the base):

AP=0.0002+0.961 BP R2=0.35, D.–W.=1.98. Standard error of the regression as a percentage of theaverage level of the American price = 2.5%

(0.0050) (0.266)BP=0.0017+0.33 AP R2=0.34, D.–W.=1.92. Standard error of the regression as a percentage of the

average level of the British price = 1.4%(0.0028) (0.089)

The correlations of the German GNP deflator with the American (0.40) and the British (0.45) areconsiderably lower, but this may be simply a reflection of the inevitable frailties of Walther Hoffman’spioneering effort to produce such a deflator, or, perhaps, a reflection of the sharp rises in German tariffs.More countries have retail price indexes (generally with weights from working-class budgets) than havereliable GNP deflators, and these statistics tell a story that is equally encouraging for the postulate ofarbitrage. The correlation matrix of annual changes in retail prices for the United States, the UnitedKingdom, Germany, France and Sweden is shown in table 4.1. The British-American correlation (0.57) isagain not markedly below the correlation of the wholesale indexes, despite the importance of suchnontraded goods as housing in the retail indexes.13

The correlation of American with British retail prices is probably not attributable to the trade in foodoffsetting a lower correlation between nontraded goods, for the simple correlation between American andBritish food prices in the years for which it is available (1894–1913) is lower, 0.49 compared with 0.57.Against this encouraging finding, however, must be put a less encouraging one. The average correlationbetween the changes in food prices in five regions of the United States (North Atlantic, South Atlantic,North Central, South Central and the West) for 1891–1913 is very high, 0.87, contrasted with the British-American correlation of only 0.49. If food prices were as well arbitraged between as inside countries theBritish-American correlation would have to be much higher than it is. Still, even with perfect unity in themarket for each item of food, one would not expect countries with substantially different budget shares toexhibit close correlations in the aggregate indexes. The lower correlation between Britain and the UnitedStates than between regions of the United States, then, may well reflect international differences of tastesand income rather than lower arbitrage.

If one proceeds in this fashion further in the direction of less traded goods the results continue to bemixed, although on balance giving support to the postulate of unity in world markets. The most obviousnontraded good is labour. The correlation between changes in wages of British and American coal miners1891–1913 is 0.42 but the correlation between those of British and American farm labourers is only 0.26.Both are lower than the correlations between changes in the wages of the two employments in each country,0.65 in Britain and 0.53 in America. The correlation between the annual changes in Paul Douglas’s index ofhourly earnings of union men in American building and the changes in A.L.Bowley’s index of wages inBritish building from 1891 to 1901 is negligible, only 0.10. On


Table 4.1 Simple correlations between annual changes in retail prices, 1880–1912

USA UK Germany France Sweden

USA 1.00 0.57 0.28 0.24 0.38

UK 1.00 0.53 0.42 0.57

Germany 1.00 0.45 0.62

France 1.00 0.32

Sweden 1.00

the other hand, the average correlation among bricklayers’ hourly wages in four cities (Boston,Cincinnati, Cleveland and Philadelphia) selected from the mass of data for 1890–1903 in the 19th AnnualReport of the US Commissioner of Labour is only 0.14. The correlations for changes in wages betweencountries are low, in other words, but there is reason to believe that they are nearly as low within ageographically large country like the United States as well.

The same is true for an unambiguously nontraded commodity, common brick. That it is nontraded, thatis, a poor substitute for traded goods, and that it enters into the production of nontraded commodities is evidentfrom the negligible correlation between changes in its average price in Britain and America. Yet from 1894,when the statistics first become available, to 1913, the average correlation between prices of common brickat the plant in seven scattered states of the United States (California, Georgia, Illinois, New York, Ohio,Pennsylvania and Texas) was only 0.11, and even between three states in the same region (New York, Ohioand Pennsylvania) it was only 0.13. This degree of correlation may be taken as an indicator of thecorrelation between regions of the United States attributable to a common experience of general inflation,technological change and growth of income rather than to the unity of markets. It is small. In any case,common brick is a good at the lower end of the distribution of goods by their correlations, and there is littleevidence of greater integration of markets within than between countries.

All these tests can be much expanded and improved, and we plan to do so in later work.14 What has beenestablished here is that there is a reasonable case, if not at this stage an overwhelming one, for the postulateof integrated commodity markets between the British and American economies in the late nineteenthcentury, vindicating the monetary theory. There appears to be little reason to treat these two countries on thegold standard differently in their monetary transactions from any two regions within each country.

Money, gold and the balance of payments

If international arbitrage of prices and interest rates was thoroughgoing and if the growth of real income in acountry was exogenous to its supply of money, then the country’s demand for money can be estimated byrelatively straightforward econometric techniques. The balance of payments—identified here with flows ofgold—predicted by the monetary theory can then be estimated as the difference between the growth in thecountry’s total predicted demand for money and the growth in its actual domestic supply. If, further, theactual flow of gold closely approximates the flow implied by the estimated change in the demand for moneyminus the actual change in the domestic supply of money, the monetary theory of the gold standardwarrants serious consideration. In fact, to a remarkable degree the monetary theory for the United States andthe United Kingdom from 1880 to 1913 passes this final test.

In table 4.2 are presented the average movements of the British and American variables to be explained(the movements, that is, in money supplies and in that part of the money supply attributable to internationalflows of gold) and the average movements of the variables with which the monetary theory would explain


them (the movements in prices, interest rates and incomes affecting the demand for money and themovements in that part of the money supply attributable to domestic forces). The average percentagechange in the money supply was decomposed in a merely arithmetical way (described in the footnote to thetable) into a part reflecting how the money supply would have behaved if all gold flows into or out of thecountry had been allowed to affect it (by way of the multiple effects of reserves on the money supply) and aresidual reflecting all other influences. Arithmetically speaking, the causes of changes in British andAmerican money supplies differed sharply; virtually all the change in Britain was attributable tointernational flows of gold while virtually all the change in America was attributable to other, domesticsources of new money. Economically speaking, the differences are less sharp. Although over these threedecades on average the rate of change of the money supply was far larger in America than in Britain, thedifference is adequately explained in terms of the monetary theory by the faster growth of Americanincome, given the similarity (in accord with the findings of the last section) in the behaviour of prices andgiven the relative fall in American interest rates.

So much is apparent from the arithmetic of the British and American experience. To go further one needsa behavioural model explaining the annual balance of payments in terms of the monetary theory. The modelis simplicity itself. It begins with a demand function for money, the only behavioural function in the model,asserting that the annual rate of change in the demand for money balances depends on the rates of change ofthe price level and of real income and on the absolute change in interest rates (asterisks signify rates ofchange):

And it ends with a domestic money supply function (literally, an identity using the observed moneymultiplier, as explained in the footnote of table 4.2) and the statement that the money not supplieddomestically was

Table 4.2 Average annual rates of change 1882–1913 of American and British money supplies (domestic andinternational), incomes, prices and interest rates (percentages; standard errors in parentheses)

United States

1 Money supply attributableto gold flows

2.22(2.41)

−0.09(2.89)

2 Money supply attributableto other influences

0.12(2.51)

5.774.56

3 Total money supply 2.35(1.78)

5.68(5.21)

4 Real income 1.84(2.33)

3.69(5.35)

5 Implicit price deflator 0.24(1.75)

0.23(3.09)

6 Long-term interest rates(absolute change in basispoints)

2.9(2.0)

−2.3(15.0)

Sources:Line 1. The rate of change of the money supply attributable to gold flows was calculated as:where M is the total money supply, H is ‘high-powered money’ (Mt,/Ht, therefore, is the so-called ‘money multiplier’)

and R is the annual net flow of gold. The figures on money supply and high-powered money for the UnitedKingdom were taken from Sheppard (1969), p. 16; and for the United States from Friedman and Schwartz(1963), pp. 704–7. The figures on gold flows for the United Kingdom were compiled from Beach (1935), pp.


United States

46 ff. These are for England alone, excluding Scotland and Ireland, but there is little doubt that they cover thegreat bulk of flows into and out of the United Kingdom. Gold flows for the United States are given in USBureau of the Census (1960), ser. U6.

Line 2.=Line 3—Line 1.Line 3. Source as in Line 1.Line 4. US real gross national product is from Simon Kuznets’ worksheets, reported in Lipsey (1963), p. 423; for years

before 1889, the Kuznets figure Lipsey used was inferred from Lipsey’s ratio of GNP to farm income and hisestimate of farm income (pp. 423–4). UK real gross domestic product is from Feinstein (1971), appendixtable 6, col. 4.

Line 5. For the US the figure is from Lipsey (1963), p. 423. For the UK the figure is from Feinstein, appendix table 61,col. 7.

Line 6. The US interest rate is Macauley’s unadjusted index number of yields of American railway bonds (US Bureauof the Census, 1960, ser. X332). The UK rate is the yield of consolidated government bonds (consols) inMitchell (1962), p. 455.

supplied through the balance of payments. It is evident that the monetary theory is simply a comparativestatics theory of money’s supply and demand, in which the balance of payments satisfies demands formoney not satisfied by domestic sources.

By virtue of the unity of world markets and the assumed exogeneity of the growth of real income to thesupply of money (which is itself a consequence of market unity and the availability of an elastic supply ofmoney abroad), there is no simultaneous equation bias in estimating the demand for money by ordinaryleast squares. It is convenient to estimate the demand in real terms. The result for the United States 1884–1913 of regressing the rate of change of real balances on the rate of change in real income and the absolutechange in the interest rate is (t-statistics in parentheses):

(M/P)*=0.030+0.61y*−0.10� i R2=0.59(4.5) (4.9) (2.6) D.−W.=2.02

And for the United Kingdom: 15

(M/P)*=0.014+0.32y*−0.005� i R2=0.27(2.4) (2.2) (1.2) D.−W.=1.89

These appear to be reasonable demand equations, although the income elasticity in the equation for theUnited Kingdom is low, perhaps an artifact of errors in the series for income, which, given the lowvariability of British income, would reduce the fitted regression coefficient. Another explanation might bethe substantial ownership of British money by foreigners, which would reduce the relevance of movementsin British income to the ‘British’ money supply. Still, both demand equations accord reasonably well withother work on the demand for money.

The acid test of the model, of course, is its performance in predicting the balance of payments as aresidual from the predicted demand for money and the actual domestically determined supply. Itsperformance is startlingly good. The good fit of the American demand equation offsets the relativeunimportance of gold flows to the American supply, while the relative importance of gold flows to theBritish supply offsets the poor fit of the British demand equation. Figures 4.1 and 4.2 exhibit the results,comparing the actual effect of gold flows on the American and British money supplies with the predictedeffect. The actual effect is calculated annually by applying the observed ratio of money to reserves


(including gold) to the actual flow of gold, the predicted effect by subtracting the domestic sources ofmoney from the demand for money predicted by the regressions. In other words, the predicted effect is theexcess demand for money predicted by the regressions in conjunction with the actual changes in the moneysupply due to domestic sources. One could just as well make the comparison of predicted with actual flowsof gold, translating the predicted excess demand for money in each country into an equivalent demand forgold imports. The result would be the same, namely, a close correspondence between the predictions of thetheory and the observed behaviour of the British and American stock of money and balance of payments.

Figure 4.1 Predicted and actual effects of gold flows on the US money stock, annual rates of change, 1884–1913

Figure 4.2 Predicted and actual effects of gold flows on the UK money stock, annual rates of change, 1884–1913


No doubt the tests could be refined and more evidence could be examined. We believe, however, that wehave established at least a prima-facie case for viewing the world of the nineteenth-century gold standard asa world of unified markets, in which flows of gold represented the routine satisfaction of demands formoney. We do not claim to have rejected decisively the view of the gold standard that depends on poorarbitrage between national markets or the view that predicts an inverse rather than a positive correlationbetween gold inflows and income or any of the other variants of the orthodox theories. Indeed, it is perfectlypossible that these variants are partly true, perhaps true in the very short run, or under specialcircumstances, such as mass unemployment—the monetary theory is, in the sense described earlier, anequilibrium theory, which could be consistent with any number of theories about how the British andAmerican economies behaved out of equilibrium. But a balance-of-payments surplus or deficit is not initself, as has often been assumed, evidence that the economy in question is in fact out of equilibrium. Themonetary theory’s central message is that a growing, open economy, buffeted by external variations inprices and interest rates, will have a varying demand for money, which would only fortuitously be suppliedexactly from domestic sources. A country’s balance of payments, in other words, could be positive ornegative over the course of a year even if all asset and commodity markets in the country were continuouslyin equilibrium, for the flow of money into the country during the year could exactly meet the year’s changein the demand for money. The source of the simplicity of the monetary theory of the gold standard is clear:the monetary theory is an equilibrium model, whereas the alternative theories are to a greater or lesser extentdynamic, disequilibrium models. We believe (as must be evident by now) that the simpler model yields apersuasive interpretation of how the gold standard worked, 1880–1913.

Notes

1 Many of these have been published in the Princeton Studies in International Finance. For example, Bloomfield(1963) and Lindert (1969). Bloomfield (1959) is seminal to this literature.

2 Keynes (1930), II, pp. 306–7.3 World official reserves at the end of 1913 of $7,100 million (16 per cent of which was foreign exchange, a good

part of it sterling) are estimated by Lindert (1969), pp. 10–12. 4 In 1964 Robert Triffin undertook to act as midwife, but as he concedes, the infant is still in poor health (see

Triffin, 1964, appendix I).5 Needless to say, these are crude estimates: to continue the metaphor above, the historical study of world income

is barely into its adolescence. The estimate of $362 billion for 1913 world income in 1955 prices begins withAlfred Maizels’ compilation of figures on gross domestic product at factor cost for twenty-one countries, given inMaizels (1965), appendix E, p. 531. Czech and Hungarian income was estimated from Austrian income(post-1919 boundaries) on the basis of Colin Clark’s ratios among the three (Clark, 1951, p. 155). Russianincome was estimated by extrapolating Simon Kuznets’ estimate for 1958 back to 1913 on the basis of his figurefor the decennial rate of growth, 1913–58 (Kuznets, 1966, pp. 65 and 360), yielding a figure of $207 per capita in1958 prices, which appears to be a reasonable order of magnitude. The Russian per capita figure was then appliedto the population of Bulgaria, Greece, Poland, Romania and Spain, completing the coverage of Europe (boundarychanges during the decade of war, 1910 to 1920, were especially important for these countries, except Spain;estimates of the relevant populations are given in Palmer (1957), p. 193). Maizels gives estimates of nationalincome for Canada, Australia, New Zealand, South Africa, Argentina and Japan in 1913. Income per head in1955 dollars was taken to be $50 in Africa except South Africa, $100 in Latin America except Argentina, $50 inIndia, and $60 in Asia except India and Japan, all on the basis of Maizels’ estimates for 1929 and an assumptionof little growth. Population figures for these groups of countries around 1910 were taken from Glass andGrebenik (1965), p. 58, with adjustments for the countries included in Maizels’ estimates, from his populationfigures (1965, p. 540).


6 The sample is described in the appendix of the longer paper, available from the authors on request.7 From 1800–2 to 1889–91 the ratio of the Berlin to the British price of wheat increased 30 per cent and remained

at the higher ratio thereafter.8 This and all subsequent regressions were subjected to the Cochrane-Orcutt iterative technique, removing in all

cases understatement of the standard errors of the coefficients resulting from any autocorrelation of the residuals.9 We have received a good deal of enlightenment on this point from H.Gregg Lewis of the University of Chicago

and Hugh Rockoff of Rutgers University. The issue is as follows. Suppose, to simplify at the outset, that onechooses the same set of weights (w1, w2,…wN) to form the two indexes of prices (IA and IB) in the two countries Aand B). What is the relationship between the weighted average of the individual correlations,

and the correlation of the weighted averages, corr (IA, IB) (where )? For the case of two prices we havewritten out both correlations in terms of the relevant covariances (expressing the prices instandardized form, thereby eliminating variances of the individual prices and making thecorresponding covariances identical to correlation coefficients), with no very illuminating results. Ifno restrictions are placed on the covariances we can generate counterexamples to the proposition thatthe two are equal. But we suspect that we are neglecting true restrictions among the covariances (oneset implying values for another set) and, further, that the case of large N would give more usefulresults.

10 For the calculation for the UK in 1913, see McCloskey unpublished MS), ch. 1, p. 18 (MS available on request).11 Angell (1926), p. 381. Later Angell conceded in part the point made below, although he believed (p. 392) that ‘it

cannot be adequate to explain the comparatively quick adjustments [of domestic to international prices] thatactually take place’.

12 Ohlin (1967), p. 104, his italics, question mark added. Contrast Viner (1924), p. 210: The prices of services andwhat may be termed “domestic commodities”, commodities which are too perishable or too bulky to enterregularly and substantially into foreign trade, are wholly or largely independent of direct relationship withforeign prices. World price-factors influence them only through their influence on the prices of internationalcommodities, with which the prices of domestic commodities, as part of a common price-system, must retain asomewhat flexible relationship’ (his italics). Although this is an improvement on the earlier formulation byCairnes (quoted by Viner on the next page) that ‘with regard to these, there is nothing to prevent the widestdivergence in their gold prices’ it falls short of a full analysis of what is meant by ‘direct’ and ‘somewhat flexible’,an analysis provided by Ohlin. In long-run equilibrium the distinction between direct and indirect is beside the pointand the relationship of domestic to international prices is not even somewhat flexible. Viner’s work, incidentally,is one of a series of books on the balance of payments published in the Harvard Economic Studies in the 1920sand 1930s under the influence, direct or indirect, of Taussig: Williams (1920); Viner (1924); Angell (1926);Ohlin (1933); White (1933); and Beach (1935). Students of the history of economic thought will find it significantthat of these Ohlin, who acknowledges explicitly his debt to the Stockholm School (among them Cassel,Heckscher and Wicksell, all of whom emphasized the intimate relationship between domestic and internationalprices), broke most sharply with Taussig on this issue.

13 The notion of an ‘Atlantic economy’, incidentally, receives support from these figures: the average correlation ofFrench with other retail price indexes, a crude measure of the appropriateness of including a country in theAtlantic economy, is 0.36, while the same statistics for the United States is 0.37; on this reading, it would be asappropriate to exclude France from the economy of Western Europe as to exclude the United States.

14 We have passed by, for example, the issue of how unified were the markets for assets. The correlation betweenthe annual changes in the British and American long-term interest rates 1882 to 1913 included in the equationsestimated in the next section was 0.36, and could no doubt be improved by a closer attention to gatheringhomogeneous data than we have thought necessary for now. Michael Edelstein (1982, p. 339) reports acorrelation coefficient of 0.77 between annual changes in the levels of yields on first-class American railwaybonds offered in London and New York from 1871 to 1913, a period including years before the refixing of the


sterling-dollar exchange rate in 1879. The discount rates of central banks may be taken as a rough measure of theshort-term interest rate. The recent revisionist literature on the gold standard has emphasized the closecorrelations between these rates in different countries. Triffin (1964, p. 9), for example, quotes Bloomfield(1959, p. 35) approvingly, to the effect that ‘the annual averages of the discount rates of twelve [European]central banks reveal the…interesting fact that, in their larger movements at least, the discount rates of virtuallyall the banks tended to rise and fall together’. Bloomfield and Triffin attribute the parallelism to a correspondingparallelism in the business cycles of the nations involved, but the finding can also be interpreted as evidence ofdirect or indirect arbitrage in the international capital market. Lance E.Davis’s finding that the internal Americancapital market was poorly arbitraged in this period, suggests that for America at least arbitrage was little betterwithin than between countries (Davis, 1971). The widely believed assertion that domestic British industry wasstarved of funds in favour of British investment in Argentine railways and Indian government bonds can be givena similar interpretation.

15 The evidence is described in the footnote to table 4.2. The interest rate on three-month bankers’ bills (Mitchell,1962, p. 460) performed better than the consol rate, and was used here.

References

Angell, J.W. (1926), The Theory of International Prices, Cambridge, Mass., Harvard University Press.Beach, W.E. (1935), British International Gold Movements and Banking Policy, 1881–1913, Harvard Economic

Studies, Cambridge, Mass., Harvard University Press.Bloomfield, Arthur I. (1959), Monetary Policy under the International Gold Standard, 1880–1914, New York, Federal

Reserve Bank of New York.——(1963), Short-term Capital Movements under the Pre-1914 Gold Standard, Princeton Studies in International

Finance, no. 11, Princeton, Princeton University Press.Clark, Colin (1951), The Conditions of Economic Progress, 2nd edn, London, Macmillan.Davis, Lance E. (1971), ‘Capital mobility and American economic growth’, in R. W.Fogel and S.L.Engerman, The

Reinterpretation of American Economic History, New York, Harper & Row, pp. 285–300.Edelstein, Michael (1982), Overseas Investment in the Age of High Imperialism, New York, Columbia University Press.Feinstein, C.H. (1971), National Income, Expenditure and Output of the United Kingdom, 1855–1965, Cambridge,

Cambridge University Press.Friedman, Milton and Anna J.Schwartz (1963), A Monetary History of the United States, 1867–1960, Princeton,

Princeton University Press.Glass, D.V. and E.Grebenik (1965), ‘World Population, 1800–1950’, in H.J. Habakkuk and M.Postan, Cambridge

Economic History of Europe, Cambridge, Cambridge University Press, vol. VI, pt 1.Keynes, J.M. (1930), A Treatise on Money, London, Macmillan.Kuznets, Simon (1966), Modern Economic Growth, New Haven, Yale University Press.Lindert, Peter H. (1969), Key Currencies and Gold, 1900–1913, Princeton Studies in International Finance, no. 24,

Princeton, Princeton University Press.Lipsey, R.E. (1963), Price and Quantity Trends in the Foreign Trade of the United States, New York, National Bureau

of Economic Research.McCloskey, D.N., ‘Markets abroad and British economic growth, 1820–1913’, unpublished MS.Maizels, Alfred (1965), Industrial Growth and World Trade, Cambridge, Cambridge University Press.Mitchell, B.R. (1962), Abstract of British Historical Statistics, Cambridge, Cambridge University Press.Ohlin, Bertil (1967), Interregional and International Trade, rev. edn (1st edn, 1933), Cambridge, Mass., Harvard

University Press.Palmer, R.R. (1957), Atlas of World History, Chicago, Rand McNally.Sheppard, D.K. (1969), ‘Asset preferences and the money supply in the United Kingdom 1880–1962’, University of

Birmingham Discussion Papers, ser. A, no. 111 (November).


Triffin, Robert (1964), The Evolution of the International Monetary System: Historical Reappraisal and FuturePerspectives, Princeton Studies in International Finance, no. 12, Princeton, Princeton University Press.

US Bureau of the Census (1960), Historical Statistics of the United States, ser. U6, Washington DC, US GovernmentPrinting Office.

Viner, Jacob (1924), Canada’s Balance of International Indebtedness 1900–1913, Cambridge, Mass., HarvardUniversity Press.

White, Harry D. (1933), The French International Accounts, 1880–1913, Harvard Economic Studies, Cambridge,Mass., Harvard University Press.

Williams, J.H. (1920), Argentine International Trade under Inconvertible Paper Money: 1880–1900, HarvardEconomic Studies, Cambridge, Mass., Harvard University Press.


5A portfolio-balance model of the gold standard*

Trevor J.O.Dick and John E.Floyd†

The decades before World War I were a remarkable period in Canada’s economic history. Althoughsettlement of the prairies began in earnest after 1896, the entire period back to 1871 witnessed substantialgrowth in per capita incomes.1 By all accounts this growth accelerated after 1896 as western developmentresponded to rising wheat prices, the development of red fife wheat and the chilled steel plow, and the lackof opportunities for further expansion on the northern subhumid plains.2 Railroad expansion acceleratedafter the Canadian Pacific Railroad monopoly expired in 1888 and a diffused process of growth was evidentin many areas of the country down to World War I. Canadian growth was accompanied throughout by inflows of capital and labor that increased dramatically early in the twentieth century.

This paper concerns the financial adjustments associated with this real growth process. During the entireperiod up to 1914, Canada was a participant in the international gold standard. The money supplyexpanded, gold flowed into the country, and a balance of trade deficit equal to the net capital inflow plus thedebt service balance arose. Yet the mechanism through which these developments occurred remains indispute. The conventional view, originally tested by Viner (1924) and elaborated by successive generationsof economists, applied the classical price-specie-flow mechanism. This interpretation has been subject tomuch criticism, the focal point of which has been the path-breaking work of McCloskey and Zecher (1976,1984).3

We develop a new interpretation of the monetary adjustment process based on modern notions ofportfolio equilibrium within the framework of an integrated world market for capital assets. The balance ofpayments adjustment process that results from our theory is substantially different from the traditionalclassical interpretation. According to the classical view, the inflow of capital caused a balance of paymentssurplus which resulted in an inflow of gold and consequent rise in the Canadian price level. This madedomestic goods more expensive in world markets and led to a deficit in the balance of trade. We establish

* This research was financed primarily by the Social Sciences and Humanities Research Council of Canada. Some ofthe work was completed while Dick was a visiting fellow at Harvard University and while Floyd was a visiting fellowat the Research School of Social Sciences, Australian National University. We thank Gordon Anderson, CharlesCalomiris, Michael Devereux, Gerry Dwyer, Stephen Easton, Barry Eichengreen, Alan Hynes, Greg Jump, MervynLewis, Don McCloskey, Ron McKinnon, Angelo Melino, Rich Simes and seminar participants at the AustralianNational University, Northwestern University, La Trobe University, the University of Melbourne, the University ofTasmania, Stanford University, the University of Toronto, the Seminar for the Application of Quantitative Methods toCanadian Economic History, and the 1988 Annual Cliometrics Conference. An earlier version of this paper waspresented at the International Economics Association 9th World Congress in Athens in September 1989. Finally, wethank Larry Neal and an anonymous referee.† From Explorations in Economic History, 1991, pp. 209–11, 213–22, 228–38, abridged. Copyright © 1991 AcademicPress, Inc.

that essentially the opposite occurred. It was the expansion of investment associated with the growth of theCanadian economy and financed by the inflow of capital, not the inflow of monetary reserves, that raisedaggregate demand and caused the domestic price level to rise. The rise in prices had two effects. First, it ledto an increase in imports relative to exports, creating a balance of trade deficit equal to the net capital inflow.Second, together with the growth of output, it led to an increase in the demand for nominal money holdingson the part of domestic residents, who maintained portfolio equilibrium by exchanging assets for money onthe international market at world interest rates. The Canadian banking system was forced to provide thedesired additions to domestic money holdings, and thereby finance the rise in the price level, because of itscommitment to maintain convertibility of banknotes and deposits into gold. In the process, the banksacquired the observed increases in their holdings of gold and secondary reserves in New York and London.

A portfolio theoretic interpretation

The traditional interpretation views the balance of payments as essentially the sum of three independentcomponents: (a) the balance of trade, deter mined by relative prices and incomes; (b) the exogenouslydetermined long-term net capital flow; and (c) the flow of short-term capital, determined by the differentialbetween domestic and foreign interest rates. Elementary versions of the classical price-specie-flowmechanism tend to ignore capital flows, taking equilibrium to be either a zero trade balance or one equal tosome exogenously determined long-term capital flow. In practical applications of the theory, short-termcapital flows in response to interest differentials were added in ad hoc fashion as forces smoothing relativeprice level and balance of trade changes in the adjustment process.4

This treatment of international capital flows involves a fallacy of composition. Individual wealth holdersmay shift their portfolios toward securities whose interest rates are rising and away from securities whoseinterest rates are falling. But when all wealth owners behave in this way, excess demand is created forsecurities whose rates are rising and excess supply for securities whose rates are falling. The higher ratesfall and the lower rates rise until wealth holders are prepared to hold the outstanding stocks of securities.Remaining differences among interest rates in a well-functioning capital market result from the riskevaluations of investors. In the portfolio theory outlined below, the total capital flow is the differencebetween savings and investment that arises in the establishment of commodity market equilibrium. Thefraction of this flow that is short-term depends on the process of international financial intermediation andwealth holders’ division of their portfolios between money and nonmonetary assets.

Due to the integrated world market for capital, Canadian interest rates were determined exogenously byinterest rates abroad and the risks of investing in Canada. This implies an adjustment mechanism completelydifferent from that postulated by the traditional interpretation. Excess money holdings in Canada do not leadto a bidding up of the domestic price level, resulting in a deterioration of the balance of trade and a goldoutflow. On the contrary, they lead to an exchange of money for assets on the international capital market,with the Canadian price level remaining unaffected. The price level is determined, as is the price level ofany region in a large unified currency area, by the supply and the demand on the part of world (includingCanadian) residents for domestic relative to foreign output. At this price level, domestic residents acquiretheir desired holdings of money by adjusting their portfolios of non-monetary assets at asset prices andinterest rates determined in the world market. These asset adjustments appear as exchanges of foreign anddomestic funds at the chartered banks who, in order to maintain convertibility, are therefore forced to createthe necessary stock of domestic money balances. Any attempt on the part of the banks to create a quantityof note and deposit liabilities different from that which the public wishes to hold will result in a gain or lossof reserves in New York and London together with an associated adjustment of actual note and deposit


holdings to the desired level. This provides the mechanism by which the banks adjust their reserves todesired levels. If reserves are deficient, for example, the banks simply tighten up on loans, puttingdownward pressure on their note and deposit liabilities. The public sells assets in the international market tomaintain cash holdings. This results in an inflow of foreign funds, thereby increasing bank reserves. Thus,while the banking system can control the division of its assets between domestic loans and short- and long-term foreign assets and gold, it cannot control the size of its note and deposit liabilities or the domestic stockof money.

This leads to a straightforward interpretation of Canadian monetary and balance of payments developmentsin the gold standard period. The massive inflow of capital resulted from expanding domestic investmentopportunities. This externally financed investment flow resulted in an excess demand for domesticresources, causing the domestic price level to be bid up. The rise in prices led to an increase in desiredholdings of money balances. The banking system, in turn, was induced to create the necessary note anddeposit liabilities to satisfy this increased demand for money by its desire to profit from domestic loans,maintain convertibility of its notes and deposits, and control its risk by holding adequate levels of reserves.The inflow of gold and secondary reserves was a response to the increase in the demand for money, ratherthan the driving force behind the expansion of money, credit, and prices as the traditional interpretationwould have us believe. The quantity of Dominion notes held by the public was determined by the demandfor banknotes of denomination under $5. The banks’ holdings of Dominion notes were determined by thelegal reserve requirements imposed by the government. The government had two avenues of influence onthe country’s stock of international reserves—it could acquire increased gold to back its Dominion noteliabilities and it could increase the legally required ratio of Dominion notes to gold reserves of the charteredbanks. The latter, unable to draw down their Dominion note reserves below required levels to meetunexpected demands for cash, would be forced to hold greater total reserves by an increase in the legalrequirement. Apart from these influences, the monetary base and the stock of international reserves wereendogenously determined. The government had no control over the stock of money.

In developing these ideas more rigorously, we first examine price level and trade balance determinationin a world where capital is internationally mobile. We then examine the determinants of asset equilibriumand balance of payments adjustment under capital mobility.5

The Canadian price level and balance of trade

A major consequence of the inflow of capital into Canada was the pressure this foreign investment put onCanadian resources. Like Viner,6 we note that the capital inflow, domestic capital formation, and risingprices of nontraded goods occurred together. The inflow of capital was not dissipated in the import offoreign goods and labor, but resulted in a net increase in aggregate demand that exerted upward pressure onthe prices of domestic nontraded goods relative to the prices of traded goods and foreign nontraded goods.Our view differs from Viner’s, however, in that this price movement, we believe, was a real rather than amonetary phenomenon. The domestic price level incorporates the prices of both traded and nontraded goodsand rose mainly in response to the upward movement of nontraded goods prices. It was mainly this elementof the Canadian price level that distinguished its movement from the movements of other countries’ pricelevels. As a small country, Canada experienced a movement in the traded goods component of her pricelevel similar to that of other countries.7

The essence of Canadian price level determination in this context can be captured by a highly simplifiedmodel that includes only one internationally traded good and two nontraded goods. The point of departure

56 PORTFOLIO-BALANCE MODEL OF THE GOLD STANDARD

for this demonstration is the equality between the total payments and the receipts of domestic residents in asmall open economy.

(1)where PU and PT are the prices of nontraded and traded goods, respectively, U and T are the domesticoutputs of the respective goods, UC and TC are the quantities of them consumed, 5 is domestic savings, andDSB is net repatriated earnings (interest and dividend receipts from abroad minus the correspondingpayments to foreigners), sometimes referred to as the debt service balance. Equation (1) says that totalearnings from the production of traded and nontraded goods plus net earnings from abroad must be eithersaved or spent on the consumption of traded and nontraded goods.

Total domestic investment expenditures can be expressed as

where UI and TI are the quantities of the two goods absorbed into investment. Adding and subtracting thelevel of investment on the right-hand side of (1) yields

(3)

Equality of the demand and supply of the nontraded good implies(4)

Combining (3) and (4) yields(5)

The balance of trade plus the debt service balance plus the net capital inflow must sum to zero as acondition of real goods market equilibrium.

The equilibrium relative price of nontraded goods can be shown to depend on a set of real factorsexogenous to the domestic economy. Equations (4) and (5) can be expanded by imposing some standardbehavioral relations.8 Consumption depends on income with the division between traded and nontradedgoods depending on the price ratio PU/PT. Investment depends on the level of output, the rate of interest andtechnological change and natural resource discoveries. Again, the division between nontraded and tradedgoods depends on relative prices. On the supply side, aggregate output depends on the stocks of labor andcapital in the economy and on technology. And the allocation of output between nontraded and tradedgoods depends on relative prices. Finally, aggregate income depends on aggregate output, the debt servicebalance, and the exogenously determined terms of trade. The equilibrium relative price of nontraded in termsof traded goods can be obtained by substituting these behavioral relations into (4) and rearranging theresulting expression to bring PU/PT to the left-hand side.9 The result is of the form

(6)where N and K are the stocks of labor and capital, r is the real rate of interest, µ is the terms of trade, and �is a portmanteau variable incorporating technological change and natural resource discoveries.

The effects of changes in N and K on the relative price variable are ambiguous because they shift the demandfor nontraded goods and the supply in the same direction. A fall in r increases investment and the demandfor the nontraded good, causing PU/PT to rise. An improvement in the terms of trade raises income andconsumption, again increasing the demand for and relative price of the nontraded good.10 The effect of theportmanteau variable is positive by construction—technological change permits the development of newland, leading to an increase in the demand for nontraded goods. A decline in the debt service balance overtime resulting from net inflows of capital reduces the income associated with each level of domestic outputand results in lower relative prices of non-traded goods than otherwise.


The domestic price level, defined in terms of the prices of goods produced, can be expressed as ageometrically weighted index of traded and nontraded goods prices. Taken in combination with (6), thisgives

(7)where � is the share of the nontraded good in domestic output. The price of traded goods can now beexpressed in terms of the price index of goods produced abroad as

(8)

where q* is the ratio of nontraded to traded goods prices in the rest of the world, P* is the index of outputprices abroad, and � is the share of nontraded goods in the rest of the world’s output. Substitution into (7)yields

(9)

Since the Canadian real interest rate equals the foreign rate plus a risk premium, (9) can be rewritten as

(10)

where � is the risk premium on domestic assets. Equation (10) reduces to (9) when there is perfect capitalmobility.

The Canadian economy was too small to have any influence on the price level and interest rates in therest of the world, so (10) is sufficient to determine the domestic price level. Domestic goods were more orless expensive than the rest of the world’s goods in accordance with the conditions of their production andthe extent of the demand for them on the part of world (including domestic) residents. Whether the pricelevel at home was high or low relative to foreign prices depended therefore on real factors relating toproduction and consumption. The supply and demand for money in the domestic economy could only affectthe price level by influencing the risk premium.

Canadian asset equilibrium under the gold standard

The outstanding feature of the balance of payments adjustment process under the gold standard was the easewith which Canadian bankers and wealth holders in general could adjust their holdings of domestic and foreignsecurities in the face of disturbances to their portfolio equilibria wrought by the activities of foreigninvestors. The existence of an international capital market and the telegraphic communication that gaveCanada a presence in important financial centers like London and New York established the type ofenvironment in which the portfolio adjustments necessary to maintain balance of payments equilibrium,given fixed gold parities, could take place with relative ease. Interest rates in Canada and abroad could thusadjust to levels at which wealth holders were content to hold the existing stocks of domestic and foreignsecurities.

Although capital was internationally mobile in the sense that there was little or no restriction on theportfolio compositions of wealth holders, it was doubtless the case that capital was less than perfectlymobile internationally. This means that wealth holders regarded domestic and foreign assets as less thanperfect substitutes. Imperfect substitutability was reflected in interest rate differentials that incorporated riskpremia. This does not mean that international trade in capital assets was in any important way artificially


restricted. Exclusive of these risk premia, it is still likely that the domestic interest rate was exogenouslydetermined since Canada was a relatively small economy participating in the world capital market.11

The keys to the portfolio adjustment model are the equations determining the demand for money and thestock of foreign exchange reserves,

(11)(12)

where M is the nominal money stock, R is the domestic stock of international reserves, L(r*, � , Y) is thedemand function for money, � (r*+� , V) is the equilibrium ratio of international reserves to the moneysupply, and V is a scale variable representing the size of the banking system.12

The demand for money depends on both domestic and foreign interest rates (or equivalently, on theforeign interest rate and the risk premium) and reduces to the standard form L(r, Y) when capital is perfectlymobile.13 In either case, the arguments in the function are exogenous and cannot adjust in the face of anyexcess demand or supply of money as they would in a closed economy. Instead, the nominal money supplymust adjust endogenously to changes in demand.

Balance of payments adjustment under the gold standard ensures this endogeneity. If, for example, thebanking system through its domestic loan and discount policies creates less money than domestic residentswant to hold, domestic residents have an easy avenue to acquire more money. They simply sellnonmonetary assets in the world market and take the foreign currency proceeds to the chartered banks forconversion into domestic currency. The banks acquire international reserves and create domestic money inequal amounts.

The chartered banks controlled their reserve levels, we argue, by appropriately adjusting their domesticcredit policies.14 Neither they nor the government through its Dominion note issue had any control over thedomestic money supply. The latter always equaled whatever domestic residents wanted to hold. The bankscontrolled only the division of their asset portfolios between domestic loans and discounts and internationalreserves.

Equation (12) taken in conjunction with (11) shows that the country’s stock of international reservesdepends on the demand for money and the banking system’s reserve ratio. In the Canadian pre-1914context, the public’s desired money holdings determined the banking system’s note and deposit liabilitiesand the outstanding stock of small denomination Dominion notes. The various factors, such as interest rates,that determined the profit maximizing reserve ratios of banks, and the government’s chosen gold reserveratio against its Dominion note liabilities translated this demand for money into a demand for internationalreserves.15 Changes in the gold stock and the stock of reserves had no causal effect on the domestic pricelevel. Indeed, the opposite was the case. Increases in the price level emanating from real forces of demandand supply and/or monetary developments in the rest of the world caused the public to hold more moneywhich in turn caused the banks to create that money and hold larger reserves.

The balance of payments surplus is the flow of increases per unit time in the country’s stock ofinternational reserves. Combining (11) with (12) and differentiating with respect to time holding the riskpremium constant, we obtain

(13)


where the � j, and Lj, (j=r*, y, v), are the partial derivatives in the functions � (r*+� , V) and L(r*, � , Y). dP/dtand dY/dt are the derivatives of the equilibrium levels of P and Y with respect to time, dV/dt is exogenous(but could possibly be represented by some function of dY/dt or dM/dt), and dr*/dt is determined abroad.

A final perspective on the nature and implications of asset equilibrium is provided by observing thataggregate savings in Eq. (5) can be split into two components—the banking system’s accumulation of goldand secondary reserves, denoted by dR/dt, and all other accumulations of assets by domestic residents on bothprivate and public account, denoted by S’.

(14)Since the prices of nontraded and traded goods and the quantities of both goods produced and consumed areunaffected by the change in the level of bank reserves, the flow of reserves has no effect on savings. Thus,S’ declines (increases) dollar for dollar with increases (declines) in dR/dt. Changes in the balance ofpayments surplus or deficit must therefore have arisen solely from changes in the proportion of theeconomy’s savings that took the form of government gold and gold and secondary reserve accumulations ofthe chartered banks. This result does not depend on whether capital mobility is perfect or imperfect. Even ifthe risk premium varies through time, imperfection of capital mobility can have no fundamental effect onthe process by which balance of payments adjustments occur.

It is evident from the above that the existence of Dominion notes does not alter our interpretation ofbalance of payments adjustment. Subject to the reserve requirement and the prohibition against smalldenomination chartered bank notes, the quantity of Dominion notes held was voluntary. As Rich (1988)points out, the government did attempt to increase the outstanding stock of Dominion notes relative to its goldbacking in the early decades of Confederation and thereby economize on the cost of servicing the publicdebt.16 But we regard this as fiscal rather than monetary policy. Although it undoubtedly reduced bankprofits, it did not affect the money supply—the public could adjust its money holdings by buying and sellingassets abroad, thereby forcing the banks to create whatever money supply it desired. The governmenttherefore could not use Dominion notes as an instrument of monetary policy as the term is usually defined.Changes in reserve requirements and in the size of the gold stock used as backing for Dominion notes wouldhave resulted in net changes in international reserves and in the observed flow balance of paymentsadjustments, but these changes must have been exogenous as far as the balance of payments adjustmentmechanism is concerned.

The portfolio theoretic reinterpretation of the Canadian evidence

The interaction of capital flows, relative prices and the balance of trade, and the behavior of banks is alsoconsistent with the portfolio approach to the mechanisms of adjustment in Canada’s balance of paymentsunder the gold standard. It remains to present the quantitative evidence that lends itself to this interpretation.

The portfolio theory argues that some part of the new capital implanted in the 1900–1913 expansionconsisted of Canadian nontraded goods, so aggregate demand increased. This should have been reflected inan increase in the price level taking the form of an increase in the prices of nontraded goods—traded goodsprices were determined almost entirely abroad. The associated rise in the costs of producing traded goodsshould have resulted in a reduction in domestic production of export and import goods, leading to anexpansion of imports relative to exports which reduced aggregate demand sufficiently to ‘make room’ forthe increased domestic production of new capital goods. The trade deficit must have been just sufficient tofinance, in real terms, the inflow of capital.

The regression results reported in Table 5.1 confirm this story. Regressions (1) and (2) show that therewas a rise in the price of domestic nontraded goods relative to traded goods and a fall in the price of


nontraded relative to traded goods abroad.17 Regression (3) shows that nontraded goods prices in Canadaalso rose relative to nontraded goods prices in the rest of the world.

Table 5.1 Capital inflows, relative prices, and the balance of trade

(1) (2) (3) (4) (5)

PNTCANa/PTCANc

PTWLDb/PNTWLD

PNTCAN/PNTWLD

PNTCAN/PNTWLD

PNTCAN/PNTWLD

Constant 96.44(107.63)a

111.30(75.84)

107.14(74.60)

108.05(34.62)

106.72(9.10)

Real NCIe+DSBf 0.08(4.91)

0.14(5.07)

0.25(8.98)

0.26(6.65)

0.26(3.54)

Canadian realGNP

−0.01(−0.33)

(−0.05)(−0.15)

Foreign real GNP 0.05(0.12)

NOBS 43.00 43.00 43.00 43.00 43.00

R-SQ 0.37 0.39 0.66 0.66 0.66

SEE 5.03 8.24 8.06 8.15 8.25

DW 0.84 0.25 0.66 0.68 0.69

Notes:a PNTCAN=price index of nontraded goods: Canada; PNWLD=price index of nontraded goods: rest of world.b PTCAN=price index of traded goods: Canada.c PTWLD=price index of traded goods: rest of worldd The figures in parentheses are t ratios.e NCI=net capital inflow.f DSB=debt service balance; real NCI+DSB=minus the trade account balance.

A more complete specification of relative price determination suggested by our theory is incorporated inregressions (4) and (5) of Table 5.1. An expansion of real output and income in a country increases both thedemand and the supply of nontraded goods and may lead to a rise or fall in their prices relative to the pricesof traded goods. To the extent that the increase in output and income is permanent, the effect on relativenon-traded goods prices depends upon the relative biases toward or away from the nontraded good of (a) theimprovement of technology that leads to the increase in output and (b) the consumption effect of theresulting increase in permanent income. If the increases in output and income are only transitory, theproduction, but not the consumption, of nontraded goods rises causing the price of the domestic nontradedgood to fall relative to the price of the foreign nontraded good. Regression (5) indicates no relationshipbetween Canadian and foreign real income and Canadian relative to foreign nontraded goods prices. Thatone of the income variables alone is not significant (column 4) suggests that the insignificance of the twovariables when included together is not due to multicollinearity.18 A plot of the deviations of real GNParound trend on the same chart with the real net capital inflow plus debt service (Figure 5.1) indicates thatthe major movements of the two series are broadly the same—addition of the real income variables to theregression thus contributes nothing.19

These results, cast above in the framework of the portfolio theory, are also consistent with the traditionalapproach. In the specie-flow mechanism, the inflow of capital also leads to a rise in the domestic pricelevel. In this case, however, monetary expansion causes the price level change.20 Again, because the prices


of traded goods are determined abroad, most of the rise in domestic prices falls on the non-traded good. Theserelative price adjustments create a trade balance deficit equal to the net capital inflow.

A more traditional presentation of the relationship between the net capital inflow and relative priceswould treat the balance of trade (equal to minus the sum of the net capital inflow and the debt servicebalance) as determined by domestic and foreign real incomes and relative nontraded goods prices. Thistraditional view of the trade balance as determined by relative prices and incomes masks the importantcausal influence of the net capital flow. The correct interpretation is that an observed increase in the tradebalance deficit and associated rise in domestic relative to foreign nontraded goods prices are bothconsequences of an increased net inflow of capital. The relative price adjustment is necessary to create asufficient deterioriation of the trade balance to finance the inflow of capital. The income variables areinsignificant in columns (3) and (4) of Table 5.1 because increases in world income affected the tradebalance and the net capital flow to the same degree—no relative price adjustment was required to maintainequilibrium.

The portfolio approach also makes international reserve flows a direct consequence of adjustments of thesupply of money to the demand together with adjustments of the banking system’s desired division of itsassets between domestic loans and discounts and primary and secondary reserves. It requires only that thepublic’s demand for money and the reserve ratios of the chartered banks be stable functions of the usualvariables.21

Regressions of the money stock on real income and interest rates are reported in Table 5.2. Although thedemand function for money fits well, there is positive serial correlation in the estimated residuals for theperiod as a whole and negative serial correlation for the subperiod prior to 1890, suggesting possiblestructural changes in the financial system between the two periods.22

The regression results reported in Table 5.3 provide further evidence of long-term structural change. Theprimary reserve ratio of the chartered banks is negatively related to the scale of the banking system(approximated by real bank liabilities) prior to 1890 and positively related thereafter. The secondary reserveratio is insignificantly related to the scale variable before 1890, but positively and significantly related to itafter 1890. A visual examination of the relevant series indicates clearly that there was a sharp increase inmoney holdings, a shift toward deposit banking, and a general decline in interest rates in Canada relative toGreat Britain in the years before 1890.23 The effects of short-term interest rates

Figure 5.1 Capital inflow and real exchange rate


Table 5.2 The demand function for money

Log of real money stock

1871–1913 1871–1889 1890–1913

Constant −0.92(−3.37)a

0.37(0.44)

−0.85(−4.52)

Log of Canadian real GNP 1.53(38.06)

1.20(9.23)

1.55(40.95)

Canadian long-term interest rate −0.15(−4.76)

−0.14(−1.90)

−0.18(−6.24)

UK treasury bill rate −0.04(−2.78)

−0.04(−2.23)

−0.07(−5.12)

NOBS 43.00 19.00 24.00

R-SQ 0.99 0.99 0.99

SEE 0.08 0.04 0.05

DW 0.89 2.88 1.90

Note:aThe figures in parentheses are t ratios.

on the bank reserve ratios are negative and significant24 for the period after 1890, but insignificant for theearlier period. Whatever may be the correct detailed interpretation of these results, they are consistent withthe banks having viewed themselves as managing a portfolio of assets with both domestic and foreigncomponents.25

Rich (1988, 1989), in his examination of the crisis of 1907, presents two additional pieces of evidenceconsistent with our portfolio interpretation of balance of payments adjustment. First, he finds no significantrelationship between seasonal variations in the balance of trade and the balance of payments surplus.26

Second, he establishes that the rise in world interest rates during the panic of 1907 led to a reduction in thedemand for money balances in Canada and consequent downward pressure on chartered bank reserves. Thebanks would have been forced to cut domestic loans to maintain their reserve ratios at satisfactory levelshad not the government, under pressure from Western interests, supplied them with advances in the form ofDominion notes against the collateral of high-grade securities. The banks then redeemed some of theseDominion notes in gold. In effect, the chartered banks were allowed to maintain their reserve levels in theface of a decline in the nominal money stock by borrowing from the government rather than reducing theirlending to grain exporters in peak season. Had the banks reduced their loans, exporters would have had toborrow abroad, with the conversion of these borrowed funds into Canadian dollars providing the desiredinflow of bank reserves. Note that the money supply is

Table 5.3 Reserve ratios of the chartered banks

Primary reserve ratio Secondary reserve ratio

1871–1913 1871–1889 1890–1913 1871–1913 1871–1889 1890–1913

Constant 0.13(8.31)a

0.26(14.13)

0.11(13.55)

0.19(6.36)

0.24(2.12)

0.13(7.80)

UK T-BILL rate 0.01(1.67)

0.01(−1.47)

−0.01(−2.07)

−0.02(−1.96)

−0.04(−1.56)

−0.02(−3.05)

Real bank liabilities −0.07 −0.80 0.05 −0.003 0.24 0.08


Primary reserve ratio Secondary reserve ratio

1871–1913 1871–1889 1890–1913 1871–1913 1871–1889 1890–1913

(−3.90) (−9.87) (3.59) (−0.10) (0.48) (2.97)

NOBS 43 19 24 43 19 24

R-SQ 0.29 0.87 0.38 0.09 0.21 0.37

SEE 0.03 0.01 0.01 0.05 0.07 0.03

DW 0.39 1.97 1.34 1.07 1.17 1.40

Notes:aThe figures in parentheses are t ratios.

unaffected by these actions—the sole issue is the composition of the chartered banks’ assets and liabilitieson the one hand and the locus of borrowing by domestic exporters on the other.

Conclusions

This study presents a reinterpretation of the Canadian monetary experience between 1870 and 1913 in termsof a portfolio model of balance of payments adjustment that treats asset markets in a worldwide generalequilibrium framework with imperfect capital mobility. The portfolio theory successfully explains theevidence. In contrast, the price-specieflow mechanism traditionally used to explain balance of paymentsadjustment in this period of Canadian history fails to explain key elements of the adjustment process.

The general equilibrium features of the portfolio model stand in marked contrast to the partial equilibriumcharacter of the traditional price-specieflow approach. Interest rates reflect relative risk evaluations ofinvestors who make long-run commitments around the world. Capital movements reflect differencesbetween local saving and investment, and international reserve flows (which reflect the composition of theoverall capital movement) represent portfolio adjustments of domestic money balances and chartered bankreserve holdings to desired levels. Interest rates are the result of aggregate portfolio choice, not the cause ofit. Capital mobility provides an avenue of adjustment that consistently links together stock and flowelements of general equilibrium. Long-term capital inflows, induced by domestic investment opportunities,raise the Canadian price level relative to the foreign price level, inducing portfolio changes which lead toreserve inflows and bring the supply and demand for money into equality. The money supply is endogenousand adjusts through the balance of payments without affecting the balance of trade.

The analysis presented here supports the general tenor of McCloskey and Zecher (1976) and significantlyextends that work. Our argument is based on a rigorous development of the implications of internationalcapital mobility while theirs is based heavily on a qualified concept of purchasing power parity. McCloskeyand Zecher (1976) devote much attention to the empirical issue of how well international commoditymarkets worked, carefully distinguishing traded and nontraded goods, because they view integratedcommodity markets as an indispensable part of their monetary approach to balance of payments adjustment.Price movements result from commodity arbitrage subject to transaction costs. For them, purchasing powerparity appears simply to be general equilibrium pricing. In the approach of the present paper, it is sufficientto assume that commodity markets exist and that some goods are traded internationally. Undoubtedly,changes occurred over time in transactions costs and, in turn, in the empirical manifestations of marketclearing and arbitrage. But the conclusions of the present paper rest on capital mobility rather thancommodity market arbitrage as the essential basis for the monetary nature of the adjustment process.


The results of this study raise serious questions about how the gold standard really worked and suggestpromising avenues for further research. How should the behavior of the Bank of England be interpreted inthe light of this theory? If the authorities in individual countries could not control their domestic moneystocks, as the portfolio theory implies, could they have been guilty of violating the ‘rules of the game’? Ifnot, why could the gold standard not have been successfully reconstituted after the War? These questionsmerit investigation.

Data appendix

The annual real output series used for Canada is based on the Urquhart (1986) new estimates of nominalnational income since 1870 deflated by an historical price index of traded and nontraded goods pricesreconstructed from existing series by Dick [Dick and Floyd (forthcoming), Appendix C]. The nationalincome estimates are based on value added calculations and incorporate a revised balance of payments.These estimates are the most exhaustive ever made for Canada and utilize data from the census, tradeaccounts, and numerous other sources.

The deflator incorporates (in its traded goods component) wholesale market price quotations and (in itsnontraded goods component) the residential construction costs series prepared in connection with the newnational income estimates of residential construction. These are weighted together with information fromthe worker surveys of the Ontario Bureau of Industries for the 1880s. The Urquhart deflator is not usedbecause it links together several indexes that are inconsistent in principle and obscures the distinctionbetween traded and nontraded goods price movement so essential to the present study. The total indexconstructed for this study nevertheless displays the same general movements as many of the other existingseries.

Real GNP data and the implicit deflator for the rest of the world are weighted averages (using as weightsGNP converted to a common currency with existing exchange rates) of the official statistics for the UnitedStates (Friedman and Schwartz, 1982); the United Kingdom (Capie and Webber, 1985), Italy (Fratianni andSpinelli, 1984); and Germany, France, Norway, and Sweden (Mitchell, 1975), except that France wasomitted from the rest of the world GNP and a French cost-of-living index was used for prices because ofFrench data deficiencies. To construct a rest of the world index of nontraded goods prices Dick [Dick andFloyd (forthcoming), Appendix A] aggregated appropriate series for the United States and the UnitedKingdom, Canada’s principal trading partners, only. These series, like the nontraded goods series forCanada, are dominated by residential rents and construction costs.

Money stock data for Canada is drawn from Rich (1988), who mostly adopts the older series of Curtis(1931) with minor revisions. Money includes Canadian dollar bank deposits held by Canadian residents,chartered bank notes held by the public and the government of Canada, and Dominion notes held by the publicin Canada. Ideally, all government deposits should be included, but Rich was only able to uncoverprovincial deposits. Also, it is assumed, not unreasonably, that most deposits in Canada were owned byCanadian residents and few Canadian bank notes were foreign owned.

The rest of the world money supply is a weighted sum of nominal money in the United States, the UnitedKingdom, France, Germany, Italy, Norway, and Sweden (Mitchell, 1975). Typically the money stock inthose countries included government or central bank notes and currency, both demand and time deposits atthe commercial banks, and savings bank deposits. The sources are the same as those for the rest of the worldincome and prices.

Ideally, one would want interest rates on Canadian securities denominated in Canadian dollars andmarketed in Great Britain, the source of most of the long-term capital before 1914, and in New York, the


other lesser source of capital inflow at this time. In lieu of this we use the average rate on Canada sterlingbonds maturing in 1903 and 1938 as reported in Neufeld (1972). World interest rates are taken to be anaverage of the rates prevailing in the United States and the United Kingdom, principally in New York andLondon. Long-term rates for US high-grade corporate bonds, for US high-grade industrial bonds, and forUK consols are taken from Friedman and Schwartz (1982). At the short end of the market, Goodhart (1969)provides monthly rates for the years 1902–1913 at Montreal and New York and other financial centers.

Notes

1 This has been documented by the recently revised estimates of national growth after 1871. See Green andUrquhart (1987).

2 For an analysis of the supply and demand factors influencing prairie wheat expansion, see Dick (1980).3 For earlier criticisms, see Bloomfield (1959, 1963, 1968), Ford (1962), Triffin (1964), and Williamson (1961,

1964). The depth of the controversy is evident in the recent retrospective volume edited by Bordo and Schwartz(1984).

4 These ideas were later incorporated explicitly by Mundell (1960, 1961, 1963, 1968) in what has come to beknown as the IS-LM-BB framework.

5 The theoretical roots of this paper extend back to Fleming (1962) and Mundell (1963). While the Fleming-Mundell approach gives essentially correct answers when capital is perfectly mobile, it cannot be generalized tomodel imperfect capital mobility correctly because of its failure to distinguish between stocks and flows.Building on the early work of Johnson (1958), more explicit treatments of stocks and flows by Floyd (1969),Harkness (1969), and others followed. This line of analysis was temporarily eclipsed by the monetary approachin the early 1970s and then reemerged in the unifying work of Frenkel and Rodriguez (1975), Mussa (1982), andFloyd (1985).

6 Viner (1924, pp. 249–50).7 Canada was a price taker in the markets for traded goods and her securities bear interest rates equal to world

interest rates on risk equivalent securities. Canada produced around 10 per cent of world wheat output at the endof the period. And wheat made about 20 per cent of Canadian exports. A change in wheat production musttherefore have had little effect on the terms of trade. While Canada produced a large fraction of the world supplyof some other goods (e.g., nickel), none of these constituted a significant fraction of total domestic exports.Canada did not import a significant fraction of the world supply of any product before 1914. The only wayCanada’s price level movements could have differed from those of other countries on account of the movementof traded goods prices would have been from the inclusion of a different set of traded commodities differentlyweighted in the relevant indexes. In the empirical work such differences are automatically incorporated in theindexes we use.

8 A more technically rigorous presentation of the argument below is given in a Mathematical Appendix availablefrom the authors on request.

9 Identical results are obtained by substituting the behavioral relations into Eq. (5).10 When the terms of trade are exogenous, exports and imports can be aggregated into a single good, so extension

of the model to four goods would complicate it unnecessarily. A change in the terms of trade can be viewed as achange in the export relative to the import price component of the index of traded goods prices without a changein PT itself.

11 In the absence of risk premia, there would be a single world real interest rate. Observed interest rates, however,which equal the real rate plus an inflation premium would differ across countries on account of different inflationrates brought about by movements in the relative prices of national outputs—i.e., by movements in real exchangerates. Differences in inflation rates arising from this source were not large [See Floyd (1985), p. 77].

12 In simple terms, the theory underlying these two equations below is as follows. Each country’s residents hold theportfolio mix of domestic assets, foreign assets, and money that maximizes utility. Any given portfolio mix in


each country generates two marginal rates of substitution—between domestic physical capital assets and moneyand foreign physical capital assets and money. These two marginal rates of substitution are the reciprocals of theinterest rates at which the country’s residents are prepared to hold that particular mix of assets in their portfolios.Assets are exchanged among domestic and foreign residents until the interest rate at which both countries’residents hold each country’s assets is the same. These levels of domestic and foreign interest rates aredetermined simultaneously with the equilibrium portfolio mixes of the residents of the two countries. The excessof the domestic over the foreign interest rate is � , the risk premium on domestic assets in the world capital market.For a more rigorous development of these ideas, the reader is again referred to the Mathematical Appendix.

13 Zero expected inflation and full employment are assumed. It is well known that there was virtually no correlationbetween interest rates and inflation rates under the gold standard. See Barsky and Summers (1988). Sargent(1973) also found evidence of very long adjustment lags in US data symptomatic of nearly static expectations.

14 For a discussion of bank lending policies, see Evans and Quigley (1990).15 The chartered banks’ gold holdings as a proportion of their total reserves also depended on interest rates and

profit maximizing considerations.16 Rich (1988), Chapter 3, p. 71.17 This need not imply that developments in Canada have an effect on world relative prices since the countries

comprising the rest of the world in our regressions are also exporting capital to places other than Canada duringthe period. In our study, the rest of the world consists of the United States, the United Kingdom, Germany, Italy,Norway, Sweden, and France.

18 The null hypothesis that the coefficients of both income variables together are zero would not be rejected at eventhe 90 per cent level.

19 Both the ratio of nontraded goods prices in Canada to nontraded goods prices abroad and the real net capitalinflow trended downward before 1897 and upward thereafter. A rerun of regression (3) with a correction for first-order autocorrelated errors, however, yields no statistical relationship between the two variables and anautocorrelation coefficient near unity. This indicates that the year to year movements of the series around thesetrends are not correlated—a not unexpected result given the general misspecification of the equation.

20 An attempt to use Granger-Sims causality analysis here was not fruitful. Variations in each of nominal money,prices, and nominal income were regressed on their own past in addition to past movements of the other twovariables. The lagged values of the other variables contributed nothing in addition to what was explained by thedependent variables’ own past. This indicates that the movements in the three variables were simultaneous.

21 We assume that the government’s ratio of gold reserves to its Dominion note liabilities was exogenous.22 The Ljung-Box Q statistic indicates autocorrelation in the residuals for the whole period but not for either

subperiod.23 See Dick and Floyd (forthcoming, Chapter 4).24 At the 5 per cent level in the case of the primary reserve ratio and the 1 per cent level in the case of the secondary

reserve ratio.25 Our results could mean that, in the post-1890 period at least, a rise in the level of short-term interest rates

increased the costs of holding both primary and secondary reserves relative to other assets. Alternatively, theycould mean that in boom periods, when short-term interest rates were high, the banks tended to hold lower ratiosof reserves to liabilities. To unravel satisfactorily the effects of interest rates on bank reserves, we would needinterest rate data on all the major assets in the banks’ portfolios.

26 Rich (1988, Chapter 6).

References

Barsky, R.B. and L.H.Summers (1988), ‘Gibson’s Paradox and the Gold Standard’, Journal of Political Economy, 96,528–50.


Bloomfield, A.I. (1959), Monetary Policy under the International Gold Standard, New York, Federal Reserve Bank ofNew York.

——(1963), Short-Term Capital Movements Under the Pre-1914 Gold Standard, Princeton Studies in InternationalFinance 11, Princeton, Department of Economics, Princeton University.

——(1968), Patterns of Fluctuation in International Investment Before 1914, Princeton Studies in International Finance21, Princeton, Department of Economics, Princeton University.

Bordo, M.D. (1984), ‘The Gold Standard: The Traditional Approach’, in M.D. Bordo and A.J.Schwartz (eds), ARetrospective on the Gold Standard, NBER Conference Report, Chicago, University of Chicago Press,pp. 23–113.

Bordo, M.D. and A.J.Schwartz (eds) (1984), A Retrospective on the Classical Gold Standard, 1821–1931, NBERConference Report, Chicago, University of Chicago Press.

Capie, F. and A.Webber (1985), A Monetary History of the United Kingdom: 1870–1982. Vol. 1, Data, Sources,Methods, London, Allen & Unwin.

Curtis, C.A. (1931), Statistics of Banking. Vol I of Statistical Contributions to Canadian Economic History, Toronto,Macmillan.

Dick, T.J.O. (forthcoming), ‘Canadian Wheat Production and Trade, 1896–1913’, Explorations in Economic History,17, 275–302.

Dick, T.J.O. and J.E.Floyd, (forthcoming), ‘Canada and the Gold Standard, 1871–1913’, Toronto, University ofToronto (mimeo).

Eichengreen, B. (1987), ‘Conducting the International Orchestra: Bank of England Leadership under the GoldStandard’, Journal of International Money and Finance, 6, 5–29.

Evans, L.T. and N.C.Quigley (1990), ‘Discrimination in Bank Lending Policies: A Test Using Data from the Bank ofNova Scotia’, Canadian Journal of Economics, 23, 210–25.

Fleming, J.M. (1962), ‘Domestic Financial Policies under Fixed and Flexible Exchange Rates’, International MonetaryFund Staff Papers, 39, 369–79.

Floyd, J.E. (1969), ‘International Capital Movements and Monetary Equilibrium’, American Economic Review, 49,472–92.

——(1985), World Monetary Equilibrium, Philadelphia, University of Pennsylvania Press.Ford, A.G. (1962), The Gold Standard, 1880–1914: Britain and Argentina, Oxford, Clarendon Press.Fratianni, M. and F.Spinelli (1984), ‘Italy in the Gold Standard Period, 1961– 1914’, in M.D.Bordo and A.Schwartz

(eds), A Retrospective on the Classical Gold Standard, 1821–1931, NBER Conference Report, Chicago, Universityof Chicago Press, pp. 405–41.

Frenkel, J.A. and G.G.Johnson (eds) (1976), The Monetary Approach to the Balance of Payments, London, Allen &Unwin.

Frenkel, J. and C.Rodriguez (1975) ‘Portfolio Equilibrium and the Balance of Payments: A Monetary Approach’,American Economic Review, 65, 674–88.

Friedman, M. and A.J.Schwartz (1982), Monetary Trends in the United States and the United Kingdom, Chicago,University of Chicago Press.

Goodhart, C.A.E. (1969), The New York Money Market and the Finance of Trade, 1900–1913, Cambridge, Mass.,Harvard University Press.

Green, A. and M.C.Urquhart (1987), ‘New Estimates of Output Growth in Canada: Measurement and Interpretation’, inDouglas McCalla (ed.), Perspectives on Canadian Economic History, Toronto, Copp Clark Pitman, pp. 182–99.

Harkness, J. (1969), ‘Monetary and Fiscal Policies in Closed and Open Economies: The Portfolio Approach’,unpublished Ph.D. dissertation, Queen’s University.

Johnson, H.G. (1958), ‘Towards a General Theory of the Balance of Payments’, in International Trade and EconomicGrowth, London, Allen & Unwin, Chap. 6.

McCloskey, D.N. and J.R.Zecher (1976), ‘How the Gold Standard Worked, 1880–1913’, in J.A.Frenkel andH.G.Johnson (eds), The Monetary Approach to the Balance of Payments, London, Allen & Unwin, pp. 357–85.


——(1984), ‘The Success of Purchasing Power Parity: Historical Evidence and Its Implications for Macroeconomics’,in M.D.Bordo and A.J.Schwartz (eds), A Retrospective on the Gold Standard, NBER Conference Report, Chicago,University of Chicago Press, pp. 121–50.

Mitchell, B.R. (1975), European Historical Statistics, 1750–1970, London, Macmillan.Mundell, R.A. (1960), ‘The Monetary Dynamics of International Adjustment under Fixed and Flexible Exchange

Rates’, Quarterly Journal of Economics, 74, 227–57.——(1961), ‘The International Disequilibrium System’, Kyklos, 14, 153–70.——(1963), ‘Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rates’, Canadian Journal of

Economics and Political Science, 29, 475–85.——(1968), International Economics, New York, Macmillan.Mussa, M. (1982), ‘A Model of Exchange Rate Dynamics’, Journal of Political Economy, 90, 74–104.Neufeld, E.P. (1972), The Financial System in Canada, New York, St. Martin’s Press.Rich, Georg (1988), The Cross of Gold: Money and the Canadian Business Cycle, 1867–1913, Carleton Library 153,

Ottawa, Carleton University Press.——(1989), ‘Canadian Bank and the Gold Crisis of 1907’, Explorations in Economic History, 26, 135–60.Sargent, T.J. (1973), ‘Interest Rates and Prices in the Long Run’, Journal of Money, Credit and Banking, 5, 385–449.Triffin, R. (1964), The Evolution of the International Monetary System: Historical Reappraisal and Future

Perspectives, Princeton Studies in International Finance 12, Princeton, Department of Economics, PrincetonUniversity.

Urquhart, M. (1986), ‘New Estimates of Gross National Product, 1870–1926: Some Implications for CanadianDevelopment’, in S.L.Engerman and R.E.Gallman (eds), Long-Term Factors in American Economic Growth,NBER Studies in Income and Wealth 51, Chicago, University of Chicago Press, pp. 9–88.

Viner, J. (1924), Canada’s Balance of International Indebtedness: 1900–1913, Cambridge, Mass., Harvard UniversityPress.

Williamson, J.G. (1961), ‘International Trade and US Economic Development 1827–1843’, Journal of EconomicHistory, 21, 372–83.

——(1964), American Growth and the Balance of Payments, 1820–1913, Chapel Hill, University of North CarolinaPress.


6The gold standard as a rule*

Michael D.Bordo and Finn E.Kydland†

Introduction

The gold standard has long been viewed as a form of constraint over monetary policy actions—as a form ofmonetary rule. The Currency School in England in the early nineteenth century made the case for the Bankof England’s fiduciary note issue to vary automatically with the level of the Bank’s gold reserve (‘thecurrency principle’). Following such a rule was viewed as preferable (for providing price-level stability) toallowing the note issue to be altered at the discretion of the well-meaning and possibly well-informeddirectors of the Bank (the position taken by the opposing Banking School).1

In this paper we survey the history of the gold standard (or to be more accurate the specie standard)treated as a rule, but our meaning of the concept of a rule differs radically from what used to be thetraditional one. We regard it as a way of binding policy actions over time. This view of policy rules, incontrast to the earlier tradition that stressed both impersonality and automaticity, stems from the recentliterature on the time inconsistency of optimal government policy.

In the simplest sense, government policy is said to be time inconsistent when a policy plan, calculated asoptimal based on the government’s objectives and expected to hold indefinitely into the future, is subsequentlyrevised. Discretion, in this context, means setting policy sequentially. This could then lead to policies andoutcomes that are very different from the optimal plan, as market agents rationally incorporate governmentactions into their planning. For that reason, the government would benefit from having access to a commitmentmechanism to keep it from changing planned future policy.

Our approach to gold-standard history posits that adherence to the fixed price of specie, whichcharacterized all convertible metallic regimes including the gold standard, served as a credible commitmentmechanism (or a rule) to monetary and fiscal policies that otherwise would be time inconsistent. On this basis,adherence to the specie standard rule enabled many countries to avoid the problems of high inflation andstagflation that troubled the late twentieth century.

* For helpful comments and suggestions on earlier drafts of this paper we thank Charles Calomiris, Barry Eichengreen,Marvin Goodfriend, Lars Jonung, Leslie Pressnell, Hugh Rockoff, Anna Schwartz, Guido Tabellini, Warren Weber, andparticipants at seminars at Carnegie-Mellon University, the Federal Reserve Bank of Richmond, Columbia University,Queens University, Carleton University, the NBER Macroeconomic History Conference, June 1989, and the NBERSummer Institute 1991. For valuable research assistance we thank Mary Ann Pastuch and Bernhard Eschweiler.

† From Explorations in Economic History, 1995, pp. 423–30, 445–64, abridged. Copyright © 1995 Academic Press,Inc.

Furthermore, we argue that the gold standard that prevailed before 1914 was a contingent rule. Under therule, gold convertibility could be suspended in the event of a well-understood, exogenously producedemergency, such as a war, on the understanding that after the emergency had safely passed convertibilitywould be restored at the original parity. Market agents would regard successful adherence as evidence of acredible commitment and would allow the authorities access to seigniorage and bond finance at favorableterms.

According to this view the core countries of the gold standard—Britain, France, the United States, andGermany, as well as a number of other Western European countries and British Dominions—adheredstrictly to convertibility rules in the period from 1880 to 1914. By contrast many peripheral countries did not.Some never adhered to the rule. Others joined it when conditions were favorable to them, ostensibly toobtain access to capital from the core countries, but they quickly abandoned it when economic conditionsdeteriorated.

Before 1880, most countries were on a form of specie standard: either bimetallism or silver or goldmonometallism. However, from our perspective the bimetallic standards that many countries followed werea variant of the gold-standard rule, since it is convertibility that defines the rule.

The interwar gold standard can be regarded as an extension of the pre-1914 system because it was basedon gold convertibility. However, it was less successful because the commitment to convertibility was oftensubordinated to other politically induced objectives.

The Bretton Woods international monetary system can be regarded as a distant relative of the classical goldstandard in that the center country, the United States, maintained gold convertibility. It was also based on acontingent rule—parities could be changed in the event of a fundamental disequilibrium. However, itdiffered from the basic specie standard rule in that a credible commitment to the fixed parity was not ofsuch primary importance.

This essay is intended to provide a new perspective for examining many of the familiar issues of gold-standard history. As such, it is hoped that it will stimulate new research on different countries’ gold-standard experience to either verify or reject our approach.

The gold standard as a contingent rule

The essence of the gold standard was that each country would define the price of gold in terms of itscurrency and keep the price fixed. This meant defining a gold coin as a fixed weight of gold called, forexample, 1 dollar. The dollar in 1792 was defined as 24.75 grains of gold with 480 grains to the ounce,equivalent to $19.39 per ounce. The monetary authority then was committed to keep the mint price of goldfixed through the purchase and sale of gold in unlimited amounts. It was willing to convert into coin goldbullion brought to it by the public, to charge a certain fee for the service—called brassage—and also to sellcoins freely to the public in any amount and allow the public to convert them into bullion or export them.2

This operational procedure applies to a pure gold-coin standard. In fact, the standard that prevailed in thenineteenth century was a mixed standard containing both fiduciary money and gold coins. Under the mixedstandard, the gold standard required that fiduciary money (issued either by private banks or by thegovernment) be freely convertible into gold at the fixed price.

Most countries, until the third quarter of the nineteenth century, maintained bimetallic systems using bothgold and silver at a fixed ratio. Defining the weight of both gold and silver coins, freely buying and sellingthem, and maintaining the ratio fixed can be viewed as a variant of the basic gold standard, since it is a fixedvalue of the unit of account that is its essence.3

THE GOLD STANDARD AS A RULE 71

This paper views the operation of the gold standard from a new perspective. It is seen as intimatelyrelated to the public-finance question: How should the government finance fluctuating expenditures so as tominimize deadweight loss to society? In a general sense, this question dates back to the pioneering work ofRamsey (1927). In the context of a timeless environment, an important result of his work was that it isbeneficial to tax relatively more heavily what is inelastically supplied.

Over the past two decades, economists more and more have come to realize that an intertemporalframework is essential for addressing such questions. Seen in this light, the principle just enunciated isprecisely what may lead governments into trouble. If one can tax an asset (or the income therefrom) that hadbeen accumulated for the purpose of yielding services or command of resources in the future, then the time-inconsistency literature shows that there will be a strong temptation to increase these taxes ex post. While theyimpose little distortion in the short run, in the sense that most of the accumulated capital will still besupplied and somebody will hold the government debt that already has been issued, the problem is theadverse effect on incentives for future accumulation. As a consequence, to reassure the public that the realvalue of the debt will not erode in the future as a result of inflationary policy, there is a need for acommitment mechanism constraining future monetary policy.

The main point of this paper is that the gold standard represented such a commitment mechanism.Moreover, we contend that nations were committed to the gold standard as a rule with war as a contingency.Operationally, the rule is to suspend the gold standard for the duration of a war plus a delay period. In allperiods with no such emergency, the gold standard is maintained unconditionally. This policy is fullyunderstood and anticipated by the public.

To understand the role of the gold standard as a commitment mechanism, we use as a framework anenvironment in which the government behaves so as to maximize an intertemporal, but unchanging,objective function.4 There are two basic categories of taxes. One consists of taxes on already accumulatedassets, such as productive capital or holdings of government bonds. For the other revenue source, incontrast, intertemporal linkages are relatively unimportant. Examples are customs duties or a proportionaltax on labor income. Households and firms decide on consumption, investment, and labor input, while thegovernment chooses tax rates and net borrowing, including the printing of high-powered money (forexample, greenbacks during the Civil War).

A government plan that maximizes the value of its objective function is called a Ramsey plan, and thecorresponding sequence of private outcomes the Ramsey allocation. The Ramsey plan takes into account theprivate-economy behavior associated with alternative plans, assuming everyone believes them to be crediblefor the entire future. In particular, decisions to accumulate assets depend critically on the tax and inflationpolicies expected to be followed. In general, private decisions today depend on the entire sequence ofpresent and future government policy.

The role of government debt is especially important if the changes in government expenditures are largeat times, such as during wars. Without the issue of new debt, the required changes in tax rates wouldseverely reduce the incentives for economic activity at a time when the need for maintaining such activity isthe greatest. Issuing debt permits smoothing of tax rates over time. Such practice generally is beneficial notonly during wars, but also in normal circumstances, as shown in Barro (1979) and Kydland and Prescott(1980). The equality between expenditures and tax revenues needs to hold only in a present-value sense, notin every time period. Net borrowing makes up the difference.

To understand the time-consistency problem, consider a government plan that is optimal (i.e., a Ramseyplan) starting at time zero. Imagine now that the analogous problem is contemplated as of a future date s.The solution to this optimal-taxation problem is different from the part of the original plan for the now-current and future periods, s, s+1, and so on. In other words, the original plan is inconsistent with the


passage of time. The reason is that while the Ramsey plan computed at time zero takes into account theeffects of government policy, for example, capital income taxation, planned for date s and subsequentperiods on private behavior at dates before s, at time s, when the new plan is computed, private behaviorprior to period s, of course, can no longer be affected.

Two points about time inconsistency should be emphasized. One is that its source is not the conflictbetween the form of utility function for the government and that of private individuals. Moreover, timeinconsistency arises despite an unchanging objective function over time. The key factor is that decisions aremade sequentially over time in environments in which future government policy affects current privatebehavior.

When issuing debt, the interest rate and therefore the price at which the government can sell its debtdepend on the expectation of future inflation. In effect, the government can partially default ex post througha surprise inflation. The time-consistency problem is that if, in a future period, the government coulddefault to a greater extent than specified by the original plan, then the need for distortionary taxes isreduced. But such an action also affects expectations of subsequent defaults and therefore the price at whichthe public is willing to hold future government debt.5

For some model environments, the Ramsey plan, along with the associated private-economy allocations,represents an equilibrium outcome.6 What supports it as part of an equilibrium is the belief that as long asthe government has followed this plan in the past, it will continue to do so.7 This observation suggests thatan explicit and transparent mechanism is needed to make it work.

To ensure the presence of a credible commitment mechanism, society in some cases has committed itselfby law. An example is the patent law. It ensures sufficient incentives for inventive activity by givingexclusive use of new inventions for a period of time without fear that the government will remove thepatent right and allow the price of the product to be driven toward the competitive price. Our thesis is that,although the gold standard is easier to change than, for example, the patent law, this institutionalarrangement was set up with a similar motivation, namely as an explicit, transparent, well-understood rule.

In an uncertain world, the Ramsey plan generally is contingent. For any model environment, it includesas many contingencies as there are variables describing the state of the economy at any point in time.Drawbacks of including many contingencies, however, are lack of transparency and possible uncertaintyamong the public regarding the will to obey the original plan. Thus, a credible rule may include only thecontingency that is considered most important—a war in the case of the gold standard. A candidate for asecond contingency, to go into effect in the event of a financial crisis, assuming the crisis has not beencaused by the monetary authorities’ own actions, may be to use temporary restrictions on convertibility ofbank liabilities to reduce the extent of a banking panic.

How was the gold-standard rule enforced? A possible explanation focuses on reputational considerationswithin each country. Long-run adherence to the rule was based on the historical evolution of the goldstandard itself. Gold was accepted as money because of its intrinsic value and desirable properties such asdurability, storability, divisibility, portability, and uniformity. Paper claims, developed to economize on thescarce resources tied up in commodity money, became acceptable only because they were convertible intogold.8

In turn, the reputation of the gold standard would constrain the monetary authorities from breachingconvertibility, except under well-understood contingencies. Thus, when an emergency occurred, theabandonment of the standard would be viewed by all to be a temporary event because from theirexperience, only gold or gold-backed claims truly served as money.

An alternative commitment mechanism, analogous to the case of patents, would have been to guaranteegold convertibility in the constitution. This was done in Sweden before 1914, where laws pertaining to the


gold standard could be changed only by two identical parliamentary decisions with an election in between(Jonung, 1984, p. 368).

To reiterate, under the contingent gold-standard rule, the sovereign maintains the standard—keeps fixedthe price of its currency in terms of gold—except in the event of a major war, in which circumstance it cansuspend specie payments and issue paper money to finance its expenditure, and it can sell debt issues interms of the nominal value of its currency on the understanding that the debt eventually will be paid off ingold. The rule is contingent in the sense that the public understands that the suspension will last only for theduration of the wartime emergency plus some period of adjustment; it assumes that, afterward, thegovernment will follow the deflationary policies necessary to return to the gold standard.9

With respect to outright suspension of convertibility, it is difficult to distinguish between a suspension aspart of the operation of a contingent rule as mentioned above, or as evidence of discretion. By discretion, wemean any purposeful deviation, under whatever guise, from the rule. The excuse could be a ‘bad outcome’,in the language of Grossman and Van Huyck (1988), which is not included as a contingency in the originalplan. Deviations from the rule (perhaps accompanied by promises not to repeat the breach of the rule) aretempting because of the immediate benefits made possible because the public had anticipated continuationof the original plan. Statements by the monetary authorities, debates in Parliament, frequency of suspension,and changes in expectations as reflected in people’s decisions all can be used to distinguish between the two.

An example of discretion less extreme than a repeal is to decide at the end of the delay period to postponefurther its resumption, perhaps as a result of the perceived current situation in terms of how much of the warhas been paid for and the undesirable effects of the taxes needed to pay back the debt. This change is all themore tempting if the public had accepted the debt at a reasonably low interest rate in the expectation that thegold standard would be resumed as scheduled. If the government breaks the rule in effect by choosing ahigh default rate, then people’s behavior, should there be another war within memory of the previous one, willbe quite different from that in the previous war, even if the situation is otherwise similar and thegovernment claims to subscribe to the same fixed-delay rule.

An international rule

The gold-standard rule also has an international dimension. Under the rule, there would be no restriction onthe nationality of individuals who presented bullion to the mint to be coined or who exported coin or bullionto foreign countries. Moreover, because every country following the rule fixed the price of its currency ingold, this created a system of fixed exchange rates, linking all countries on the same standard. Theinternational aspect of the gold standard may have been particularly important to the countries that wererelatively less developed and therefore depended on access to international debt markets. The thesis of thispaper, however, is that the essence of the gold standard rule was as a domestic commitment mechanism. Tothe extent that the commitment was honored in relation to other countries, it served to strengthen thecredibility of the domestic commitment.10

The enforcement of the international gold standard seems to have taken a particular form that was conduciveto making it credible. A key factor may have been the role of England—the leading financial andcommercial center of the gold-standard era. The financial institutions of London provided the world with awell-defined and universally accepted means, based on gold, of executing bilateral trades and obtainingcredit. As we shall argue later, the gold standard provided England with the necessary benefits to enforce itand for many other countries to follow England’s lead. Exchange in both goods and capital was facilitated ifcountries adhered to a standard based on a rule anchored by the same commitment mechanism. This


arrangement may also have contributed to making the commitment mechanism a transparent one, acondition that we think is important for its success.

Evidence for the gold standard as a rule

In this section we survey the recent literature for evidence on the credibility of the commitment to goldconvertibility and of the contingent aspect of the gold standard rule. We focus on three separate lines ofinquiry: the gold standard as a credible commitment mechanism for core countries, commitment to the goldstandard as facilitating access to capital by peripheral countries, and the gold standard as a contingent rule inwartime.

The gold standard as a credible commitment mechanism for core countries

A number of recent research efforts have used statistical methods to reveal the credibility of the commitmentby the monetary authorities of the core countries to preserving gold convertibility at fixed parities during theperiod 1880–1914. These studies include evidence on the persistence of inflation, on the credibility bands ofshort-term interest rates within the gold points, on the reaction functions of gold-standard central banks, andon the durability of the commitment to gold convertibility based on speculative attack models.

Inflation persistence

Barsky (1987) presents evidence for the UK and the United States that inflation under the gold standard wasvery nearly a white-noise process. This is compared to the post-World War II period, when the inflation rateexhibited considerable persistence. Evidence for the absence of inflation persistence does not prove thatcountries followed the gold-standard rule. It is, however, consistent with the suggestion that market agentsexpect that the monetary authorities will not continuously follow an inflationary policy—an expectationthat is also consistent with belief in following a convertibility rule.

Bordo and Kydland (1995) extended Barsky’s approach using annual wholesale prices for three corecountries: the UK 1730–1938, the United States 1793–1933, and France 1803–1938, and one peripheralcountry—Italy 1861–1913. Their results that inflation in all four countries was very nearly white-noiseconfirm those of Barsky. Moreover the results hold for different subperiods when the countries concernedfollowed the bimetallic variant of the rule and for subperiods when they departed from convertibilityfollowing the contingent aspect of the rule. As did Klein (1975) and Barsky (1987), they observe negativeserial correlation at a large number of lags in all the subperiods.11 This is consistent with the commoditymoney adjustment mechanism of the gold standard discussed by Rockoff (1984) and Barsky and Summers(1988). They also tested for a unit root in the inflation series using the Dickey-Fuller test. In only oneepisode—the United States, 1862–78—could one be detected at the 10 per cent significance level.

In sum, they interpret this evidence as consistent with agents’ beliefs in the credibility of the commitmentto the gold-standard convertibility rule. However because the power of the tests is admittedly low, thisevidence is suggestive only.

More recently, Alogoskoufis and Smith (1991) show, based on AR(1) regressions of the inflation rate,that inflation persistence in the United States and UK increased between the classical gold-standard periodand the interwar period and between the interwar period and the post-World War II period.12 Table 6.1presents the inflation rate coefficient from the type of AR(1) regressions on annual CPI inflation estimatedby Alogoskoufis and Smith, for the 21 countries. The regressions, as well as the standard errors and the


Dickey-Fuller tests for a unit root,13 are run on four successive monetary regimes since 1880: the classicalgold standard 1880–1914, the interwar period 1919–39, Bretton Woods 1946–70, and the recent managedfloat 1974–90.

The results, as in Alogoskoufis and Smith, show an increase in inflation persistence for most countries(both core and peripheral) between the classical gold standard and the interwar period. This evidence isconsistent with the view that the credibility of commitment to the gold-standard rule was strong under theclassical gold standard but weakened after World War I. For 12 countries, most notably the UK, Germany,and Japan,

Table 6.1 Persistence of CPI inflation: 21 countries, 1880-1990

AR1 S.E. t AR1 S.E. AR1 S.E. AR1 S.E. t

1 Core Countries

United States United Kingdom Germany France

Gold standard 0.27 (0.18) 4.05 0.30 (0.17) 4.03 0.51 (0.16) 3.06 −0.22 (0.18) 6.78

Interwar 0.45 (0.17) 3.18 0.35 (0.19) 3.37 0.51 (0.21) 2.33 0.42 (0.24) 2.42

Bretton Woods 0.49 (0.19) 2.68 0.33 (0.20) 3.38 −0.03 (0.21) 4.90 0.56 (0.16) 2.75

Floating exchange 0.68 (0.18) 1.76 0.69 (0.19) 1.67 0.83 (0.14) 1.21 0.85 (0.16) 0.94

2 British dominions

Australia Canada

Gold standard 0.39 (0.19) 3.19 0.08 (0.18) 5.11

Interwar 0.23 (0.19) 4.13 0.35 (0.20) 3.25

Bretton Woods 0.60 (0.19) 2.09 0.39 (0.19) 3.21

Floating exchange 0.60 (0.18) 2.27 0.75 (0.17) 1.47

3 Latin America

Argentina Brazil Chile

Gold Standard 0.29 (0.30) 2.37 0.47 (0.17) 3.11 0.23 (0.31) 2.48

Interwar −0.17 (0.19) 6.14 0.53 (0.19) 2.53 0.16 (0.23) 3.68

Bretton Woods 0.27 (0.20) 3.61 0.81 (0.12) 1.58 0.56 (0.16) 2.66

Floating exchange 0.74 (0.22) 1.18 1.29 (0.08) −3.62 0.85 (0.14) 1.11

4 Southern Europe

Greece Italy Portugal Spain

Gold standard na — 0.28 (0.14) 5.14 0.14 (0.19) 4.54 −0.22 (0.11) 10.60

Interwar 0.18 (0.24) 3.37 0.28 (0.18) 4.00 0.63 (0.36) 1.02 0.14 (0.24) 3.63

Bretton Woods 0.72 (0.26) 1.08 0.21 (0.12) 6.58 −0.13 (0.17) 6.77 0.13 (0.23) 3.75


5 Scandinavia

Denmark Finland Norway Sweden

Gold standard 0.27 (0.17) 4.21 0.46 (0.17) 3.08 0.44 (0.17) 3.23 0.42 (0.17) 3.53

Interwar 0.24 (0.25) 3.04 0.64 (0.25) 1.45 0.31 (0.31) 2.21 0.50 (0.11) 4.44

Bretton Woods −0.12 (0.16) 7.20 0.30 (0.30) 2.37 0.10 (0.20) 4.42 0.13 (0.21) 4.26



AR1 S.E. t AR1 S.E. t AR1 S.E. t AR1 S.E. t

6 Other Western Europe

Belgium Netherlands Switzerland

Goldstandard

0.11 (0.18) 4.83 −0.36 (0.26) 5.14 na

Interwar

0.49 (0.20) 2.57 0.34 (0.16) 4.14 0.16 (0.15) 5.77

BrettonWoods

−0.09 (0.22) 5.07 0.31 (0.15) 4.58 0.24 (0.20) 3.75

Floatingexchange

0.78 (0.16) 1.40 0.88 (0.11) 1.05 0.69 (0.18) 1.80

7 Japan

Goldstandard

0.22 (0.18) 4.33

Interwar

0.70 (0.25) 1.20

BrettonWoods

0.52 (0.18) 2.67

Floatingexchange

0.70 (0.19) 1.58

Source: See Data Appendix to Bordo and Schwartz (1994).Note. Annual data: coefficient of AR1 regression (standard error), and t statistic for unit root test. Five per cent

significance level for unit root test with 25 observations is 3.00.

persistence declined under Bretton Woods. This is consistent with a restoration of credibility upon thereturn to some link with gold convertibility. However, the fact that it rose in nine countries, most notablythe United States and France, suggests that credibility under Bretton Woods was weaker than under the goldstandard (evidence consistent with Giovannini’s (1993) findings). Finally for virtually all countries inflationpersistence increases between Bretton Woods and the float. This suggests that whatever the commitment tolow inflation that remained under Bretton Woods evaporated upon cutting the final link to goldconvertibility.

Credibility bands

Under the classical gold standard, the official parity rate was bounded by the gold points, upper and lowerlimits reflecting the arbitrage (transactions, transportation, interest, and risk) costs of shipping gold betweeninternational centers in the face of balance of payments disequilibria. Evidence by Officer (1986) for theLondon-New York exchange rate on demand bills showed that, although exchange rates frequently departedfrom par, violations of the gold points were rare—making the case for the efficiency of the gold standard.14


More recently Giovannini (1993) has calculated credibility bands (bands within which uncovered interestarbitrage prevails consistent with gold point arbitrage efficiency). For sterling, the mark, and the franc in theperiod 1899–1909, these bands lie within the limits set by the gold points—evidence which he interprets asconsistent with market agents’ expectations of a credible commitment by the core countries to the gold-standard rule in the sense of this paper. For the United States, credibility bands pierce the gold export pointin 1893 and 1895–6 at the height of agitation over silver. Officer (1993) made similar calculations for theinterwar dollar sterling exchange rate, finding serious violations only in 1931, at the very end of the goldexchange standard.

The durability of the commitment to gold convertibility based on speculative attackmodels

The recent literature on speculative attacks on fixed exchange rates (Flood and Garber, 1983) has beenapplied to the gold standard. According to that approach any fixed exchange rate regime will collapse oncethe fundamentals of money supply relative to money demand predict an exchange rate that departs from thefixed exchange rate consistent with purchasing power parity.

Based on this approach, Grilli (1990) calculated the probability of a speculative attack on the gold dollarat the height of the agitation over silver in 1893 (before the repeal of the Sherman Silver Purchase Act) asnot much greater than 6 per cent. Calomiris (1993) disputes these calculations. Using the interestdifferential between short-term interest rates in New York and London as an estimate of the expected rate ofexchange rate depreciation, he finds no evidence of significant short-term depreciation risk in the periodNovember 1893 to late 1895. Only in June 1893 is there a slight risk of a temporary suspension of goldconvertibility. This evidence buttresses the case for the United States as a credible core gold-standardadherent.

Reaction functions

A substantial literature in the 1980s tested whether core country central banks violated the rules of the gamein the sense that the discount rate and other policy tools responded to domestic economic objectives at theexpense of maintaining convertibility (see Bordo, 1986). Eschweiler and Bordo (1994) estimated a reactionfunction for the Reichsbank using monthly data over the period 1883–1913. They found that the centralbank’s pursuit of an interest rate smoothing policy (an obvious violation of the rules of the game) wassubordinate to its commitment to keep the exchange rate within the gold points. This evidence as well as theearlier evidence by Dutton (1984), Pippenger (1984), and Giovannini (1986) of significant violation of therules in the short term, with paramount importance placed on convertibility in the long run, is consistentwith the view that the gold points represented a form of target zone within which central banks could, overshort horizons, pursue domestic objectives without threat to convertibility (Svensson, 1994).

Access to foreign capital as a commitment mechanism by peripheral countries

One of the enforcement mechanisms of the specie standard rule for peripheral countries was presumablyaccess to the capital needed for their economic development from the core countries. Adherence to theconvertibility rule would be viewed by lenders as evidence of financial probity—i.e., membership in theinternational gold standard would be like ‘the good housekeeping’ seal of approval. It would signal that acountry 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 considerablelengths to avoid defaulting on externally held debt. It would also presumably be a signal to the lenders inLondon and other metropolitan areas that the groups in power observed similar standards of financialrectitude.

This suggests that adherence to the specie standard rule, ceteris paribus, would make a difference in thevolume 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 abrogationof the gold clauses or of total default on the debt. That eventuality would be reflected in a risk premium onthe loan. In that case it would be attractive to a potential borrower to adhere to the specie standard rule as asignal of financial responsibility to induce the lender to lower the risk premium. But a more fundamentalproblem 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 borrowerswilling to engage in unduly risky projects. Lenders faced with imperfect information on the borrowers’likely actions would then be reluctant to lend at any price. A credible commitment to the specie standardrule, as evidenced by the holding of substantial gold reserves, would provide a signal to lenders of the costsborrowers would be willing to bear to avoid default and hence would circumvent the aversion to lendingimposed by asymmetric information.

There is considerable evidence that access to capital was indeed a key determinant of peripheralcountries’ attempts to maintain convertibility. This was the case both for developing countries seekingaccess 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 1905–6 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). Bycontrast, Spain, according to Martin Acena (1993), suffered a significant loss in potential output because itsfailure to join the gold standard precluded it from access to foreign capital at favorable terms. Finally, thefact that England, the most successful country of the nineteenth century, and ‘progressive’ countries wereon the gold standard, was probably a powerful argument for joining (Friedman, 1990; Gallarotti, 1993).

Bordo and Schwartz (1994) provide limited evidence based on a descriptive analysis of Davis andHuttenback’s (1986) capital calls on new issues on the London Stock Exchange—a measure of access tocapital—for the United States and Argentina in the period 1865 to 1914. Their narrative suggests thatadherence to the rule by Argentina may have had some marginal influence on capital calls on new issues ofsecurities in London before 1890, but that the key determinant was the opening up of the country’s vastresources to economic development once unification and a modicum of political stability were achieved.The 1890 Baring crisis was a major shock to investor confidence and it took years of austerity, therestoration of convertibility, and the establishment of a currency board before British investors’ confidencewas restored.

For the United States, events suggesting the restoration of convertibility during the suspension episode(the Resumption Act of 1875) and threats to convertibility during adherence (the 1893 Silver Crisis) areassociated with increases and decines in capital calls on new issues of securities in London.

For Italy, Fratianni and Spinelli (1984) provide evidence for the credibility of the commitment to the goldstandard in the risk premium on Italian government long-term securities relative to their French counter—parts over the period 1866 to 1912. When Italy adhered to gold, from 1884– 94, it averaged close to zero.After suspending convertibility in 1894 it rose to 2 per cent, but at the turn of the century and a return tomonetary and fiscal conservatism (Toniolo, 1990, p. 188), with the monetary authorities acting as if theywere on the gold standard, the premium fell to 0.5 per cent. In a similar vein, Della Paolera (1993) shows


that the spread between Argentine government bonds and both British consols and the Rendita Italianabegan a steep decline, from its post Baring crisis peak in 1894, to low levels once orthodox financialpolicies were restored and the inten tion of restoration was made known—well before convertibility wasactually restored in 1899.

Finally, a case study of Canada during the Great Depression provides evidence for the importance of thecredible commitment mechanism of adherence to gold. Canada suspended the gold standard in 1929 but didnot allow the Canadian dollar to depreciate nor the price level to rise for two years. Canada did not takeadvantage of the suspension to emerge from the depression because of concern for its credibility withforeign lenders (Bordo and Redish, 1990).

The gold standard as a contingent rule in wartime

A key aspect of the gold-standard rule was the contingency in the event of a well-understood emergency.Britain, France, and the United States all took advantage of it during wartime. They suspended specieconvertibility, issued fiat currency, and sold government bonds denominated in fiat currency. They alsorestored parity once the emergency passed.

An important question that arises from these experiences is whether during the suspension periods marketagents believed that they were only temporary, i.e., whether the commitment to restore convertibility at theoriginal parity was credible. We survey three pieces of evidence for the credibility of the contingent rule:the behavior of interest rates and exchange rates during suspension periods in the United States and Britain,the pattern of shares of finance in government expenditures during wartime periods, and evidence onrevenue smoothing.

The behavior of interest rates and exchange rates during suspension periods

Calomiris (1988 and 1993), following a study by Roll (1972), presents evidence of expected appreciation ofthe greenback dollar well before resumption of specie payments occurred in January 1, 1879, based on anegative interest differential between bonds that were paid in greenbacks and those paid in gold. Calomiris(see Table 6.2) calculates the appreciation forecast error on a semi-annual basis from January 1869 toDecember 1878, defined as the difference between his calculation of expected appreciation and actualappreciation. The errors are close to zero for most of the periods, with three exceptions: January to June1869, when the error is 0.43, January to June 1871, when the error is −1.00, and January to June 1876, whenit is −1.49. The first positive exchange rate surprise reflects the credibility of the government’s commitmentto the redemption of bond principal in gold; the second followed a change in treasury policy which


Table 6.2 Expected and actual appreciation of the greenback dollar, 1869– 1878

(1) (2) (3) (4) (5)

Averagedifferentialbetween gold andgreenbacks yielda

Expectedappreciation(currentdifferential lessdifferential forJuly-December1878)

Average actualrate ofgreenbacksappreciation to1881b

Appreciationforecast error (2)-(3)

Appreciationforecast errorallowing time-varying riskpremiumc

January-June1869

1.33 3.53 2.00 1.53 0.43

July-December1869

0.49 2.69 1.85 0.84 −0.26

January-June1870

−0.52 1.68 0.93 0.75 −0.35

July-December1870

−0.42 1.78 0.93 0.85 −0.15

January-June1871

−1.01 1.19 1.09 0.10 −1.00

July-December1871

−0.95 1.25 1.10 0.15 −0.95

January-June1872

−0.02 2.18 1.26 0.92 -0.18

July-December1872

0.01 2.21 1.40 0.81 −0.29

January-June1873

−0.09 2.11 1.90 0.21 −0.89

July-December1873

−0.26 1.94 1.39 0.55 −0.55

January-June1874

−0.65 1.55 1.60 −0.05 −0.05

July-December1874

−0.45 1.75 1.50 0.25 0.25

January-June1875

0.07 2.27 2.36 −0.09 −0.09

July-December1875

0.09 2.29 2.30 −0.01 −0.01

January-June1876

−1.19 1.01 2.50 −1.49 −1.49

July-December1876

−1.07 1.13 1.76 −0.63 −0.63

January-June1877

−1.22 0.98 1.36 −0.38 −0.38

July-December1877

−1.21 0.99 0.84 0.15 0.15

January-June1878

−1.32 0.88 0.40 0.48 0.48


(1) (2) (3) (4) (5)

Averagedifferentialbetween gold andgreenbacks yielda

Expectedappreciation(currentdifferential lessdifferential forJuly-December1878)

Average actualrate ofgreenbacksappreciation to1881b

Appreciationforecast error (2)-(3)

Appreciationforecast errorallowing time-varying riskpremiumc

July-December1878

−2.20 0.00 0.10 0.10 0.10

Source: Calomiris (1988, Table 5, and 1993, Table 4.1). Reprinted with the permission of Oxford University Press andCambridge University Press.

Notes:a

b The average of monthly exchange rate closings for the period was used to measure current gold price of greenbacks.The 6s of 1881 were redeemable June 1, 1881.

c This calculation sets the risk premium equal to 1.10 for 1869–73, and 2.20 for 1874–78.

was perceived as a signal of a possible significant increase in the supply of greenbacks; the third negativesurprise reflects the temporary threat to resumption by the election of 1876.15

In the case of Britain’s return to gold in 1929, Smith and Smith (1990) and Miller and Sutherland (1992,1994), using models from the stochastic process-switching literature, find that the increasing likelihood thatresumption would occur at the original parity gradually altered the path of the dollar-pound exchange ratetoward the new ceiling several months in advance. Their models differ, however: Smith and Smith’s is statedependent—driven by an expectation of the restoration of purchasing power parity—while Miller andSutherland’s is time dependent—driven by a strong belief that resumption would occur by expiration of theGold and Silver (Export Control) Act at the end of 1925.

The pattern of shares of finance in government expenditure during wartime

According to the gold-standard contingent rule, a temporary suspension of convertibility allows thegovernment to follow a feasible fiscal policy in which it smooths revenue by financing its expenditures witha combination of taxes, bond finance, and seigniorage. This is predicated on the assumption of a crediblecommitment to restore convertibility once the emergency has passed.

In a case study comparing British and French finances during the Napoleonic Wars, Bordo and White(1993) show that Britain was able to finance its wartime expenditures by a combination of taxes, debt, andpaper money issue—to smooth revenue, because after previous wars in the eighteenth century governmentdebt was successfully serviced. By contrast, France had to rely primarily on taxation. France had to rely ona less efficient mix of finance than Britain because she had used up her credibility by defaulting on outstandingdebt at the end of the American Revolutionary War and by hyperinflating during the Revolution. Napoleonultimately returned France to the bimetallic standard in 1803 as part of a policy to restore fiscal probity, butbecause of the previous loss of reputation France was unable to take advantage of the contingent aspect ofthe bimetallic standard rule.

One implication of the theory is that after the emergency has passed and successful resumption and debtretirement has been achieved, access to bond finance on favorable terms would be possible in the event ofanother war. Indeed, this may have been the case for the British where it appears that the successful


resumption of 1821 may have been a factor enabling them to finance an even larger share of World War Iexpenditures by debt finance and the issue of fiat money than during the Napoleonic Wars. See Table 6.3.

Revenue smoothing

The theory of tax smoothing implies that an optimizing government will set tax rates over time so as tominimize deadweight losses (Barro, 1979). The

Table 6.3 The financing of wartime expenditures in the French wars and World War I

Per cent of total wartime expenditures financed by:

A. French wars 1793–1815 (GB) B. World War I 1914–1918 (UK)a

(1) Taxes 58.0 31.8

(2) Bonds 40.5 64.4

(3) High-powered money 1.5 3.8

Sources: (1) 1793–1815: O’Brien (1967), table 4; 1914–1918: Mitchell and Deane (1962), pp. 392–5, 396–8. (2) 1793–1815: O’Brien (1967) table 4; 1914–18: Mitchell and Deane (1962), ibid. (3) 1793–1815: Mitchell andDeane (1962), pp. 441–3; 1914–18: Capie and Webber (1985), table 1 (1), pp. 52–9.

Note:a Wartime expenditures are calculated as total government expenditures less 1903–1913 annual average of total

government expenditures.

theory also predicts that in an uncertain world, tax rates should follow a martingale process. An extension ofthis theory to the case of two fiscal instruments—taxes and seigniorage (the theory of revenue smoothing—implies that both should follow a martingale process and furthermore that a regression of the inflation rateon the average tax rate should have a positive and significant coefficient (as Mankiw, 1987; Poterba andRotemberg, 1990; and Trehan and Walsh, 1990, found for the post-World War I United States).16

Bordo and White (1993) tested the theories of tax smoothing and revenue smoothing for Great Britainduring the Napoleonic War suspension period. See Table 6.4. The evidence does not reject the hypothesis oftax smoothing but it does reject the hypothesis of revenue smoothing. Similar results are found in a study byGoff and Toma (1993) for the United States under the classical gold standard and by Lazaretou (1995) forGreece during periods of inconvertibility in the nineteenth century. As Goff and Toma (1993) argue,seigniorage smoothing would not be expected to prevail under a specie standard where the inflation ratedoes not exhibit persistence (as was the case during the 1797–1815 period and during the Greekinconvertibility episodes). The Bordo and White and Lazaretou results suggest that, although speciepayments were suspended, the commitment to resume prevented the government from acting as it wouldunder the pure fiat regime postulated by the theory.

Table 6.4 A. Tax smoothing in Great Britain 1700–1815

Coefficients of independent variables

Variable Period B0 B1 R2 SEE D-W

T/Y=B0+B1 · (T/Y)t−1 +et

Taxes/commodityoutput

1715–1815 0.876(0.918)

0.978(25.94)a

0.87 2.81 1.84


Coefficients of independent variables

Variable Period B0 B1 R2 SEE D-W

Taxes/nationalincome

1700–1815 1.69(1.46)

0.947(22.39)a

0.81 3.16 1.85

� T/Y=B0+B1� (T/Y)t−1+et

Taxes/commodityoutput

1715–1815 0.312(1.09)

0.049(0.47)

0.002 2.82 1.92

Taxes/nationalincome

1700–1815 0.29(0.98)

0.017(0.177)

−0.008 3.20 1.19

Note. t values in parentheses.at value not significantly different from 1 at the 5 per cent level of significance.

(1) � log Pt=−1.81 −0.0008 (T/Y)+0.001 (time)

(−0.12) (1.27) (0.14)

R2=0.016 SEE=0.119 D−W= 1.60 p=0.33

(2) � log pt=−7.77−0.007 (T/CO)+0.004 (time)

(−0.40) (−1.27) (0.42)

R2=0.004 SEE=0.120 D−W=1.56 p=0.35

(3) � 2 log pt=0.007−0.01 A (T/Y)

(0.207) (−1.87)

R2=0.121 SEE=0.137 D−W= 1.77

(4) � 2 log pt=0.01–0.01 A (T/CO)

(0.31) (−2.03)*

R2=0.148 SEE=0.135 D−W= 1.74

Source: Bordo and White (1993). Reprinted with the permission of Cambridge University Press.Notes. (T/Y) represents tax revenues divided by national income. (T/CO) represents tax revenues

divided by commodity output.* Signifies statistically significant at the 5 per cent level.

Suggestions for future research

Although theoretical research in the past decade in the literature on time inconsistency has mushroomed(Persson and Tabellini, 1990), historical research is still in its infancy. This essay has explored the historicalimplication of just one aspect of this subject, the use of the gold (specie)-standard rule as a form ofcommitment mechanism. Historical research on other aspects of time-inconsistent policies, especially fiscalpolicy, is also on the rise (see, e.g., Calomiris, 1993 and Motomura, 1994).

Within the framework of convertibility rules as commitment mechanisms we can offer some suggestionsfor future research. The first is to devise a test which can empirically distinguish a time-consistentgoldstandard policy from a time-inconsistent policy that switches on and off a rigid gold standard wheresuspensions are not permitted.

The second is to examine the convertibility history of diverse countries in more detail than the briefoverview we present earlier. It would be of great interest to know why particular countries adapted andabandoned convertibility rules when they did. Detailed case studies such as those of Reis (1991), MartinAcena (1993), Fritsch and Franco (1992), LlonaRodriquez (1992), and Lazaretou (1995) could be done forother countries. Of interest in this regard would be an exegesis of official views, statements from


parliamentary debates, newspaper editorials, and the like, to ascertain the tone of feeling for theconvertibility rule during periods of suspension. Finally, an analysis of central bank laws and constitutionsmay reveal explicit statements of the contingencies under which the rule could be suspended.

The third suggestion is to extend the type of empirical analysis pioneered by Roll (1972) and Calomiris(1988), using available data on interest rates and exchange rates for diverse periods of suspension ofconvertibility, to ascertain market agents’ expectations of their permanence. One possible approach is to usea pooled time-series cross-country sample of a number of countries’ experiences. Related to this would bemore effort devoted to measuring risk premia on long-term securities during periods of convertibility andsuspension for both core and peripheral countries. Can a significantly different pattern be detected?

A fourth suggestion would be serious studies on long-term capital flows from the core to the periphery. Didadherence to gold convertibility make a significant contribution in addition to that of the fundamentaldeterminants of capital movements (such as expected real rates of return, levels of real activity, the terms oftrade, and the phase of the business cycle) as isolated in earlier studies by Ford (1962) and Edelstein (1982).Alternatively, were risk premia significantly related to specie standard adherence?

Finally more research would be of value in testing with historical data the target zone approach to peggedexchange rates. Although one early investigation comparing the implications of Krugman’s (1991) modelfor the gold standard, Bretton Woods, and the EMS (Flood et al, 1991) found only limited support, thestudies by Giovannini (1993) and Officer (1993) deriving credibility bonds suggest the opposite. Empirical/historical extensions of Svensson’s (1994) analysis of the compatibility of stabilization policy andconvertibility rules under a commitment regime, as in Eschweiler and Bordo (1994), may also provefruitful, as would theoretical and historical application of the target zone methodology to the contingentaspect of the gold-standard rule.

The lessons from history

The history of the gold standard suggests that the gold (specie) convertibility rule was followedcontinuously by only a few key countries—the best example being England from 1821 to 1914. Most majorcountries, however, did follow the rule during the heyday of the classical gold standard, 1880–1914.Peripheral countries and several fairly important nations—Italy and Argentina—alternately followed andthen departed from the rule, but even they were constrained in a looser sense.

The gold-standard rule also proved to be successful as a commitment mechanism for England, the UnitedStates, and France in preventing default on debt and ensuring that paper money issues during periods ofwartime suspension were not permanent. It may have been successful as a commitment device because ithad the virtues of being simple and transparent.

We have suggested a number of reasons why the gold-standard rule was so successful as a commitmentmechanism before 1914. First, as a contingent rule it permitted nations to have access to revenue in times ofwartime emergency. The commitment to return to gold parity after the war would enable the authorities toissue debt and to collect seigniorage at more favorable terms than otherwise.17

Second, in England and possibly in other countries, gold emerged early on as a way of certifyingcontracts. This certifying characteristic of gold carried forward to the relationship between the private andpublic sectors. Abandoning gold convertibility was viewed as a serious breach of contract. The goldstandard emerged in the stable political environment of England after the seventeenth century, where therule of law sanctified private contracts.18 Only a few countries had comparable stability. Countries fraughtwith more unstable internal politics found it more difficult to refrain from running budget deficits,


ultimately financed by paper money issue (for example, Italy and Argentina), although the benefits ofconvertibility placed some constraints on their behavior.19

The gold standard was also successful as an international rule: by pegging their currencies to gold,countries became part of a fixed exchange rate system. The international aspect of the gold standard mayhave reinforced the domestic commitment mechanism because of the perceived advantages of morefavorable access to international capital markets, by the operation of the rules of the game, by theimportance of England as a hegemonic power, and by international cooperation between central banks.

The advantages accruing to England as the center of the gold-standard world—the use of sterling as areserve asset and the location in London of the world’s key asset and commodity markets—made the costsof not following the gold-standard rule (except in wartime emergency) extremely high. Furthermore,because England was the most important country in the gold-standard era and access to the London capitalmarket was considered to be of great benefit to developing countries, it is likely that many countries adheredto the gold standard that otherwise would not have, given the high resource costs of maintaining goldreserves. Also, because of the Bank of England’s leadership role, other central banks may have beenprevented from using discretionary policies, threatening adherence to the rule.

A comparison of the pre-1914 period with the subsequent period is of great interest. The gold exchangestandard, which prevailed for only a few years from the mid-1920s to the Great Depression, was an attemptto restore the desirable features of the classical gold standard while allowing a greater role for domesticstabilization policy.

As an application of the contingent rule it was much less successful. Because monetary policy was highlypoliticized in many countries, the commitment to credibility was not believed and devaluation would haveled to destabilizing capital flows. Unlike the prewar gold standard, central bank cooperation wasineffective. The system collapsed in 1931 and subsequent years in the face of the shocks of the GreatDepression.20

Bretton Woods was our last convertible regime. It can be viewed within the context of the specie standardrule, although it is a distant variant of the original specie standard. The architects of Bretton Woods wantedto combine the flexibility and freedom for policy makers of a floating rate system with the nominal stabilityof the gold-standard rule. Under the rules of Bretton Woods, only the United States, as central reservecountry and provider of the nominal anchor, was required to peg its currency to gold; the other memberswere required to peg their currencies to the dollar. They were also encouraged to use domestic stabilizationpolicy to offset temporary disturbances.

The Bretton Woods system had a contingency for its members—a change in parity was allowed in theface of a fundamental disequilibrium, which could encompass the contingencies under the specie standardrule—but it was not the same as under the specie standard because it did not require restoring the originalparity. The rule for members (other than the United States) was enforced, as under the gold standard, byaccess to US capital and to the IMF’s resources. For the United States, there was no explicit enforcementmechanism other than reputation and the commitment to gold convertibility. Capital controls were viewedas a method to contain market pressures.

Ultimately the system was successful as long as the United States—the nominal anchor to the system—maintained its commitment to convertibility, i.e., maintained price stability. As events turned out, byfollowing highly expansionary monetary and fiscal policies to finance the Vietnam War beginning in themid-1960s, the United States attached greater importance to domestic concerns than to its role as the centerof the international monetary system, thus weakening the system and contributing to its collapse.21

While the gold-standard rule was widely upheld before 1914, it has not been since, although, as webriefly document, to a lesser extent both the short-lived gold exchange standard and the Bretton Woods


system incorporated a number of its features. The breakdown in the rule may reflect a number of importantunderlying factors including the rise in democracy and especially the move to universal suffrage in a numberof countries—this in turn implied a decline in the power of groups in society who benefited from adherenceto the gold standard (Gallarotti, 1993; Frieden, 1994) and an increase in the influence of previouslydisenfranchised groups who may have suffered in the face of external shocks brought on by adherence tothe rule; a growing understanding of the use of the tools of monetary policy to shield the domestic economyfrom external disturbances (Eichengreen, 1992b); the breakdown in overall political stability (including theimportant role played by England) which in turn fostered the high degree of mobility of capital, labor, andcommodities so complementary to the smooth operation of the gold standard.

Today, one could characterize most nations as following a discretionary standard, although rhetoric overthe importance of rules abounds. This may seem surprising, since the benefits of having a commitmentmechanism seem more relevant today than 100 years ago. On the other hand, there may have been theperception that government debt was, and is, less important as an emergency source of funds than it was inthe gold-standard era. For example, the stocks of physical and human capital have risen substantially. Thetime-inconsistency literature has taught us that the incomes therefrom have broadened the scope for policymakers to use discretionary policy. For example, marginal tax rates for people with above-average humancapital rose dramatically during and after World War II. In the absence of a commitment mechanism, theserates were not returned to prewar levels.

The gold-standard rule was simple, transparent, and, for close to a century, successful. Even though itwas characterized by some defects from the perspective of macroeconomic performance, a bettercommitment mechanism has not been adopted. Despite its appeal, many of the conditions that made thegold standard so successful vanished in 1914, and the importance that nations attach to immediateobjectives casts doubt on its eventual restoration.

Notes

1 For a discussion of the Currency-Banking School debate, see Viner (1937), Fetter (1965), and Schwartz (1987).2 Strictly speaking, the government needed to define a gold coin only in terms of the unit of account. Private mints

could then supply the demand for coin. Indeed, this was the case shortly after the California gold discoveries(Bancroft, 1890, p. 165). In most countries, however, the mint was under government authority.

3 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 cushionof a switch from one metal to the other. See Friedman (1990b). Garber (1986) regards bimetallism as a goldstandard with an option.

4 If everyone were alike, a natural objective for the government would be to maximize the expected present valueof a representative household’s utility subject to the government budget constraint.

5 In Prescott’s model (1977), for example, the government finances a given stream of expenditures either throughtaxes on labor income (abstracting from capital) or by selling debt. He finds that if the government has nocommitment mechanism for future actions, the government will always default on outstanding debt to avoidlevying distorting taxes. As a consequence, the equilibrium implies that the value of government debt is zero andthat the government always runs a balanced budget. This policy and the implied allocation are, of course, inferiorto the Ramsey allocation for that model.

6 Chari and Kehoe (1989, 1990) consider this issue for environments in which time-consistency problems ariseeither because of capital taxation or because of the presence of government debt. A well-written overviewappears in Chari (1988).


7 The idea that reputation may support optimal policy has been explored in a different context by Barro andGordon (1983).

8 Goodfriend (1989) describes how the evolution of contractual arrangements in the financial system in eighteenth-and nineteenth-century England had to overcome the problem of fraud. Private markets developed an elaboratesystem of monitoring financial arrangements, but ultimately convertibility into gold lay behind them.

9 This description is consistent with a result from a model of Lucas and Stokey (1983), in which financing of warsis a contingency rule that is optimal. In their example, where the occurrence and duration of the war are uncertain,the optimal plan is for the debt not to be serviced during a war. Under this policy, people realize when they purchasethe debt that effectively it will be defaulted on in the event the war continues.

10 The role of spillover effects on reputation through multiple relationships is discussed in Cole and Kehoe (1991).11 Klein (1975) also presents evidence for mean reversion of the price level under the gold standard.12 Also see Alogoskoufis (1992), who attributes the increase in persistence to the accommodation by the monetary

authorities of shocks. This evidence is also consistent with the results of Klein (1975).13 The regression run was � log Pt=B0+B1� log Pt−1+� t. We ran the same regression for the GNP deflators, with

similar results.14 Officer’s calculations effectively overturned those in earlier studies by Clark (1984) and Morgenstern (1959).15 Also, see Smith and Smith (1993) who demonstrate, using a stochastic process-switching model, that movements

in the premium on gold from the Resumption Act of 1875 until resumption was established in 1878 were drivenby a credible belief that resumption would occur.

16 However, according to Chari et al. (1991) this result applies strictly to partial equilibrium models that assume aconstant rate of return on debt and a loss function for the government which depends directly on the tax ratesrather than the allocations. Further assumptions are required for it to hold in a general equilibrium model.

17 Grossman’s (1990) interpretation of the historical record, though emphasizing different factors, accepts thisview. Thus, according to him, the ratio of government debt to gross national product increased during majorwartime episodes in Britain and the United States from the mid-18th century until after World War I, reflectingintertemporal substitution. Such borrowing represented a temporary effort to shift resources from the future to thepresent. Following each war, the ratio of debt to income would then be reduced by contractionary fiscal policyaccompanied by deflationary monetary policies that maintained the real rate of return on outstanding bonds.According to Grossman, such a policy was an investment in the credibility capital of the sovereign borrower—areputation for responsible repayment of the principal and for preservation of the real value of interest paymentsthat enhanced the probability of being able to borrow heavily again at favorable rates in the event of a future war.

18 According to North and Weingast (1989), this process was complete by the Glorious Revolution of 1688. Afterthat date, capital markets developed in an environment free of the risk of sovereign appropriation of capital.

19 An alternative and complementary explanation to that offered in this paper relates to political economyconsiderations and the distribution of income. The configuration of political interest groups in the 19th centurywas favorable to the hard-money, pro-gold-standard-rule position. This may have been related to the more limiteddevelopment of democracy and less-than-universal suffrage. Thus, a comparison of the debates over resumptionin England from 1797 to 1821 and in the United States from 1865 to 1878 suggests that the more limited suffragein England in the early period served as a brake on the soft-money forces favoring permanent depreciation. In theUnited States, the soft-money forces favoring redistribution of income to debtors and other groups (such asMidwestern manufacturers) almost carried the day.

20 As is well known, the gold exchange standard suffered from a number of fatal flaws (Kindleberger, 1973;Eichengreen, 1992b; Temin, 1989). These include the use of two reserve currencies, the absence of leadership byhegemonic power, the failure of cooperation between the key members, and the unwillingness of its twostrongest members, the United States and France, to follow the rules of the game, instead exerting deflationarypressure on the rest of the world by persistent sterilization of balance-of-payment surpluses.

21 For a detailed discussion of the collapse of Bretton Woods see Bordo (1993) and Garber (1993).


References

Akerlof, G. (1970), ‘The Market for Lemons: Quality Uncertainty and the Market Mechanism’, Quarterly Journal ofEconomics, 84 (Aug), 488–500.

Alogoskoufis, G.S. (1992), ‘Monetary Accomodation, Exchange Rate Regimes and Inflation Persistence’, EconomicJournal, 2 (May), 461–80.

Alogoskoufis, G.S. and R.Smith (1991), ‘The Phillips Curve, the Persistence of Inflation and the Lucas Critique:Evidence from Exchange-Rate Regimes’, American Economic Review, 81(2), 1254–73.

Bancroft, H.H. (1890), History of California, Vol. VII, San Francisco, History.Barro, R.J. (1979), ‘On the Determination of the Public Debt’, Journal of Political Economy, 87, 940–71.Barro, R.J. and D.B.Gordon (1983), ‘Rules, Discretion and Reputation in a Model of Monetary Policy’, Journal of

Monetary Economics, 12, 101–21.Barsky, R.B. (1987), ‘The Fisher Hypothesis and the Forecastability and Persistence of Inflation’, Journal of Monetary

Economics, 19(1), 3–24.Barsky, R.B. and L.H.Summers (1988), ‘Gibson’s Paradox and the Gold Standard’, Journal of Political Economy, 96

(3), 528–50.Bloomfield, A. (1959), Monetary Policy Under the International Gold Standard, 1880–1914, New York, Federal

Reserve Bank of New York.Bordo, M.D. (1986), ‘Money, Deflation and Seigniorage: A Review Essay’, Journal of Monetary Economics, 18(3),

337–46.——and F.E.Kydland (1995), ‘The Gold Standard as a Commitment Mechanism’, in T.Bayoumi, B.Eichengreen and

M.Taylor (eds), Economic Perspectives on the Gold Standard, Cambridge, Cambridge University Press.——and A.Redish (1990), ‘Credible Commitment and Exchange Rate Stability: Canada’s Interwar Experience’,

Canadian Journal of Economics, 23(2), 357–80.——and Schwartz, A.J. (1994), ‘The Specie Standard as a Contingent Rule: Some Evidence For Core and Peripheral

Countries, 1880–1990’, NBER Working paper 4860, September.——and E.N.White (1993), ‘British and French Finance During the Napoleonic Wars’, in M.D.Bordo and F.Capie

(eds), Monetary Regimes in Transition, Cambridge, Cambridge University Press.Calomiris, C.W. (1988), ‘Price and Exchange Rate Determination During the Greenback Suspension’, Oxford Economic

Papers (Dec).——(1993), ‘Greenback Resumption and Silver Risk: The Economics and Politics of Monetary Regime Change in the

United States, 1862–1900’, in M.D.Bordo and F.Capie (eds), Monetary Regimes in Transition, Cambridge,Cambridge University Press.

Chari, V.V., L.J.Christiano and P.J.Kehoe (1991), ‘Optimal Fiscal and Monetary Policy: Some Recent Results’, Journalof Money, Credit and Banking, 23(2), 519–39.

——and P.J.Kehoe (1989), ‘Sustainable Plans and Debt’, Federal Reserve Bank of Minneapolis Staff Report 125.Clark, T. (1984), ‘Violations of the Gold Points, 1890–1908’, Journal of Political Economy, 92, 791–823.Cole, H.L. and P.J.Kehoe (1991), ‘Reputation With Multiple Relationships: Reviving Reputation Models of Debt’,

Federal Reserve Bank of Minneapolis Staff Report 137.Davis, L.E. and R.A.Huttenback (1986), Mammon and the Pursuit of Empire: The Political Economy of British

Imperialism, 1860–1912, Cambridge, Cambridge University Press.Della Paolera, G. (1993), ‘Monetary and Banking Experiments in Argentina: 1861– 1930’, Mimeo, Universidad

Torcuato Di Tella Buenos Aires.Dutton, J. (1984), ‘The Bank of England and the Rules of the Game Under the International Gold Standard: New

Evidence’, in M.D.Bordo and A.J.Schwartz (eds), A Retrospective on the Classical Gold Standard, 1821–1931,Chicago, University of Chicago Press.

Edelstein, M. (1982), Overseas Investment in the Age of High Imperialism: The United Kingdom, 1850–1914, NewYork, Columbia University Press.


Eichengreen, B. (1992a), ‘The Gold Standard Since Alec Ford’, in S.N.Broadberry and N.F.R.Crats (eds), Britain in theInternational Economy: 1870–1939, Cambridge, Cambridge University Press.

——(1992b), Golden Fetters: The Gold Standard and the Great Depression, 1919–1939, Oxford, Oxford UniversityPress.

Eschweiler, B. and M.D.Bordo (1994), ‘Rules, Discretion and Central Bank Independence: The German Experience1880–1989’, in Pierre Siklos (ed.), Varieties of Monetary Reform: Lesson and Experiences on the Road toMonetary Union, Boston, Kluwer Academic.

Fishlow, A. (1987) ‘Market Forces or Group Interests: Inconvertible Currency in Pre-1914 Latin America’, Mimeo,University of California at Berkeley.

——(1989), ‘Conditionality and Willingness to Pay: Some Parallels from the 1989s’, in B.Eichengreen and P.Lindert(eds), The International Debt Crisis in Historical Perspective, Cambridge, Mass., MIT Press.

Flood, R.P. and P.M.Garber (1983), ‘A Model of Stochastic Process Switching’, Econometrica, 51(3), 537–51.Flood, R.P., D.J.Mathieson and A.K.Rose (1991), ‘An Empirical Exploration of Exchange Rate Target Zones’,

Carnegie Rochester Conference Series on Public Policy, 35(Autumn), 7–66.Ford, A.G. (1962), The Gold Standard 1880–1914: Britain and Argentina, Oxford, Clarendon Press.Fratianni, M. and F.Spinelli (1984), ‘Italy in the Gold Standard Period, 1861– 1914’, in M.D.Bordo and A.J.Schwartz

(eds), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press.Frieden, J.A. (1994), ‘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 theInternational System, Colorado, Westview Press.

Friedman, M. (1990), ‘Bimetallism Revisited’, Journal of Economic Perspectives, 4(4), 85–104.Fritsch, W. and G.H.B.Franco (1992), ‘Aspects of the Brazilian Experience Under the Gold Standard’, Mimeo, PUC

Rio de Janeiro.Gallarotti, G.M. (1993), The Scramble for Gold: Monetary Regime Transformation in the 1870s’, in M.D.Bordo and

F.Capie (eds), Monetary Regimes in Transition, Cambridge, Cambridge University Press.Garber, P. (1986), ‘Nominal Contracts in a Bimetallic Standard’, American Economic Review, 76, 1012–30.——(1992), ‘The Collapse of the Bretton Woods Fixed Exchange Rate System’, in M.D. Bordo and B.Eichengreen

(eds), A Retrospective on the Bretton Woods System, Chicago, University of Chicago Press.Giovannini, A. (1986), “‘Rules of the Game” during the International Gold Standard: England and Germany’, Journal of

International Money and Finance, 5, 467–83.——(1993), ‘Bretton Woods and Its Precursors: Rules Versus Discretion in the History of International Monetary

Regimes’, in M.D.Bordo and B.Eichengreen (eds), A Retrospective on the Bretton Woods System, Chicago,University of Chicago Press.

Goff, B.L. and Toma, M. (1993), ‘Optimal Seigniorage, The Gold Standard, and Central Banking Financing’, Journalof Money, Credit and Banking, 25(1), 79– 95.

Goodfriend, M. (1989), ‘Money, Credit, Banking and Payments System Policy’, in D.B.Humphrey (ed.), The U.S.Payments System: Efficiency, Risk, and the Role of the Federal Reserve, Boston, Kluwer Academic.

Grilli, V. (1990), ‘Managing Exchange Rate Crises: Evidence from the 1890s’, Journal of International Money andFinance, 9, 258–75.

Grossman, H.I. and J.B.Van Huyck (1988), ‘Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation,and Reputation’, American Economic Review, 78, 1088–97.

Grossman, H.I. (1990), ‘The Political Economy of War Debts and Inflation’, in W. S.Haraf and P.Cagan (eds),Monetary Policy for a Changing Financial Environment, Washington, DC, American Enterprise Institute.

Hawtrey, R. (1926), Monetary Reconstruction, London, Longman.Hayashi, F. (1989), ‘Japan’s Saving Rate: New Data and Reflections’, NBER Working Paper 3205.Jonung, L. (1984), ‘Swedish Experience Under the Classical Gold Standard, 1873– 1914’, in M.D.Bordo and

A.J.Schwartz (eds), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of ChicagoPress.

Kindleberger, C.P. (1973), The World in Depression, 1929–1939, Berkeley: University of California Press.


Klein, B. (1975), ‘Our New Monetary Standard: Measurement and Effects of Price Uncertainty, 1880–1973’, EconomicInquiry 13, 461–84.

Krugman, P. (1991), ‘Target Zones and Exchange Rate Dynamics’, Quarterly Journal of Economics, 56, 669–82.Kydland, F.E. and E.C.Prescott (1977), ‘Rules Rather than Discretion: The Inconsistency of Optimal Plans’, Journal of

Political Economy, 85, 473–91.——(1980), ‘A Competitive Theory of Fluctuations and the Feasibility and Desirability of Stabilization Policy’, in

S.Fischer (ed.), Rational Expectations and Economic Policy, Chicago, University of Chicago Press for theNational Bureau of Economic Research.

Lazaretou, S. (1995), ‘Government Spending, Monetary Policies and Exchange Rate Regime Switches: The Drachma inthe Gold Standard Period’, Explorations in Economic History, 32, 28–50.

Lindert, P. (1969), Key Currencies and Gold, 1900–1913, Princeton Studies in International Finance 24, Princeton,Princeton University Press.

Llona-Rodriguez, A. (1992), ‘Chile During the Gold Standard: A Succesful Paper Money Experience’, Mimeo, InstitutoTorcuato Di Tella.

Lucas, R. E, Jr. and N.L.Stokey (1983), ‘Optimal Fiscal and Monetary Policy in an Economy Without Capital’, Journalof Monetary Economics, 12, 55–93.

Mankiw, G.N. (1987), ‘The Optimal Collection of Seigniorage: Theory and Evidence’, Journal of MonetaryEconomics, 20, 327–42.

Martin, Acena, P. (1993), ‘Spain During the Classical Gold Standard Years, 1880– 1914’, in M.D.Bordo and F.Capie(eds), Monetary Regimes in Transition, Cambridge, Cambridge University Press.

Miller, M. and A.Sutherland (1992), ‘Britain’s Return to Gold and Entry into the ERM’, in P.Krugman and M.Miller(eds), Exchange Rate Targets and Currency Banks, Cambridge, Cambridge University Press.

——(1994), ‘Speculative Anticipations of Sterling’s Return to Gold: Was Keynes Wrong?’ Economic Journal, 104(July), 804–12.

Mitchell, B.R. and P.Deane (1962), Abstract of British Historical Statistics, Cambridge, Cambridge University Press.Morgenstern, O. (1959), International Financial Transactions and Business Cycles, Princeton, Princeton University Press.Motomura, A. (1994), ‘The Best and Worst of Currencies: Seigniorage and Currency Policy in Spain, 1597–1650’,

Journal of Economic History, 54(1), 104–27.North, D. and B.Weingast (1989), ‘Constitution and Commitment: Evolution of Institutions Governing Public Choice’,

Journal of Economic History, 49(4), 803–37.Officer, L. (1986), ‘The Efficiency of the Dollar-Sterling Gold Standard, 1890– 1908’, Journal of Political Economy, 94

(Oct), 1038–73.——(1993), ‘Gold-Point Arbitrage and Uncovered Interest Arbitrage Under the 1925–1931 Dollar-Sterling Gold

Standard’, Explorations in Economic History, 30(1), 98–127.Persson, T. and G.Tabellini (1990), Macroeconomic Policy, Credibility and Politics, Chur, Harwood Press.Pippenger, J. (1984), ‘Bank of England Operations, 1893–1913, in M.D.Bordo and A.J.Schwartz (eds), A Retrospective

on the Classical Gold Standard, 1821– 1931, Chicago, University of Chicago Press.Poterba, J.M. and J.Rotemberg (1990), ‘Inflation and Taxation with Optimizing Government’, Journal of Money, Credit

and Banking, 22, 1–28.Prescott, E. (1977), ‘Should Control Theory Be Used for Economic Stabilization?’ Carnegie-Rochester Conference

Series on Public Policy, 7, 13–38.Ramsey, F. (1927), ‘A Contribution to the Theory of Taxation’, Economic Journal, 37, 47–61.Reis, J. (1991), ‘The Gold Standard in Portugal, 1854–1891’, Mimeo, Universidade Nova de Lisboa.Rockoff, H. (1984), ‘Some Evidence on the Real Price of Gold, Its Cost of Production, and Commodity Prices’, in

M.D.Bordo and A.J.Schwartz (eds), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago,University of Chicago Press.

Roll, R. (1972), ‘Interest Rates and Price Expectations During the Civil War’, Journal of Economic History, 32,478–98.


Schwartz, A.J. (1987) ‘Banking School, Currency School, Free Banking School’, in New Palgrave Dictionary ofEconomics, London, Macmillan.

Sharkey, R.D. (1959), Money, Class, and Party, Baltimore, Johns Hopkins University Press.Smith, G. and T.Smith (1993), ‘Wesley Mitchell and Irving Fisher and the Greenback Gold Reforms 1865–1879’,

Mimeo, Queens University.Smith, W.S. and R.T.Smith (1990), ‘Stochastic Process Switching and the Return to Gold’, Economic Journal, l00

(March), 164–75.Stiglitz, J. and A.Weiss (1981), ‘Credit Rationing in Markets with Imperfect Information’, American Economic Review,

71(6), 393–410.Svennson, L.E.O. (1994), ‘Why Exchange Rate Bands? Monetary Independence in Spite of Fixed Exchange Rates’,

Journal of Monetary Economics, 33(1), 157– 99.Temin, P. (1989), Lessons from the Great Depression, Cambridge, Mass., MIT Press.Toniolo, G. (1990), An Economic History of Liberal Italy: 1850–1918, New York, Routledge.Trehan, B. and C.E.Walsh (1990), ‘Seigniorage and Tax Smoothing in the United States: 1914–1986’, Journal of

Monetary Economics, 25, 97–112.Unger, I. (1964), The Greenback Era: A Social and Political History of American Finance, 1865–1879, Princeton,

Princeton University Press.Viner, J. (1937), Studies in the Theory of International Trade, Chicago: University of Chicago Press.Yeager, L.B. (1984), ‘The Image of the Gold Standard’, in M.D.Bordo and A.J. Schwartz (eds), A Retrospective on the

Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press.


Part II

The gold standard in history

Introduction

The gold standard’s evolution and operation differed in practice from the predictions of economists’theoretical models. The selection from the work of John Maynard Keynes suggests that central banksactually possessed significant policy autonomy under the gold standard owing to the presence of thefluctuation bands which were demarcated by the gold import and export points. Robert Triffin describesother aspects in which the gold standard’s operation diverged from the predictions of standard models,emphasizing asymmetries and arguing that the system operated to the benefit of Europe but not theperiphery. A.G.Ford documents Triffin’s point, contrasting the gold standard’s operation in Britain with itseffects in Argentina. Barry Eichengreen synthesizes the work of Triffin and Ford, offering evidence fromBritish experience which supports their interpretations of the gold standard. Finally, Jeffry Frieden providesa political-economy analysis of the emergence of the gold standard, emphasizing domestic political factors.

7The significance of the gold points

John Maynard Keynes*

We have seen that, if rates of exchange were absolutely fixed, so that it cost nothing to exchange the moneyof one country into the money of another, exactly the same rate of interest would always prevail in bothcountries for loans of the same type and believed to offer the same security. If such conditions were toprevail between all countries, each change occurring anywhere in the conditions of borrowing and lendingwould be reflected by a change in the Bank-rate and Bond-rate everywhere. Every puff of wind, that is tosay, would travel round the world without resistance. If its force were spread over a large area, it would, ofcourse, be less than if it were boxed up within a small area. If, on the other hand, most countries were toerect wind-screens of greater or less effectiveness, then any country which remained exposed would, unlessit were large relatively to the rest of the world, be subjected to perpetual instability.

If, therefore, a country adopts an international standard, it is a question just how international it wishes tobe—just how sensitive to every international change. The device which we have now to consider isexpressly directed towards damping down this sensitiveness without departing from effective conformitywith an international standard.

A loan in terms of one currency is not identical with a loan in terms of another, even when both broadlyconform to the same international standard, unless the currencies are interchangeable without cost and at arate which is known for certain beforehand. If there is an element of expense or an element of doubt in theconditions of exchange of one currency for another, then the rate of interest on loans in terms of the first canfluctuate, within limits set by the amount of the cost and the degree of the doubt, independently of the rate ofinterest on loans in terms of the second. The possible range, between the terms on which one currency canbe exchanged for another and the terms on which the exchange can be reversed at a later date, is determinedby what is called, in the technical language of the foreign exchanges, the distance between the gold points.The greater the distance between the gold points, the less sensitive to short-period external changes acountry’s rate of foreign lending will be.

Thus the degree of separation of the gold points is a vital factor in the problem of managing a country’scurrency and ought to be the subject of very careful decision. It has not, however, been treated as suchhitherto, but has been governed by influences some of them historical and some of them purely fortuitous,though, doubtless, there has been over extended periods a sort of empirical survival of the fittest.

One of the most effective means of keeping short-term foreign lending insensitive is to allow an elementof doubt as to the future terms of exchange between currencies. This was the traditional method of the Bankof France for many decades before the [First World] war. Silver five-franc pieces remained legal tender, andthe Bank of France gave no guarantee that they would always exchange them for gold at their legal parity. It

* A Treatise on Money, Volume II: The Applied Theory of Money, London, Macmillan, 1930, pp. 319–31. Reproducedcourtesy of Macmillan Ltd.

was not necesssary for the efficacy of the method that the threat should be exercised often or at all, its mereexistence being enough to hinder the activities of the arbitrageur, who is essentially concerned withcertainties and narrow margins. In many other European countries besides France doubt used to exist, forone reason or another,1 as to whether gold would be freely obtainable for export in all circumstances; whilstin the United States the fact that there was no Central Bank imported an element of uncertainty into theatmosphere. Even since the post-war restorations of the Gold Standard there are several countries whichprotect themselves, in one way or another, from their obligation to redeem their notes, gold being of tooabsolute a character.

In Great Britain, however, this form of protection has never been used2 (apart from the war and post-warperiod of the suspension of gold convertibility); nor is it in use in the United States since the foundation ofthe Federal Reserve System. These countries depend, not on an element of doubt, but on an element of cost;though—it should be added—the element of cost is also present, as an additional protection, in all othercountries. In what follows I will take Great Britain as typical, the practices of most other countries being thesame in principle though quantitatively different.

The element of cost is made up of two factors, both of which are familiar enough. The first is thedifference between the Bank of England’s buying and selling prices for gold,3 i.e., the difference betweenthe rate at which the Bank will give notes for gold and the rate at which it will give gold for notes.Historically this difference was based on the convenience and economy of getting notes for gold at onceinstead of taking gold to the Mint and waiting for it to be coined; and before the war the effective differencewas in fact tied down to the actual measure of this convenience, since holders of gold had an option betweenselling it to the Bank at the Bank’s price or taking it to the Mint and waiting. The Currency Act of 1928,however, abolished this option, and the precise amount of the difference between the Bank’s buying andselling prices (namely the difference between £3:17: per oz. and £3:17:9, which amounts to 0.16 per cent) isnothing but an historical survival.4

The second factor represents the actual cost, in freight-charges, insurance and loss of interest, oftransporting gold from one place to another. The amount of this is variable, not only because somedestinations are nearer than others, but also in accordance with the varying rates of interest and insurancecharges and the time occupied by different means of transport. On balance, however, the amount of thissecond factor tends in modern conditions to be reduced.5 As between London and New York, the extremevariation between the best and worst terms on which dollars and sterling can be exchanged is about percent. As between London and Paris, on the other hand, the variation is obviously smaller.6 But in the case ofIndia I calculated (before the war) that the range was nearly per cent.7 Generally speaking, the maximumextent of the range for different pairs of countries varies from to per cent.

Whilst these are the extreme limits, the organization of the ‘forward exchanges’ usually enablesborrowers to make a better bargain than this as to the terms on which they will be able to exchange onecurrency for another three months ahead.8 But even in the most favourable circumstances some cost will beincurred or some risk will have to be ran, which an owner of one currency who wishes to effect a loan interms of another will have to take into consideration in calculating whether a given transaction offers aprofit; and, as the pressure of transactions in one direction drives the rate of exchange towards one of thegold points, the prospective cost of a further transaction in the same direction will probably increase.

Let us, for the sake of illustration, assume that the anticipated cost is per cent. Now in the case of a loanof long duration this does not appreciably affect the net rate of interest obtainable after allowing for theexchange of one currency into another. For example, with a loan for ten years, the factor only reduces the rateobtainable by per cent per annum. But in the case of a short-period loan the position is materially different.


For example, with a loan for three months, the cost of exchange would, on the above hypothesis, reduce therate obtainable by 2 per cent per annum.

Thus this factor enables a substantial inequality to exist between the rates of interest obtainable in twodifferent currencies respectively if the rate of exchange existing at the moment cannot be relied on to lastfor more than a short period. For example, if the dollar-sterling exchange stands at the gold-export point forsterling, an interest rate in London higher than the New York rate for comparable loans will attract lendingfrom New York to London, since loans can be remitted from New York to London with a certainty that theycan at any future date be brought home again without exchange loss; that is to say, the dollar-sterlingexchange can always be kept above gold-export point for sterling, and the flow of gold from London to NewYork thereby prevented, by keeping interest-rates in London higher than in New York. The same, of course,is also true in the opposite direction. But if, on the other hand, the rate of exchange lies somewhereintermediate between the gold points, then there is no necessity for interestrates to be equal in the twocentres, though there are still limits beyond which the inequality cannot go; if, for example, the gold pointsare per cent apart, the rate for three months’ loans in New York could conceivably be 3 per cent perannum higher in New York than in London with the exchange at the gold-export point for sterling, or—equally well—the three-month rate could be 3 per cent per annum higher in London than in New York withthe exchange at the gold-export point for dollars. The mathematical expectation, however, or probable costof reversing a remittance three months hence, will seldom, or never, equal its maximum possible value. Themarket’s estimation of this probability is given by the ‘forward exchange’ quotations, so that in equilibrium:

Three months’ interest at the London rate, plus (or minus) an allowance for the discount (or premium) onforward dollars

= Three months’ interest at the New York rate.Thus if there is a fair distance between the gold points, there is a fair margin of difference which can

exist between short-money rates in two centres—provided always that the money-market cannot rely on along continuance of this margin of difference. It is this distance, therefore, which protects the money-market of one country from being upset by every puff of wind which blows in the money-markets of othercountries.

It follows that the magnitude of the difference is a matter of great importance for the stability of acountry’s internal economy. One might have supposed, therefore, that it would have been fixed after carefulconsideration at a safe amount. But this has not been the case hitherto. Moreover, the amount of thedifference is liable to be upset by air-transport, or, for example, by a bank being willing to forgo interest ongold during transit.

I believe that there is room here for a reform of real importance.9 I suggest that the difference between aCentral Bank’s obligatory buying and selling prices of gold should be made somewhat greater than hitherto,say 2 per cent, so that there would be at least this difference between the gold points irrespective of theactual costs of transporting gold (double the amount of which would have to be added on to the 2 per cent togive the difference between the gold points). But a Central Bank would be free at any time, if it wished toencourage the movement of gold inwards or outwards, to quote closer prices within the legal limits. Further,a Central Bank should be in a position to control when necessary, within the limits set by the gold pointsand the relative rates of interest at home and abroad, the premium or discount of the forward exchange onthe spot exchange; whereby short-period interest-rates at home could stand temporarily in such relation(within limits) to similar rates abroad as the Central Bank might deem to be advisable.

The object of this reform is to enable a Central Bank to protect the credit structure of its own countryfrom the repercussions of purely temporary disturbances abroad, whilst the laws of long-period equilibriumwould remain as before. Let us give an example where the proposed arrangements would have been useful.

THE SIGNIFICANCE OF THE GOLD POINTS 97

In the autumn of 1928 local conditions in the United States convinced the Federal Reserve Board that theshort-period interest-rate should be raised in the interests of business stability; but local conditions in GreatBritain were of a precisely opposite character, and the Bank of England was anxious to keep money ascheap as possible. The Federal Reserve Board did not desire that its high rates should attract gold fromGreat Britain; for this, if it occurred, would have tended to defeat its efforts. Nor did the Bank of Englanddesire to impose high rates in Great Britain—to which it might be driven—in order to prevent its gold fromflowing out. Such a situation could be handled by the above plan. The Federal Reserve Banks would reducetheir buying price for gold to a figure nearer to their legal minimum, whilst the Bank of England wouldraise its selling price for gold nearer to its legal maximum.10 If the Central Banks had also the practice, assuggested above, of influencing the rates for forward exchange, then these rates too would have to bemoved correspondingly. This would permit temporarily the maintenance of materially different short-money rates in the two centres. It would not, of course, permit a permanent difference, since it is theexpectation or the possibility of fluctuations between the gold points within a short period which permits thedifference to exist. Thus a lasting divergence in interest-rates in favour of New York would cause the dollar-sterling exchange to reach the gold point corresponding to the Bank of England’s legal maximum sellingprice for gold and so cause gold to flow.

I would propose, therefore, to furnish Central Banks with a trident for the control of the rate of short-termforeign lending—their bank-rate, their forward-exchange rate, and their buying and selling rates for gold(within the limits of the lawful gold points). I conceive of them as fixing week by week, not only theirofficial rate of discount, but also the terms on which they are prepared to buy or sell forward exchange onone or two leading foreign centres and the terms on which they are prepared to buy or sell gold within thegold points. This would have the effect of putting Central Banks in the same short-period position as that inwhich they would be in if they were to feel themselves able to suffer larger fluctuations in their stock ofgold without inconvenience. The reader must also notice in particular that a Central Bank would be able, byfixing appropriately its forward-exchange rates relatively to the spot rates prevailing in the market, to fix ineffect different short-term rates of interest for foreign funds and for domestic funds respectively.

The anomalies of the present situation and the instinctive striving of Central Banks to widen the goldpoints are well illustrated by two events which occurred during 1929–30, whilst these pages were beingpassed through the press.

The reader will have perceived that the distance between the gold points is narrower for a pair of countrieswhich are geographical neighbours than for a pair which are more distant from one another. It follows thatthe gold points are particularly narrow as between Paris and London. Thus, unless the sterling rate is, e.g.,well above its parity on New York, the gold export point to Paris, when France is on the point of importinggold, will be reached sooner in London than in New York; so that the mere propinquity of London to Pariswill tend to throw on her more of the short-period burden of the French gold requirements than on NewYork. Under pressure of such circumstances, however, the Bank of England and the Bank of France found aroundabout way within the letter of the existing law for making the cost of transporting gold from NewYork to Paris more nearly equal to the cost of transporting it from London to Paris. The Bank of Englandexercised its statutory right to deliver only standard gold and the Bank of France exercised its statutory rightto accept only fine gold, which had the effect of widening the gold points by the addition of the expenses ofrefining and of the consequent delay; and at the same time ways were found of minimizing the loss ofinterest during the period occupied in transporting gold from New York to Paris. The result was to transferthe bulk of the gold exports from London which was an unwilling exporter, to New York which was awilling one, without the change in the relative short-money rates in the two centres, which would have beenrequired otherwise as a necessary accompaniment of the movement of the sterling-dollar exchange towards


the gold export point from New York to London. This is an excellent example of the technique the habitualuse of which I am advocating; but it is somewhat absurd that the possibility of employing it on the aboveoccasion should have depended on an accidental feature of the existing statutes in the two countriesconcerned.11

The other recent example is supplied by Canada. The gold points between Canada and the United Statesare—for geographical reasons as in the previous case—particularly narrow, with the result that goldmovements between the two countries are frequently brought about by transient causes. In September 1929,the combined effects of a slow movement of the Canadian wheat crop out of the country and of the highrates for short-term loans in Wall Street, led to a tendency for gold to leave Canada on a greater scale thanwas convenient. The position was dealt with by lowering the gold-export point in virtue of an informalagreement between the Minister of Finance and the ten Chartered Banks not to export gold for the sake of asmall arbitrage profit. The rate of exchange inevitably went outside the lower gold point presumed by theterms of Canada’s adherence to the gold standard; but the expedient was justified by the event, since thehigh rates for short-term loans in New York proved to be temporary, and by the summer of 1930, when theyhad evaporated, the Canadian exchange had not only recovered to parity but had gone beyond it to the gold-import point. Thus this informal widening of the gold points was probably the wisest way of preventing theabnormal position in Wall Street from reacting unduly on the domestic credit conditions in Canada.

The objections likely to be raised against the proposals of this section are two. It is likely to be urged thatthe extra latitude which it would allow might be abused and become the occasion and the excuse for aCentral Bank to omit to take measures to remedy what was not a passing phase but a cause of persistingdisequilibrium. It is true, of course, that every increase in the discretion allowed to a Central Bank, so as toincrease its power of intelligent management, is liable to abuse. But in this case the risk is slight. For theeffect on any element in the situation, other than the international short-loan position, would beinsignificant, and the effect on this would be strictly limited and incapable of cumulative repetition.

Or it may be urged that an expedient of this kind, whilst well enough for a country which is not adepository of part of the international short-loan fund, is against the interests of a financial centre whichaspires to be an important depository of such funds. Some force must be allowed to this. It is a question ofhow high a price in the shape of domestic instability it is worth while to pay in order to secure internationalbanking business. From the standpoint of the latter the ideal would be to narrow the gold points until they wereidentical. It is a matter of finding a just and advantageous compromise between the competing interests. Butthere are also reasons for not attaching much importance to this objection. For I have coupled with thisproposal a further suggestion by which the Central Bank would quote rates of forward exchange; and bythis means the foreign depositor can always be accorded, for three-monthly periods at a time, such degree ofsecurity and economy in the movement of his funds, backwards and forwards, as the Central Bank deems itsafe and advantageous to accord to him. That is to say, we should only be loosening the tightness of thelegal obligation laid on the Central Bank—we should not be preventing or hindering it from working thesystem, in practice and as a rule, exactly as at present.

Moreover, London could overcome the awkwardness due to geographical propinquity without weakeningher competitive position against New York as a depository of foreign funds, if she were to fix her new goldpoints on any third country no wider than New York’s.

The ideal system, however—and one which would entirely overcome the competitive argument—wouldbe an arrangement for uniform action by all the leading countries. It would be better to supplant the existinghaphazard and fluctuating gold points, different between every pair of countries and open to all kinds ofminor uncertainties, with a fixed and uniform system between every pair of countries. There is a way inwhich, in the ideal international currency system of the future, we could ensure this, namely, by every

THE SIGNIFICANCE OF THE GOLD POINTS 99

country substituting, for its Central Bank’s existing legal obligation to buy and sell its local money onprescribed terms in exchange for gold, an obligation to buy and sell its local money on prescribed terms inexchange for balances at a Supernational Bank. A reasonable measure of domestic autonomy could then beobtained by every country’s buying rate being fixed 1 per cent below the parity of its local money and itsselling rate 1 per cent above. This would not be incompatible with the maintenance of a gold standard,provided that balances at the Supernational Bank were encashable in gold—it would merely mean that therewould be no movements of gold between pairs of individual countries, but only between individual countriesand the Supernational Bank. We shall return in Chapter 38 to the contribution which such a Bank mightmake to the solution of the currency problems of the world.

At any rate, we may conclude, the precise magnitude of the stretch between the gold points deservesmore scientific consideration than it has yet received. The present system of an international gold standard,in combination with separate national systems of Central Banks and domestic money, would be unworkableif the gold points coincided (as they do, in practice, within a country); and if this is agreed, it follows thatthe degree of their separation should not be a matter of material costs of transport or historical survivals.

Notes

1 E.g., the possibility of the Central Bank meeting its legal obligations by paying out light-weight gold coins.2 The option to the Bank of England to pay in fine gold or in standard gold will be dealt with below.3 To which must be added, to get the extreme limits of fluctuation, the corresponding charge made by the other

country whose currency is being exchanged.4 In 1929 the Bank of England revived a practice, which had existed up to 1912, of paying more than its statutory

minimum price whenever it is specially anxious to secure gold.5 Some interesting calculations, bearing on this, have been published by Dr P. Einzig in articles published in the

Economic Journal, March 1927, September 1927 and December 1928. Dr Einzig calls attention to the narrowingof the gold points made possible by air transport. The most up-to-date calculations are to be found in Dr Einzig’sInternational Gold Movements, Appendix 1. For example, in 1913 the dollar-sterling gold points were $4.89 and$4.8509, or 0.81 per cent; they widened in 1925 to $4.8949 and $4.8491, or 0.96 per cent; and they narrowed in1928 to $4.8884 and $4.8515, or 0.76 per cent. If the rate of interest was to fall to 3 per cent (as against 5 percent assumed in the above) the difference between the gold points would narrow to 0.7 per cent.

6 The range between London and Amsterdam is about 0.8 per cent, between London and Berlin about 0.7 per cent,and between London and Paris about 0.5 per cent (reckoning interest at 5 per cent in each case). Cf. Einzig, op.cit.

7 For a detailed examination of this point in the particular case of India see my Indian Currency and Finance,Chapter 5.

8 I have described in detail the machinery of the ‘forward exchange’ market, and have analysed the factors whichdetermine the rates quoted, in my Tract on Monetary Reform, Chapter 3, § 4.

9 The proposal which follows is substantially the same in principle as that which I made in my Tract on MonetaryReform, pp. 189–91.

10 Owing to the absence of any such provisions, the Bank of England was, in the middle of 1929, resorting to theunsatisfactory expedient of putting moral pressure on banks and finance houses to forgo the small profit whichwould have been obtainable by exporting gold.

11 The technical details of this episode have been described by Dr Paul Einzig in the Economic Journal Sept. 1930.Since some Central Banks, e.g., the Reichsbank, were prepared to accept standard gold and pay out fine gold (atthe usual margin between their buying and selling prices), the theoretical limit to the movement of the franc-sterling exchange was set temporarily by the cost of a triangular transaction over some third country.


8The myth and realities of the so-called gold standard

Robert Triffin*

The monetary traditions and institutions of the nineteenth century provided a remarkably efficientmechanism of mutual adjustment of national monetary and credit policies to one another, essential to thelong-term maintenance of exchange-rate stability between national currencies.

The reasons for this success, and for the breakdown of the system after World War I, are very imperfectlyreflected in most of our textbooks. Most of all, however, over-concentration on the mechanism ofintercountry adjustments fails to bring out the broader forces influencing the overall pace of monetaryexpansion on which individual countries were forced to align themselves.

The mechanism of adjustment among countries

Textbook abstract

Starting from an initial position of balance-of-payments equilibrium, the emergence of a fundamentaldeficit is generally described in terms of divergent movements of exports—downward—and imports—upward— in the deficit countries, with opposite, and equally divergent, movements in the surplus countries.

The money flows associated with the international settlement of such imbalances if not offset bydomestic ‘neutralization’ policies, should then tend to prompt downward price readjustments in the deficitcountries, and upward readjustments in the surplus countries. This would restore a competitive price andcost pattern among them, and bring their balances of payments back into equilibrium.

These ‘automatic’ adjustment forces were strengthened and speeded up by central banks through the so-called ‘rules of the game’. Discount-rate policy and open-market interventions would raise interest rates andtighten credit in the deficit countries, while lowering interest rates and expanding credit in the surpluscountries’ This would both cushion balance-of-payments and monetary transfers in the short term, bystimulating compensatory capital movements from the surplus to the deficit countries and accelerate thedesirable downward readjustment of prices and costs in the latter countries and their upward readjustment inthe first.

The ‘rules of the game’ were widely violated after World War I. The surplus countries adopted‘neutralization’ policies which increasingly concentrated upon the deficit countries the burdens ofadjustment previously distributed between surplus and deficit countries alike. At the same time, the

* Our International Monetary System: Yesterday, Today and Tomorrow, New York, Random House, 1968, ch. 1. Alsoappears in The Evolution of the International Monetary System: Historical Reappraisal and Future Perspectives,Princeton, Princeton University Press, 1964, pp. 2–20.

development of stronger resistance to downward price and wage adaptations—particularly as a result of thegrowing strength of the trade unions—blocked the price-adjustment mechanism in the deficit countries,transfering its impact to fluctuations in economic activity and employment. The resulting social andpolitical strains gradually became unbearable, particularly during the world depression of the 1930s, andinduced governments to abandon the harsh gold standard disciplines in favour of fluctuating exchange ratesand/or trade and exchange restrictions.

Historical abstract

The highly simplified digest of the theory of international adjustment under the actual gold standardcertainly meets the first test of an economic theory, i.e., the test of logical consistency. Does it meet equallywell the second test by which a theory should be judged, i.e., its conformity to the major facts calling forexplanation?

It undoubtedly fits some of the facts. Comparative price—or exchange-rate—movements obviously playa role in the fluctuations of balances of payments on current account, and are themselves influenced by thetightening or expansion of money flows arising both from international settlements and from domestic policiesor lack of policies.

Other facts, however, must also be taken into account if we are to develop a general and politicallymeaningful theory of balance-of-payments adjustments.

1 First of all, the most cursory look at international trade statistics reveals an enormous degree ofparallelism—rather than divergent movements—between export and import fluctuations for any onecountry, and in the general trend of foreign-trade movements for the various trading countries. Over theeighty years from 1880 to 1960, all significant increases or decreases in the exports of Western Europe weremarked by parallel increases, or decreases, for the eleven major trading countries of the world in 91 per centof the cases, and by simultaneous increases, or decreases of exports and imports for each country, takenseparately, in 88 per cent of the cases. These proportions fall to 77 and 73 per cent, respectively, forfluctuations of one year only, but rise to 95 and 92 per cent for fluctuations of more than a year’s duration,and to 98 and 100 per cent for movements extending over more than four years.1

2 Equally impressive is the overall parallelism—rather than divergence—of price movements, expressedin the same unit of measurement, between the various trading countries maintaining a minimum degree offreedom of trade and exchange in their international transactions. In spite of wide differences andfluctuations in the composition of each country’s exports, the indexes of export unit values—measured incurrent dollars—for the same eleven countries over the period 1870–1960 moved in the same direction in 89per cent of the observed fluctuations, and in the opposite direction in only 11 per cent of the cases.2

This solidarity of national price movements—when measured in a common unit of account—is notincompatible, of course, with sharp divergences in national price levels, offset by opposite divergences inexchange-rate fluctuations. One does find indeed that any large variations in the evolution of national pricesare invariably offset, more or less rapidly, by exchange-rate fluctuations, and vice versa. Such variationswere, however, eschewed—except in wartime—by most industrial countries in the nineteenth century, butwere relatively frequent in the countries of the so-called ‘periphery’, and particularly in Latin America.

3 Downward wage adjustments rarely reached any sizeable amplitude, even in the nineteenth century,among the countries which maintained exchange-rate stability, and it may be doubted whether they wouldhave proved much more acceptable at that time, economically, politically and socially, than they are today.Wherever substantial inflation had been allowed to develop, international cost competitiveness was nearlyinvari ably restored through devaluation rather than through downward price and wage adjustments.

102 MYTH AND REALITIES OF THE GOLD STANDARD

Standard statistical series for the United States, the United Kingdom, France and Germany show onlyfour or five instances of actual declines in any broad-based indexes of money wages during the fifty yearspreceding World War I. Such declines were, moreover, usually confined to one or a few percentage pointsonly. They were far exceeded, in post-gold-standard days, by the much sharper wage drops of the 1920–2recession—37 per cent in the United Kingdom—and of the first years of the great depression—22 per centin the United States and Germany.3

4 The ‘neutralization’ policies stigmatized by Ragnar Nurkse as another major cause—along withincreasing price and wage rigidity—of the downfall of the gold standard (Nurkse, 1944, pp. 66–8) were byno means a postwar innovation. Using exactly the same techniques of measurement as Nurkse, Arthur I.Bloomfield found that ‘central banks in general played the rules of the game just as badly before 1914 asthey did thereafter!’4 It might be noted in passing, however, that Nurkse’s method defines as neutralizationthe cases where fluctuations in a central bank’s domestic portfolio offset only a fraction—no matter how small—of the changes in its international assets. In many cases, however, there remained a positive correlationbetween the latter and changes in the central bank’s sight liabilities. The impact of the latter changes uponthe country’s money supply would most often be magnified, in turn, several times by the operation of theprivate banking system under customary cash and liquidity requirements. Nurkse’s ‘neutralization’ policies,therefore, could still permit a multiple impact of international gold—or foreign-exchange—movementsupon money supply, as contrasted with the mere one-to-one impact which would have resulted under thepure gold coin system of monetary circulation assumed in the most abstract formulations of gold standardtheory (Triffin, 1947, pp. 52–3).

5 The impact of discount rates on cushioning capital movements and on corrective changes in costcompetitiveness was also far less general and uniform than is usually assumed.

The first seems indeed to have been particularly effective for the well-developed money and capitalmarkets of the major creditor countries and financial centres, and most of all in the case of the UnitedKingdom. Discount and interest-rate changes could accelerate, or slow down, the normal, or average, paceof capital exports, and had to be resorted to frequently by the Bank of England to defend its very slendergold reserves. The much higher reserve levels of the Bank of France enabled it, on the other hand, tocushion temporary deficits out of its own reserves, with much rarer recourses to discount-rate changes.Most of all, however, capital-importing countries were far less able to influence in the same way the pace oftheir capital imports, these being primarily determined by the ease or stringency prevailing in the majorfinancial centres.

The impact of Britain’s international surpluses and deficits on British bank reserves was cushioned,moreover, by the ample use of sterling balances as cash reserves by overseas banks, particularly throughoutthe British Empire. Surpluses and deficits between Britain and its Empire—and even, to some extent, withother countries—merely led to a reshuffling of British bank deposits, rather than to an overall expansion orcontraction in their amount and to correlative gold inflows or outflows.

Finally, the enormous role played by the London discount market in the financing of the food and raw-materials exports of the less developed countries probably imparted to the Bank of England’s discount-ratepolicy an influence on British terms of trade—and balance of payments—which has escaped the attention ofeconomic theorists. Increases in discount rates did—as is usually pointed out—tend to reduce British pricesand costs, improving the competitiveness of British exports in world markets and of home-made import-substitute goods on the domestic market. What is forgotten, however, is that the tightening of the Londondiscount market also affected, most directly and overwhelmingly, the ease with which inventories of staplefoods and raw materials could be financed, thus forcing also a quicker liquidation and attendant pricedeclines in Britain’s chief import goods. Such declines could be expected to be far larger than those in the


less sensitive and volatile prices of British industrial exports. Thus, the favourable impact of discount-rateincreases on British competitiveness (lowering British prices in relation to foreign prices in competingindustrial nations) would be reinforced in its balance-of-payments effects by a simultaneous improvementof Britain’s terms of trade (i.e., by decreases in the prices of foreign suppliers of complementary goods toBritain, larger than the decreases in British export prices to them). (See Triffin (1947, pp. 60–3), and Kenen(1960, pp. 59–62).)

6 The importance of international capital movements, and of their fluctuations, is often obscured by thedisproportionate emphasis often placed on comparative price and cost fluctuations as the major factor inbalance-of-payments disequilibria and their correction. Attention is thereby centred on the current-accountitems of the balance of payments, and tends to suggest that most disturbances arose in this area and had tobe corrected promptly by the restoration of equilibrium between receipts and expenditures on current—oreven merely merchandise—account.

In fact, however, international capital movements often did cushion—and even stimulate—vast andenduring deficits, or surpluses, on current account without calling for any correction whatsoever, except inan extre mely long run indeed. Developing countries, such as the United States, Canada, Argentina andAustralia, could maintain, over an average of years, large and persistent deficits on current account,financed by correspondingly large, persistent and growing capital imports from the more advancedcountries of Western Europe. Rough estimates, compiled by the United Nations (1949, p. 2), place at about$40.5 billion, on the eve of World War I, the gross long-term foreign investments of the principal creditorcountries of Western Europe, and at $3.5 billion those of the United States. Of this $44 billion total, $12billion had been invested in Europe itself, $6.8 billion in the United States—which was still a net debtorcountry at the time—$8.5 billion in Latin America, $6.0 billion in Asia, $7.4 billion in Africa, $3.7 billionin Canada, and $2.3 billion in Australia and New Zealand.

The lion’s share of these investments was that of the United Kingdom ($18 billion), followed by France($9 billion) and Germany ($5.8 billion). The United Kingdom had indeed been running persistent andgrowing surpluses on current account for more than a century, without any tendency whatsoever towardsequilibrium. On the contrary, these surpluses rose continually from about $35 million a year, on theaverage, over the years 1816–55 to more than $870 million a year in the last years before World War I(1906–13). Nobody could ever dream of explaining this favourable balance—and its fluctuations—in termsof the cost-competitiveness adjustment mechanism depicted in the textbooks, since it arose primarily fromBritain’s earnings on its swelling foreign-investment portfolio, and coincided with large and increasingdeficits on merchandize account—close to $670 million a year over the period 1906–13—offset themselves,for the most part, by net receipts on services and remittances account.

These current-account surpluses were nearly fully absorbed by Britain’s investments abroad, which roseover the same period from an average of less than $30 million a year in 1816–55 to more than $850 milliona year in 1906–13, and indeed more than a billion dollars a year in the last three prewar years, i.e., about athird of the British export level at the time, and 10 per cent of net national income (Imlah, 1958, pp. 70–5).

Foreign investments on such a scale undoubtedly accelerated economic development and helped at timesrelieve balance-of-payments pressures in the recipient countries. In the case of the United States, forinstance, net capital inflows from Europe—primarily Britain—financed large and growing deficits oncurrent account throughout most of the nineteenth century. They reached a peak of close to $300 million in1888, tapering off afterwards, and shifting to net capital exports around the turn of the century, as theUnited States finally turned from chronic deficits to equally chronic surpluses on current account (USBureau of the Census, 1960, pp. 562–6).


7 The cyclical pattern of international capital movements, however, had a very different impact upon thecapital-exporting and the capital-importing countries.

A mere slowdown of capital exports could help relieve, in the first countries, any pressures on central-bank—and private-bank—reserves arising from unfavourable developments in other balance-of-paymentstransactions. In the British case, for instance, capital exports dropped year after year, from their 1872 peakof roughly $480 million to $60 million in 1877, recovered again to $480 million in 1890, and declined oncemore in the following years to $110 million in 1898, rising nearly uninterruptedly afterwards to $250million in 1904, and booming to $400 million in 1905, $570 million in 1906, to reach finally close to $1,100million in 1913 (Imlah, 1958, pp. 73–5).

The borrowing countries, on the other hand, were far less able to control the rate of their capital importswhich tended, on the whole, to swell in boom times and dry up in hard times, contributing further to theeconomic instability associated with their frequent dependence on one or a few items of raw material orfoodstuff exports, themselves subject to wide quantity and/or price fluctuations. All in all, therefore, thebalance of payments of the countries of the so-called ‘periphery’ would be assisted, over the long run, bythe large capital imports available to them from the financial markets of industrial Europe, but thesecountries would pay for this dependence through perverse fluctuations in the availability of such capital andin their terms of trade over the cycle. The exchange-rate instability of most underdeveloped countries—otherthan those of colonial or semi-colonial areas tightly linked to their metropolitan country’s currency andbanking system—finds here one of its many explanations.5

8 Another important qualification of the traditional theory of balance-of-payments adjustments relates tothe international timing of reserve movements and discount-rate changes. The textbook explanationsuggests that rate increases were undertaken by the deficit countries in order to relieve a drain of theirreserves to the surplus countries. As noted by Bloomfield, however:

the annual averages of the discount rates of twelve central banks [England, Germany, France,Sweden, Finland, Norway, Denmark, Belgium, Switzerland, the Netherlands, Russia, and Austria-Hungary] reveal the…interesting fact that, in their larger movements at least, the discount rates ofvirtually all the banks tended to rise and fall together. … To some degree, and certainly for many ofthe banks, this broad similarity reflected competitive or ‘defensive’ discount rate changes…. But amore important explanation lies in the fact that discount rates in most…of the individual countriestended…to show a positive correlation, though generally not a very marked one, with domesticbusiness cycle fluctuations. Since, as is well known, major cyclical fluctuations tended to be broadlysynchronous in all countries, discount rate movements thus generally tended to exhibit a broadparallelism over the course of the world cycle—although there were, of course, many dissimilarities withrespect to short-term movements in the various countries. (Bloomfield, 1959, pp. 35–7)

This importance of parallel movements, associated with the international business cycle—as againstdivergent movements between surplus and deficit countries—brings us back to the first two points made onp. 142, and to the comparative neglect of this parallelism in textbook discussions centred nearly exclusivelyon intercountry balance-of-payments adjustments.

Reinterpretation and conclusions

1 The nineteenth-century monetary mechanism succeeded, to a unique degree, in preserving exchange-ratestability—and freedom from quantitative trade and exchange restrictions—over a large part of the world.


2 This success, however, was limited to the more advanced countries which formed the core of the system,and to those closely linked to them by political, as well as economic and financial ties. The exchange ratesof other currencies—particularly in Latin America—fluctuated widely, and depreciated enormously, overthe period. This contrast between the ‘core’ countries and those of the ‘periphery’ can be largely explainedby the cyclical pattern of capital movements and terms of trade, which contributed to stability in the firstgroup, and to instability in the second.

3 The adjustment process did not depend on any tendency towards equilibrium of the national balances ofpayments on current account. Vast and growing capital movements cushioned over many years, up to acentury or more, correspondingly large and increasing surpluses—and deficits—on current account.

4 The preservation of exchange-rate stability depended, however, on the impact of international monetarysettlements—of the combined current and capital accounts—upon domestic monetary and creditdevelopments. Large or protracted deficits or surpluses had to be corrected, residually, by a slowdown oracceleration of bank-credit expansion sufficient to bring about—through income and/or price and costadaptations, and their impact on exports and imports—a tenable equilibrium in overall transactions, and acessation of persistent drains in the deficit countries’ stock of international money (i.e., gold and silverinitially, and increasingly gold alone as all major countries shifted from the silver or bimetallic standard tothe gold standard).

5 This residual harmonization of national monetary and credit policies, depended far less on ex postcorrective action, requiring an extreme flexibility, downward as well as upward, of national price and wagelevels, than on the ex ante avoidance of substantial disparities in cost competitiveness in the monetarypolicies which would allow them to develop.

As long as stable exchange rates were maintained, national export prices remained strongly boundtogether among all competing countries, by the mere existence of an international market not broken downby any large or frequent changes in trade or exchange restrictions. Under these conditions, national priceand wage levels also remained closely linked together internationally, even in the face of divergent rates ofmonetary and credit expansion, as import and export competition constituted a powerful brake on theemergence of any large disparity between internal and external price and cost levels.

Inflationary pressures could not be contained within the domestic market, but spilled out directly, to aconsiderable extent, into balance-of-payments deficits rather than into uncontrolled rises of internal prices,costs, and wage levels.6 These deficits led, in turn, to corresponding monetary transfers from the domesticbanking system to foreign banks, weakening the cash position of domestic banks and their ability to pursueexpansionary credit policies leading to persistent deficits for the economy and persistent cash drains for thebanks. (Banks in the surplus countries would be simultaneously subject to opposite pressures, which wouldalso contribute to the harmonization of credit policies around levels conducive to the re-equilibration of theoverall balance of payments.)

Central banks could, of course, slow down this adjustment process by replenishing through their discountor open-market operations the cash reserves of the commercial banks. As long as exchange controls ordevaluation were effectively ruled out from their horizon, however, they would themselves be responsive tosimilar pressures, arising from the decline in the ratio of their own reserves to liabilities. While theirliabilities were internal, and thus easy to expand, their reserves were—and still are today—limited tointernational assets over which they had no direct control.

6 These pressures for international harmonization of the pace of monetary and credit expansion wereindeed very similar in character to those which continue today to limit divergent rates of expansion amongprivate banks within each national monetary area.


They were further reinforced, as far as central banks were concerned, by the fact that a substantial portionof the domestic monetary circulation itself was in the form of commodity money—gold and silver—whollyor partly international in character, rather than in credit money. Expansionary credit policies were thusaccompanied by an outflow of gold and silver assets from the coffers of central banks into internalcirculation and commercial banks’ reserves, as well as to foreign countries. This movement of specie intointernal circulation was all the more pronounced, as the lowest denomination of paper currency was usuallymuch too high—often equivalent to several times the level of monthly wages—to be usable in householdand wage payments. Central bank credit expansion was therefore limited not only by foreign deficits andgold losses, but also by internal gold and silver losses, very much as commercial banks’ credit and depositexpansion may be limited today by the drain on their paper currency reserves. While the latter can bereplenished by central-bank credit, central banks themselves did not have access to any gold or silver‘lender of last resort’.

The overall pace of advance of commercial banks’ credit and depositmoney creation in a nationaleconomy was and remains subject today to the policies of the central bank. Similarly, the overall pace ofcredit creation by the central banks as a group was limited, in the nineteenth century’s international economy,by their ability to increase simultaneously their international reserves.

7 This latter observation brings once more into the limelight a most important question left unansweredby the theory of balance-of-payments adjustment among countries: granted the need for mutualharmonization of national monetary policies among the gold standard countries, what were the factorsdetermining the international pace on which such alignments did take place? The question is all the moresignificant in view of the size and parallelism of major fluctuations in national price, export, and importlevels over the period 1815–1914 as a whole.

The international pace of adjustment

A gentle reminder of the apostles of gold money

1 The gold standard is often credited with having reconciled, to an unprecedented degree, price stabilitywith a high rate of economic growth over the nineteenth century. Contemporary advocates of a return togold rarely miss the opportunity of quoting, in this respect, Gustav Cassel’s observation that ‘the generallevel of prices in 1910 was practically the same as in 1850’.7 This stability is then attributed to thesafeguards erected against inflation by the small size of new gold production and monetary gold increases inrelation to existing stocks, and, more generally and optimistically, to the response elasticity of new goldproduction to any substantial decreases or increases in the price level: price declines or increases would bekept in check by their impact on gold-mining costs and profitability, and the resulting stimulation orslowdown of new gold production and monetary expansion.

2 As pointed out by Cassel himself, however, price fluctuations were by no means inconsiderable in thenineteenth century. Increases and decreases of 30 to 50 per cent, or more, accompanied the famousKondratieff cycles (Kondratieff, 1926), and have been attributed by many writers—including Cassel—tofluctuations in gold production, following new mining or refining discoveries.

The evidence of long-term stability—or rather reversibility—of prices seen in the return of the 1910index to its 1850 level is, to say the least, extremely misleading. Such an arbitrary choice of dates wouldallow us, for instance, to demonstrate equally well the ‘stability’ of the price level over the period from1913 to the early 1930s, since the precipitous fall of prices during the Great Depression brought back boththe US and the UK price indexes down to approximately their 1913 level in 1931–2!


The starting point of Cassel’s comparison—1850—is taken close to the very bottom of a long depressionduring which prices had fallen by 50 per cent or more, while the end year—1910—comes at the end of afifteen-year upward trend during which the index used by Cassel had risen by more than 30 per cent.

Making the same comparison from peak to peak, or from trough to trough, we would find a ratherpronounced downward long-run trend of wholesale prices in all major countries (Table 8.1). Prices declined,for instance, by 25 per cent in the United States from 1814 to 1872, and by 25 per cent again from 1872 to1913, adding up to a cumulative 44 per cent decline over the century, from 1814 to 1913. In the UnitedKingdom, price declines of 30 per cent from 1814 to 1872, and 20 per cent from 1872 to 1913 also add upcumulatively to a similar 44 per cent decline for the century as a whole.

3 The influence of fluctuations in gold production upon these broad price trends seems far more plausiblethan the supposed inverse relationship from commodity prices to gold production. The significance of anysuch relationship as may have existed was certainly dwarfed by the gold avalanche unleashed by thediscovery of new gold fields and the improvement of mining and refining techniques, both after 1848 andafter 1888. On

Table 8.1 Wholesale price indexes, 1814–1913

US UK Germany France Italy

Indexes(1913= 100)

1814 178 178 129 1321

1849 80 90 71 96

1872 133 125 111 124

1896 67 76 71 71 74

1913 100 100 100 100 100

Changes (in %)

1814–1849 −55 −49 −45 −272

1849–1872 +66 +39 +56 +31

1872–1896 −50 −39 −36 −43

1896–1913 +49 +32 +41 +41 +35

1814–1913 −44 −44 −22 −242

Sources:(a) For the United States(a) Warren and Pearson index until 1890(b) BLS index since 1890(b) For the United Kingdom(a) Gayer, Rostow and Schwartz index until 1849(b) Rousseaux index from 1844 to 1871(c) Board of Trade index since 1871(c) For Germany, France and Italy: France (1952), pp. 513–15.Notes:1 18202 since 1820

both occasions, current production just about doubled, over twenty-four or twenty-five years, the gold stockaccumulated over the previous three-and-a-half or four centuries. The yearly rate of growth in the estimated


monetary gold stocks—after deduction for hoarding, industrial and artistic uses—rose abruptly from 0.7 percent in the first half of the nineteenth century to 4.3 per cent over the years 1849–72, declined precipitouslyto only 1.3 per cent in 1873–88, and rose again to 3.2 per cent in 1889–1913.

4 The neat mechanistic explanation derived by some authors from this broad parallelism between goldproduction and long-run trends in commodity prices fails, however, to give a full account of the complexfactors involved in the process of nineteenth-century economic growth. The Kondratieff long waves werecertainly influenced also to a major degree by the clustering and spread of technological discoveries andinnovations in production, transportation, etc., by the vast migrations from old to new settlement areas, and—last but not least—by the preparation, waging and aftermath of wars. These powerful influences,brilliantly analysed by Schumpeter (1934, 1939) among others, obviously cannot be reduced to anymechanistic monetary explanation. It would be equally absurd, on the other hand, to deny that monetary andbanking developments also had a role—even if primarily permissive, rather than initiating—on theacceleration or retardation of price trends and production growth. Schumpeter himself insisted abundantlyon the role of bank credit in the process of capitalistic development.

One might well wonder, indeed, whether the unprecedented stability of the major currencies in terms ofgold—and exchange rates—in the nineteenth century was not due to the spectacular growth of bank moneyor ‘credit money’—in the form of paper currency and bank deposits—rather than to the residual, and fastdeclining, role of gold and silver ‘commodity money’. Certainly, full dependence of the monetary system ongold and silver, in pre-nineteenth-century days, to the exclusion or near-exclusion of credit or paper money,did not prevent wide inflationary excesses—through debasement of the coinage—and wide fluctuations inexchange rates. The pound sterling lost three-fourths of its gold value and the French franc more than nine-tenths, from the middle of the thirteenth century to the end of the eighteenth century.

5 It is rather ludicrous to reflect that the vast literature devoted to the so-called nineteenth-century goldstandard is practically devoid of any quantitative estimates of the enormous changes that modified, out ofall recognition, the actual structure of the volume of money, or means of payments, as between gold, silver,currency notes, and bank deposits, between the end of the Napoleonic Wars and the outbreak of World WarI.

Yet, according to the League of Nations estimates, paper currency and bank deposits already accounted in1913 for nearly nine-tenths of overall monetary circulation in the world, and gold for little more than one-tenth. Comprehensive estimates for earlier periods are practically nonexistent and can only be piecedtogether from disparate sources, the reliability of which is most difficult to assess. Yet, some broad facts andorders of magnitude can hardly be in doubt. Bank currency and demand deposits probably constituted lessthan a third of total money supply at the beginning of the nineteenth century, but close to nine-tenths by1913. Silver exceeded gold in actual circulation by about two or three to one until well into the second halfof the century, but dropped considerably behind in the latter part of the period, the previous proportionbeing just about reversed by 1913. Increases in credit money—paper currency and demand deposits—accounted, in the major and more developed countries, for two-thirds or more of total monetary expansionafter the middle of the century, and more than 90 per cent from 1873 to 1913.

These facts can hardly be reconciled with the supposed automaticity still ascribed by many writers—particularly in Europe—to the so-called nineteenth-century gold standard. The reconciliation of high ratesof economic growth with exchange-rate and gold-price stability was made possible indeed by the rapidgrowth and proper management of bank money, and could hardly have been achieved under the purely, orpredominantly, metallic systems of money creation characteristic of the previous centuries. Finally, the term‘gold standard’ could hardly be applied to the period as a whole, in view of the overwhelming dominance ofsilver during its first decades, and of bank money during the latter ones. All in all, the nineteenth century


could be far more accurately described as the century of an emerging and growing credit-money standard,and of the euthanasia of gold and silver moneys, rather than as the century of the gold standard.

Monetary expansion and international reserves before World War I

A more precise assessment of the nature of the nineteenth-century international monetary mechanism and ofits relation to production and price fluctuations must await the development of better monetary and reservestatistics than are now available, not only for the world as a whole, but even for the major countries whichformed the basic core of the so-called gold standard.

The task should not prove impossible, if two limitations are accepted from the start. The first relates tothe dearth of meaningful and reasonably reliable statistics for many countries. This should not prove toodamaging for an appraisal of the international monetary mechanism in the few major countries whichformed in the nineteenth-century—and still form today—the core of the system. I have assembled somerough estimates of this sort, running back to 1885, for eleven such countries (the present so-called Group ofTen, or Paris Club, plus Switzerland). They account in 1885 and 1913 for 60 to 80 per cent of the worldmoney supply and monetary reserves. Earlier estimates—back to 1815—are for three countries only (theUnited States, the United Kingdom, and France) but accounted for about half the world money and reservesin 1885 and 1913, and for about two-thirds to three-fourths of the eleven core countries.8 Table 8.2 givesfurther indications in this respect, revealing an encouraging parallelism between the estimates in the threegroups.

The second limitation lies in the incompleteness and lack of full comparability of available data even forthe major countries. Yet, this could hardly be more damaging than similar—and often far worse—limitations

Table 8.2 Comparative evolution of money and reserve structure, 1885 and 1913

End of Three countries1 Eleven countries2 World

1885 1913 1885 1913 1885 1913

(in billions of US dollars)

1 Money supply 6.3 19.8 8.4 26.3 14.2 33.1

a Gold 1.4 2.0 1.8 2.7 2.4 3.2

b Silver 0.7 0.6 1.0 1.2 3.0 2.3

c Credit money 4.1 17.2 5.6 22.4 8.8 27.6

i Currency3 1.6 3.8 2.3 5.9 3.8 8.1

ii Demand deposits 2.6 13.3 3.3 16.5 5.0 19.6

2 Monetary reserves 1.0 2.7 1.5 4.0 2.0 5.3

a Gold 0.6 2.1 0.9 3.2 1.3 4.1

b Silver 0.4 0.6 0.6 0.8 0.7 1.2

3 Total gold and silver 3.1 5.4 4.3 7.9 7.4 10.8

a Gold 2.0 4.1 2.7 5.9 3.7 7.3

b Silver 1.1 1.2 1.6 2.0 3.7 3.5

(in % of money supply)

1 Money supply 100 100 100 100 100 100

a Gold 23 10 21 10 17 10

b Silver 11 3 12 5 21 7


End of Three countries1 Eleven countries2 World

1885 1913 1885 1913 1885 1913

c Credit money 66 87 67 85 62 83

i Currency3 25 19 27 22 27 25

ii Demand deposits 41 67 39 63 35 59

2 Monetary reserves 16 14 18 15 14 16

a Gold 9 11 11 12 9 12

b Silver 7 3 7 3 5 4

3 Total gold and silver 49 27 51 30 52 33

a Gold 32 21 32 22 26 22

b Silver 17 6 19 8 26 11

Notes:1 United States, United Kingdom and France2 United States, United Kingdom, France, Germany, Italy, Netherlands, Belgium, Sweden, Switzerland, Canada and

Japan3 Including subsidiary (non-silver) coinage, except in last column

on the validity of other nineteenth-century estimates, in the field of national accounting for instance. Theycertainly remain, moreover, very minor in relation to the broad orders of magnitude involved in theenormous shifts in the monetary structure. […] In any case, imperfect as they are bound to be, suchestimates are essential to an understanding of the ‘nineteenth-century international monetary mechanism,and far better than the implicit and totally unwarranted assumptions that underlie most of past and currenttheorizing about the so-called gold standard.

With these qualifications in mind, the following observations can be derived from these tables:1 Although the 1816–48 estimates are particularly venturesome, there can be no doubt about the very

slow growth of monetary gold stocks—just about nil, if we can trust the estimates—and of total moneysupply—about 1.4 per cent a year—over this period. Monetary expansion was sustained, not by goldaccretions, but by an approximate doubling of silver stocks, accounting for about two-thirds of the totalincrease in the money supply, and for the remaining third by the incipient increase in internal creditmonetization.9

2 The gold avalanche of the next twenty-four years produced an average increase of 6.2 per cent yearly inthe total stock of monetary gold. This rate of growth declined sharply, to about 1.4 per cent a year, from1873 to 1892, but recovered to about 3.7 per cent in the last twenty years preceding the outbreak of WorldWar I.

These enormous fluctuations in gold-stock increases were significantly smoothed down by concurrentadaptations in the functioning of the monetary and banking system. The yearly rate of growth of moneysupply declined only from 4.2 per cent in 1849–72 to 3.3 per cent in 1873–92, and recovered to 4.3 percent, on the average, in the period 1893–1913.

This smoothing down was due, to a minor extent, to the partial offsetting of gold fluctuations by oppositefluctuations in the monetary silver stocks. These contracted substantially in the two periods of fastest goldexpansion, but more than doubled during the leaner gold years from 1873 through 1892. Far moresignificant is the dwarfing of gold and silver stock changes by the spectacular growth of credit money,which fed more than 70 per cent of total money increases over the years 1849–72, to about 34 per cent (seeTable 8.3).


3 Credit money—i.e., paper currency and bank deposits—did not, however, normally circulate beyondthe national borders of the issuing country and banking institutions. Exchange-rate stability thus dependedon their ready convertibility—directly by the issuing banks, or ultimately through a national central bank—into the foreign currencies required, or into metallic currencies or bullion of international acceptability.Silver bullion lost its previous role in this respect around 1872, and silver coin settlements remainedacceptable only among the countries of the Latin Monetary Union. Silver, however, was no longer ‘full-bodied’ money, as the commercial value of silver coins fell well below their nominal value.10 Gold thusemerged increasingly as the primary guarantor of international exchange stability even for the countrieswhich remained on a so-called ‘limping’ bimetallic standard.

Table 8.3 Composition of money and reserve increases, 1816–1913: United States, United Kingdom and France

1816–1913 1816–48 1849–72 1873–92 1893–1913

(in millions of US dollars)

1 Money increases 18,791 581 2,688 3,863 11,659

a Gold 1,673 −55 913 81 734

b Silver 287 379 −167 132 −57

c Credit money 16,831 257 1,942 3,650 10,982

i Currency and coin 3,551 44 1,044 461 2,002

ii Demand deposits 13,280 213 898 3,189 8,980

2 Reserve increases 2,675 81 215 1,046 1,333

a Gold 2,097 62 218 379 1,438

b Silver 578 19 −3 667 −105

3 Total gold and silver

increases 4,635 405 961 1,259 2,010

a Gold 3,770 7 1,131 460 2,172

b Silver 865 398 −170 799 −162

4 Internal credit monetization(1–3=lc—2)

14,156 176 1,727 2,604 9,649

(in % of money increases)

1 Money increases 100 100 100 100 100

a Gold 9 −9 34 2 6

b Silver 2 65 −6 3 —

c Credit money 90 44 72 95 94

i Currency and coin 19 8 39 12 17

ii Demand deposits 71 37 33 83 77

2 Reserve increases 14 14 8 27 11

a Gold 11 11 8 10 12

b Silver 3 3 — 17 −1

3 Total gold and silver

increases 25 70 36 33 17

a Gold 20 1 42 12 18

b Silver 5 69 −6 21 −1


1816–1913 1816–48 1849–72 1873–92 1893–1913

4 Internal credit monetization 75 30 64 67 83

(% absorption of new gold into)

1 Reserves 56 886 19 82 66

2 Circulation 44 −786 81 18 34

Three factors explain the maintenance of stable exchange rates in the face of growing issues of nationalcredit moneys, side by side with fast declining proportions of international gold and silver moneys.

The first is the de facto harmonization of the national rates of monetary and credit expansion among thegold standard countries. This harmonization itself, however, depended, as pointed out on pp. 143–4, on thereaction of the issuing banks to the fluctuations in their reserve ratio arising from cyclical movements ininternal circulation, as well as from external settlements of balance-of-payments disequilibria.

The overall pace of expansion, in turn, could not but be strongly influenced by the ability of the nationalbanking systems to accumulate sufficient gold reserves to guarantee the convertibility of their nationalcredit money issues into the gold through which foreign currencies could be acquired at stable exchangerates. The maintenance of relatively fast rates of monetary expansion after 1848 was thus conditioned bytwo further factors which the tables bring clearly into light.

The first was the spectacular spurt in gold production that followed the discovery of new gold fields andimproved mining and refining techniques, and was of course predominantly accidental in character.

The second lay in the resiliency and adaptability of monetary and banking institutions, and the enormouseconomy of the precious metals which resulted from their increasing transfers from actual circulation in thepublic to the reserve coffers of commercial banks and of national central banks—or Treasury in the case ofthe United States.11 The proportion of monetary gold and silver stocks absorbed in centralized monetaryreserves rose from about 10 per cent in 1848 to 16 per cent in 1872, 41 per cent in 1892, and 51 per cent in1913.12 Even more significant is the relative proportion of new gold accretions absorbed by central reserves,on the one hand, and by the public and banks on the other. During the first gold avalanche of 1849– 72, 81per cent of the new gold was dispersed among the public and banks, only 19 per cent being accumulated inreserves. These proportions were nearly exactly reversed in the leaner gold years from 1873 through 1892,82 per cent of the new gold feeding the increase of central reserves, with a multiple impact on overallmoney creation. When gold production rose again at a faster pace in the period 1893–1913, the proportionabsorbed by central reserves declined to 66 per cent, while that of private holdings rose from 18 to 34 percent (see Table 8.3).

These spectacular changes in the structure of money and reserves thus contributed powerfully both to themaintenance of relatively fast rates of monetary expansion, and to a considerable smoothing out of moneysupply fluctuations in relation to fluctuations in the available gold stocks.

4 There was nothing inherently stable, however, in a process of monetary creation so heavily dependenton the accidents:

(a) of gold and silver discoveries and production rates;(b) of uncoordinated—and largely irrational—national decisions regarding the adoption, retention or

abandonment of silver, gold or bimetallism as the basic monetary standard; and (c) of compensatory adaptations in banking structure, the scope of which would inevitably taper off

over time, especially when central banks could no longer replenish their own reserves from thedwindling—relatively, if not yet absolutely—amounts of gold still in circulation.


In any case, the slow evolution which had adjusted gradually the international monetary system of thenineteenth century to the economic requirements of peacetime economic growth, but had also changed itout of all recognition between 1815 and 1913, was brutally disrupted by the outbreak of World War I. Theensuing collapse of the system ushered in half a century of international monetary chaos, characterized bywidespread exchange-rate instability and/or trade and exchange controls, with only brief interludes ofnostalgic and vain attempts to fit upon the twentieth-century economy the monetary wardrobe of thenineteenth-century world.

Notes

1 The above percentages are derived from 287 observations of national increases or decreases for eleven countries(the United States, the United Kingdom, France, Germany, Italy, Belgium, the Netherlands, Switzerland, Sweden,Austria and Canada), in the course of seventeen upward or downward movements of more than 1 per cent in WesternEuropean exports, in the period 1880–1960. The estimates used in these calculations are those of Maddison(1962), pp. 179–81.

2 Based on estimates from Maddison (1962), pp. 189–90.3 See, for instance, US Bureau of the Census (1960), pp. 90–2; Mitchell (1962), pp. 343–5; and France (1939), pp.

443–4.4 Bloomfield (1959), p. 50. The evidence of neutralization, measured by Nurkse’s formula, was present in 60 per

cent of total observations, in the period 1880–1913, coinciding exactly with Nurkse’s results for the 1922–38period.

5 Another, closely connected with the main topic of this study, lies in the retention of a silver standard long afterthe effective abandonment of silver or bimetallic standards in Europe and the United States.

6 This is still true today, in the absence of major changes in exchange rates and/ or trade and exchange restrictions.See Triffin and Grubel (1962), pp. 486–91.

7 See Cassel (1930), p. 72. The calculation is based on the Sauerbeck-Statist index of wholesale prices, and carriedback to 1800 on the basis of Jevons’s index. See also Kitchin (1930), pp. 79–85.

8 World totals, however, are somewhat incomplete and particularly unreliable.9 The latter being measured, indifferently, by the excess of money supply increases over the increase of monetary

gold and silver stocks, or by the excess of credit-money increases over the increase of monetary reserves. 10 The valuation of silver at nominal par in the tables thus understates the importance of credit money, since silver

coinage included in effect a substantial credit money component. Its acceptance at par among the countries of theLatin Union demonstrates the feasibility of international credit-money settlements, even under the very imperfectarrangements negotiated to this effect among the countries of the Latin Union.

11 The reserve estimates of the tables refer to the centralized holdings of central banks and treasuries only. The goldand silver components of money supply estimates include, therefore, gold and silver held by other issuing banksand commercial banks, thus overstating once more the metallic component of money supply in the modern senseof the word—coin, currency and demand deposits in the hands of the public—and understating the proportion ofcredit money in circulation outside banks.

12 The proportion of gold alone temporarily dropped from 31 per cent in 1848 to 20 per cent in 1872, rising later to35 per cent in 1892, and 51 per cent in 1913. The 1848–72 decline, however, was more than compensated by theincreased absorption into centralized reserves of silver which could still be regarded at that time as a valid reservecomponent. After 1872, the movements of gold alone are more significant than those of gold and silvercombined.


References

Bloomfield, Arthur I. (1959), Monetary Policy under the International Gold Standard: 1880–1914, New York, FederalReserve Bank of New York.

Cassel, G. (1930), ‘The Supply of Gold’, in Interim Report of the Gold Delegation of the Financial Committee, Geneva,League of Nations.

France (1939), (1952), Annuaire Statistique for 1938, 1951, Paris, Institut National de la Statistique et EtudesEconomiques.

Imlah, A.H. (1958), Economic Elements in the Pax Britannica, Cambridge, Mass., Harvard University Press.Kenen, Peter B. (1960), British Monetary Policy and the Balance of Payments, Cambridge, Mass., Harvard University

Press.Kitchin, J. (1930), ‘The Supply of Gold Compared with the Prices of Commodities’, in Interim Report of the Gold

Delegation of the Financial Committee, Geneva, League of Nations.Kondratieff, N.D. (1926), ‘The Long Waves in Economic Life’, Review of Economic Statistics (November 1935)

(abridged in English by W.Stolper).Maddison, A. (1962), ‘Growth and Fluctuations in the World Economy’, Banca Nazionale del Lavoro Quarterly Review

(June).Mitchell, B.R. (1962), Abstract of British Historical Statistics, Cambridge, Cambridge University Press.Nurkse, Ragnar (1944), International Currency Experience, Geneva, League of Nations.Schumpeter, Joseph A. (1934), The Theory of Economic Development, New York, McGraw-Hill.——(1939), Business Cycles, Cambridge, Mass., Harvard University Press.Triffin, R. (1947), ‘National Central Banking and the International Economy’, International Monetary Policies,

Washington, DC, Federal Reserve System.Triffin, R. and H.Grubel (1962), ‘The Adjustment Mechanism to Differential Rates of Monetary Expansion among the

Countries of the European Economic Community’, Review of Economics and Statistics (November).United Nations (1949), International Capital Movements during the Inter-war Period, New York, United Nations.US Bureau of the Census (1960), Historical Statistics of the United States, Washington, DC, US Government Printing

Office.


9Notes on the working of the gold standard before 1914

A.G. Ford*

The pre-1914 gold standard may be seen as a sterling standard largely, with London as its pivot, which hadgradually developed in the nineteenth century. Despite the use of a common international currency, somestriking features of this standard and its operation were the differences in the economic behaviour of thecountries who were members of this ‘club’,

Much of the theoretical and practical discussion of the gold standard has centred on the repercussions ofgold movements and of the consequential banking measures on international capital movements, incomes,and prices in the countries concerned. Less attention has been paid to the following two questions: (a) whydid the exchanges move to either of the gold points and gold move? (b) what equilibrating forces might beset in motion by the factors causing the gold movement? It will be argued that it is of crucial importance toask these questions in order to understand the actual operation of the gold standard, for the same factors (e.g.,a fall in export values or a fall in receipts from foreign borrowing) causing the export of gold or the loss offoreign balances may automatically set in motion income movements which reduce the initial discrepancybetween foreign-currency receipts and payments. Two cases will be considered: (a) Great Britain, and (b)Argentina, an economy which had a chequered history of adherence (and otherwise) to the gold standardbefore 1914. Contrasts will emerge between the operation of the gold standard for the central country andfor a ‘periphery’ country, both of which were geared together by trade and capital flows. This approach willthus follow up the suggestion of P.Barrett Whale, who, after emphasizing that the suggestions in his well-known article1 were tentative, stated: ‘They provide a hypothetical view of the working of the goldstandard…; but it is admitted that my view requires a more exhaustive testing, with regard to both the casesin which the gold standard has worked successfully and those in which it has broken down, before it canclaim acceptance.2

*This section will attempt to set up a general model of adjustment for gold standard conditions, although

certain portions may well be absent or lack importance when actual conditions of particular economies areconsidered. The principal answers to the question ‘Why did an excess of foreign currency payments overreceipts occur in an economy and gold move out?’ may be classified as follows:3

1 An increase in import purchases due to an increase in incomes generated by domestic forces (not by anincrease in exports), or at the expense of savings with given incomes.

2 An increase in import purchases, more being spent from a given income on foreign goods and less onhome goods—for example, because of changes in tastes, tariffs or prices.

* Oxford Economic Papers, February 1960, pp. 52–76.

3 A fall in the net proceeds of foreign borrowing transferred to the country and being used to financedomestic expenditure.

4 An increase in lending abroad at the expense of domestic purchases of consumption goods or at theexpense of home investment.

5 A fall in export values, either because foreign markets were lost, world prices fell, or more exportableswere purchased at home.

6 A transfer of bank deposits from one international monetary centre to another. This might well be anunrepeated ‘stock’ transaction, rather than a change in a continuing flow of payments, and the loss of goldwould cease automatically in this case.

First, if gold is lost because of factors 2, 3, 4, 5, these disturbing factors will set in motion contractionaryincome movements in the country, which will be equilibrating in the sense that they will reduce thediscrepancy between foreign-currency receipts and payments by reducing import purchases4 (vice versa foran influx of gold). These equilibrating income movements, I suggest, deserve more attention when the actualoperation of the gold standard is considered—they account for the speedy adjustment of imports noted incertain cases—and will be dubbed ‘automatic’ forces for the purposes of this paper.

Secondly, if the loss of gold is allowed to affect the domestic supply of money (i.e., no ExchangeEqualization Account), then the contraction of the quantity of money will bring rising interest rates5 and acontraction in bank credit, both of which will tend to reduce incomes and import purchases. This forms asecond ‘automatic’ equilibrating force.

Thirdly, these may very well be ‘induced’ measures undertaken to check the loss of gold—an increase in‘Bank Rate’ and the measures taken to make it effective. These will also tend to reduce incomes and importpurchases. Moreover, in countries where there is a substantial gold coin circulation, falling domestic activityand incomes will reduce the demand for this type of money so that there will be an ‘internal reflux’ of goldcoins to swell the banking system’s or the central bank’s gold reserves.

In summary, then, the loss of gold is seen to be associated with equilibrating income movements incertain cases, or to bring about such movements, and falling incomes and activity may very well causeprices to fall and further promote adjustment.6 Furthermore, the fact that domestic interest rates have risenand that the ‘exchange risk’ is zero, assuming perfect confidence in the stability of the mint par rate ofexchange, will provide an additional speedy but temporary means of stopping the loss of gold through aninduced influx of short-term international capital which is seeking maximum remuneration. This capitalmovement will provide a ‘cushion’, whilst other forces are bringing about a more fundamentalreadjustment.

All this analysis may be reworked for the cases of net imports of gold and associated expansionary incomemovements.

In order to decide what parts of this ‘general’ model are applicable to particular economies it is necessaryto examine their financial, commercial, and political structures before 1914. It will be found that this willlay low any hopes of a general explanation such as that proffered by the old ‘rules of the game’ school,except that changes in incomes provide the basic longer-run equilibrating mechanism. Indeed, not all thecountries participating in the gold standard followed these rules—many ‘limped’—and devices to preventgold exports varied from country to country. Again, actual adjustment processes varied: a creditor countryfound adjustment to a loss of gold easier than a debtor, for it was (and is) always easier to refrain fromlending abroad or to recall funds from use abroad than to increase foreign borrowings to staunch the drainof gold. Indeed, differences in structure and nature of a country’s economy are crucial, when attempting toassess why the gold standard worked so successfully for some countries and failed for others, and may besummarized briefly. (1) A creditor or a debtor country. (2) An industrial or a primary-producing country. In

THE WORKING OF THE GOLD STANDARD BEFORE 1914 117

this case it is of vital importance to realize the differences in export price formation—on a cost basis, ordetermined in the world commodity markets. (3) Whether domestic investment activity or export receiptswas the principal determinant of fluctuations in national income. (4) The size of the economy relative to theworld markets. (5) The size of the marginal propensity to import, for on this would depend the speed, orotherwise, of balance-of-payments adjustments in response to income changes. (6) The financialorganization of the country. Where was the banking system based? (7) The political framework and socialsystem. Were any particular economic interests (e.g., exporters or landowners or rentiers) dominantpolitically? Was there any past history of depreciation? (8) Gold-mining or non-gold-mining countries.

This list does not pretend to be exhaustive, but it does emphasize the need for examining the positions ofparticular countries before making any wide theoretical generalizations.

*In this period an efflux of gold from Britain was stopped first by short-term international capital

movements, which had been induced by the Bank of England’s actions, and secondly by longer-runequilibrating changes in imports, exports and long-term loans, which occurred as a result of the automaticand induced forces associated above with a loss of gold. The short-term capital movements respondedspeedily and provided a ‘cushion’ whilst the latter forces, which needed time to make their effect felt,gradually changed items in the balance of payments so that monetary pressure might be relaxed without thedanger of a renewal of the gold loss. This adjustment mechanism will now be considered in more detail.

When confronted with a net export of gold (which it desired to stop) the Bank of England’s main objectwas to lever up the London market rate of discount in order to bring about equilibrating international capitalflows to London, and indeed quick action was necessary because of low gold holdings (these wereunremunerative and disliked by profit-seeking British bankers) as compared with British internationaltransactions. The main weapon used was an increase in Bank Rate, together with measures to make it‘effective’ by reducing the available pool of short-term credit in London. However, on occasion devicesother than Bank Rate were employed if the Bank of England did not want internal trade to be affected, butonly international capital transactions.7

A rise in the market rate of discount had a speedy effect on the exchanges by bringing short-term funds intoLondon, partly because of the complete confidence in the stability of the link between sterling and gold andpartly because of institutional reasons peculiar to Britain at this period. For Britain was a short-term creditor—to the tune of £150–200 million in 1909 according to Hartley Withers’s estimates—and higher Bank Rate,when effective, speedily induced the repatriation of British funds employed abroad to the (now) morelucrative home uses.8 This unique technical position of London provides much of the explanation of theefficacy of Bank Rate in speedily staunching gold losses before 1914. Secondly, foreign-owned sterlingbalances were provided with an added inducement to stay, and fresh foreign money was attracted, moreespecially as the exchange risk was zero with sterling at the gold export point. Furthermore, the demand forforeign currency was curtailed as borrowing in London became more expensive and difficult, whether short-term loans, raised by discounting finance bills, or long-term issues on the Stock Exchange,9 whilst someforeign bills were diverted to less expensive money markets for discounting. Thus high market rates ofdiscount by affecting international capital movements served to float the exchanges off the gold exportpoint and to staunch the efflux of gold.

It was indeed important that the main world gold market was located in the same country and controlledby the same financial organization that also regulated these international capital funds. Thus London could,by extending or restricting short-term credits, considerably enhance or restrict the ability of non-residents tobuy and ship gold abroad. In other words, such operations enabled London to retain more of the new-minedgold passing through and, if these capital movements shifted the exchanges to the gold import point, to


attract gold from other monetary centres. However, it was a feature of the gold standard before 1914 thatthis latter movement of gold was never ‘automatic’: for when the German and French exchanges were attheir gold export points, on occasion no gold moved because foreign central banks dissuaded bankers byone means or another.10 Ultimately, though, gold did move from Paris and Berlin, perhaps through fear ofwhat might happen if Bank Rate had to go yet higher.

It should be pointed out that international capital movements of this nature did not form a permanent flow,but were equivalent to ‘stock’ transactions, being temporary and very likely to be reversed as soon as theinterest differential disappeared. This mechanism, however, would eliminate gold losses which had arisenfrom excessive speculative short-term capital movements (too many ‘finance’ bills). The longer-run forceswhich provided a more basic cure for the loss of gold will now be considered and included here will be adiscussion of the internal effects of the gold loss and higher interest rates. Much of what follows willbe rather tentative in the sense that detailed empirical studies of particular crises would need to be made tocome to any definite conclusions. Nevertheless, I believe that the underlying theoretical structure andapproach would prove fruitful for such studies and would stand justified.11

Why, then, did the sterling exchanges sag to the gold export point? Let us consider the case where exportvalues fell relatively to imports, other items in the balance of payments remaining constant, because of adecline in foreign buying. (It is assumed that there was a positive marginal propensity to save and that therewere lags in the income process so that a fall in exports did not bring an equivalent fall in importsimmediately and thus eliminate the gold loss.) Besides bringing a gold loss, the fall in exports caused incomes,activity and import purchases to fall, thereby alleviating the strain on the balance of payments, the more soif falling British demand for primary produce reduced their prices so that import payments fell on this countalso. If this decline in British activity induced a fall in domestic investment, then incomes and imports werereduced still more.

Furthermore, the loss of gold reduced the supply of money, the more so if the gold exported had beenpaid for by cheque on a joint-stock bank and thus a contraction of bank deposits had been set in motion.Secondly, the increase in Bank Rate brought increases in domestic interest rates conventionally tied to it, aswell as the rise in the London market rate of discount. Borrowing, besides being more expensive, became moredifficult, so that domestic expenditure was checked and import purchases were reduced—provided that themonetary stringency lasted long enough. Furthermore, a sustained spell of stringency affected the ability tohold stocks of primary produce, so that stock liquidation or the lack of ability to buy brought falling primaryproduct prices, thereby reducing import payments and reinforcing the effects on import prices noted above.

In so far as in such a crisis incomes and activity declined the internal reflux of sovereigns from domestictransactions use helped to swell the Bank of England’s gold reserves. But without a decline in activity thereflux in response to interest-rate changes alone was slight.12

In this case export prices also tended to fall both because of falling activity in Britain and because of thefalling prices of raw materials (e.g., raw cotton prices and their influence on textile prices). This would onlyhelp adjustment on those occasions when the foreign price elasticity of demand was greater than one. Lastly,if the prices of home goods fell relatively to imported goods there would be some expenditure-switchingeffects which helped to promote adjustment. However, these price effects, when they occurred, were ofminor importance as compared with the automatic income effects. It will be noticed in this outline that wehave neglected the repercussions of variations in British activity and import purchases on the rest of theworld and the reaction on purchases of British exports. In any detailed case study it would be improper toneglect such reactions, but the intention of this section is to sketch out certain adjustment forces which havenot received enough attention in the past.


If gold was lost because of an increase in import purchases, other balance-of-payments items remainingthe same, two cases must be distinguished: (a) rising domestic activity and expenditure (because of adomestic investment boom) which brought rising imports; (b) a switch in expenditure from home-producedgoods to imports, money incomes initially remaining the same.13 In the first case the only ‘automatic’effects were those resulting from the influence of the loss of gold on the quantity of money, whilst the‘induced’ monetary effects also helped to promote balance-of-payments adjustment as noted above. In thesecond case incomes and activity automatically fell as expenditure was switched away from home goods,thereby reducing import purchases, to which force the monetary effects were added to promote adjustment.

Consider now the case where gold was exported (net) because an increase occurred in the flow of fundswhich overseas borrowers were seeking to transfer (in the form of sterling bills or gold) to finance domesticexpenditure in their countries, and the exchanges sagged to the gold export point.14 Now the spending ofsuch funds in the borrowing countries raised their incomes and their purchases of imports, particularly ofconsumption goods. Since Britain was an important supplier of such goods, this increased activity abroadeventually increased the demand for British exports and provided an offsetting force to adjust the balance ofpayments after short-term capital movements had provided a temporary cushion.15 To this generaladjustment force may be added the automatic and induced monetary effects of the gold loss. Again, if goldwas lost because more funds were transferred abroad from a given volume of overseas borrowing and lessspent directly on British capital goods for export, then incomes and import purchases declined in Britain;likewise if more was lent to overseas countries at the expense of domestic spending.

It was very important for her operation of the gold standard pre-1914 that ex ante British foreign lending(i.e., overseas issues in London or undistributed profits retained abroad) generated increases in Britishexports relative to imports (either by increased sales of investment goods or by induced sales ofconsumption goods as overseas activity rose) and so facilitated its transfer without balance-of-paymentsdisruption and prolonged gold loss.16 Such a process would be facilitated if the pressure of overseas demandturned the terms of trade in Britain’s favour by pushing up export prices.

Why, however, in these periods of active lending abroad did rising export values not bring roughly similarincreases in imports through rising incomes, so that the tendency of exports to rise relatively to imports andso to bring an increased current account surplus would have been eliminated? First, a sustained increase inexports per period will only bring eventually, ceteris paribus, an equivalent increase in imports per period ifthe marginal propensity to save is zero—a condition certainly untrue for Britain at that time. Secondly,although home and foreign investment tended to move together in the short run, nevertheless the longer-runtendency was for increased foreign investment to take place at the expense of home investment. If the seriesfor these are smoothed with nine-year moving averages, it is clear that the long-run waves of foreigninvestment have their peaks opposite the troughs in the waves of home investment, so that when foreigninvestment was relatively high, home investment was low, and vice versa. Thus the income-generatinginfluence of increased export sales was offset by the decline in home investment—so far as these underlyingtrends are concerned. Again, British lenders, besides switching from home to foreign investment, mightperhaps have been tempted to decrease the proportion of their incomes spent on consumption goods so thatthey could employ their increased savings to purchase the now attractive foreign securities being offered.17

All these factors would militate against the rise in imports which otherwise might have been expected fromrising export values.

It is possible thus to explain why these bursts of overseas investment by Britain were managed with solittle disruption to the British balance of payments and her adherence to the gold standard. The necessary(increased) current-account surplus was created by inducing directly and indirectly increased export salesrelative to import values, and thus the danger of increased foreign lending leading to a severe loss of gold


averted. However, if British foreign lending had not influenced export values, as in the French case,18 thenless might have been heard about the smooth operation of the gold standard in this period.

In summary, then, apart from the transfer of short-term deposits and the ‘domestic investment boom’cases the other factors causing balance-of-payments disequilibria and the loss of gold brought equilibratingtendencies through income movements either in Britain or in borrowing countries, whilst present in all caseswere the automatic and induced monetary effects. The speedy staunching of gold losses from Britain is tobe attributed partly to the peculiar institutional position of Britain which made for rapid changes in themovement of short-term funds and partly to the income forces (automatic in many cases) which were set inmotion and brought about a more permanent balance-of-payments adjustment so that Bank Rate could thenbe relaxed. The automatic and induced monetary effects of the gold loss, to which earlier writers assignedmuch importance in explaining balance-of-payments re-adjustment in the longer run, are thus seen to beonly part of the adjustment mechanism. (The above discussion has concentrated on the export of gold, butthe arguments may be reworked for the import of gold.)

In conclusion of this section it should be pointed out that not all countries and territories ‘took gold’ whenthey had a balance-of-payments surplus with Britain. Those in which the ‘Anglo’ banks were of paramountimportance increased their sterling holdings or ‘London balances’ in such cases—the embryo of the sterlingarea. Balance-of-payments adjustment was thus facilitated for Britain when it meant in some cases not theloss of gold, but the transfer of bank deposits in London from domestic to overseas ownership.

*Argentina, of vast land area and with various climates and vegetations, and sparsely populated, advanced

rapidly as a primary producer in the nineteenth century as transport barriers were removed. The federalconstitution, adopted in 1862, resembled the American, but gave much more power to the President, whilstthe dominant political group was the conservative landowning (and export-producing) oligarchy, whose rulewas maintained until 1910 either by fraud or by force at the ‘free’ elections. The fact that power was seatedhere will enable us to understand certain economic actions of the national governments.

Export values expanded sevenfold within the period 1880–1913, and their composition changed sharply,the main exports in 1880 being wool, fleeces, and hides, whilst after 1890 grains (i.e., wheat, maize,linseed) rose in importance. The twentieth century saw the decline of sheep farming and the rapid rise ofcattle ranching and export of frozen beef. It is important to note that Argentina was a relatively smallproducer of these products, whose prices were determined in centralized commodity markets, and thus shehad little or no control over her export prices. Additionally, Argentina was heavily dependent on export salesfor her well-being, since roughly half of her primary produce was exported. Imports, which expanded sixtimes in value, were composed of food, drink, tobacco, manufactures of all kinds, fuel, machinery and rails,whilst their prices were largely determined by suppliers, since Argentina was a relatively small part of theworld market for these products. Moreover, since Argentine production comprised exportable goods mainlyand little import-competing industry existed, there was little substitution between home-produced andimported goods.

Foreign investment by European centres—especially London—together with immigration, from Italy andSpain particularly, had brought about this expansion with the railway as the basic factor which hadpermitted this ‘export-biased’ development to take place. Other important destinations of foreign fundswere governmental borrowings, public utilities, land mortgage bonds, tramways and land companies. Oneimportant consequence of this borrowing must be noted—the foreign debt-service charges incurred had alarge element fixed and payable in gold or sterling.

The domestic currency, the paper peso (a gold peso was an oddity), had a history of depreciation in thenineteenth century, whilst Argentine banking was anarchic, there being no central bank, although in later


years after 1900 the Bank of the Nation tended to be dominant through sheer size. The commercial bankswere sharply divided into international banks, based on other monetary centres, which specialized in foreigntrade and remittance business, and domestic banks, which looked to local business. Argentina, however,hardly belonged to the English monetary area, as much of the banking system was purely local.

In summary, then, Argentina was an economy predominantly dependent on the production of exportablesand the level of foreign-currency receipts for the determining of her income and well-being. Her export andimport prices together with prices of domestically consumed exportables were determined externally, whilstthere was low substitutability between imports and home-produced non-exportables, so that the scope forprice changes to promote balance-of-payments adjustment was severely limited.19 Furthermore, fluctuationsin foreign-currency receipts provided the principal cause not only of fluctuations in income, but also ofbalance-of-payments disequilibria, so that in this setting the question ‘Why did gold move?’ is to beanswered frequently by pointing to changes in export proceeds or in the flow of funds from abroad.Changes in either of these variables brought directly associated income movements and hence equilibratingchanges in import purchases.

In 1881 monetary reforms were enacted in Argentina to remedy the previously chaotic condition of thecurrency; the main provisions were the definition of the new gold peso, the replacement of the existingpaper currencies by a new national paper peso (at par with the gold peso) whose issue by approved bankswas subject to regulation by the national government, and the convertibility of paper pesos into gold. In1883 these were put into operation and Argentina had formally rejoined the gold standard. Yet the internalbanking system, although sizeable, was rudimentary, whilst the foreign-based banks were not dominant nordid they hold the major part of the Argentine gold and foreign-exchange reserves. This situation may becontrasted with the institutional arrangements prevailing in New Zealand and Australia, for example, whereoverseas banks did provide considerable internal facilities and where London branches (or head offices)held the countries’ foreign-exchange reserves in the form of sterling balances. Sanctions on excessive noteissue were weak; indeed the country had a tradition of these being ignored or removed by governments indifficulties, as indeed it had a tradition of depreciated paper money, so that popular confidence in thestability of the paper peso-gold peso link was not high.

Furthermore, the dominant political interests, the export producers and landowners, were not adverselyaffected by a depreciating exchange; rather a depreciating exchange shifted the distribution of a given realincome in their favour whilst an appreciating exchange moved it against them. For, with constant worldprices, depreciation increased the paper peso prices of exportables pari passu, and hence producers’ receiptsin paper currency for a given output. Wage-rates (both urban and rural), however, were sluggish, showingno comparable increase, and the landowners’ mortgage debts were for the most part fixed in terms of papercurrency, so that the gap between their receipts and costs widened by more than the depreciation. On theother hand, an appreciating exchange rate with constant world prices cut the paper prices of exportableproduce pari passu, shifted the income distribution in favour of wage-earners, for wages were ‘sticky’ inpaper currency, and increased the real burden of a given paper peso debt for landowners. Lastly, less moralshame (such as prevailed widely in Britain) was felt at a depreciating exchange. Accordingly, it is notsurprising to find in times of balance-of-payments deficit a bias in favour of a depreciating exchange rate,and in times of persistent surplus a distinct preference for a stable exchange rate rather than an appreciatingrate. Indeed, these effects are of the utmost importance in explaining Argentina’s lapse from the gold standardin 1885 and her decision to rejoin in 1900.

With this background it would be reasonable to expect that a small disturbance in the balance ofpayments, which an older member of the gold standard club could have withstood either from its ownresources or by means of short-term foreign loans, might have more serious repercussions. Any sustained


export of gold would provoke internal speculation, the populace becoming bulls of gold, and thus the note-issuing banks’ gold reserves (in any case none too plentiful) would be subjected to a further drain as goldwas absorbed into private hoards in the expectation of a breakdown of the system and the emergence of agold premium.

Indeed, as early as January 1885 a gold premium had appeared despite the attempts of the issuing banksto preserve paper convertibility at par. In their attempts the Banco Nacional and the Bank of the Province ofBuenos Aires had lost 77 million pesos worth of gold and foreign exchange, of which the major share,perhaps 50 million pesos, was absorbed internally, the rest internationally. The international drain in 1884 hadarisen because the proceeds of foreign borrowing declined whilst imports and foreign debt-service chargeshad increased and exports barely expanded—these items giving rise to a deficit of 14 million pesos, ascompared with a surplus of 7 million in 1883. Doubtless this drain encouraged some domestic speculativehoarding of gold, but this is not the whole story, for as early as June 1884 the Bank of the Province ofBuenos Aires suspended specie payments. This would seem early for international forces to have made theirimpact; rather, one suspects, it was the victim of some autonomous domestic speculation, perhaps even adeliberate run on gold by special interests who stood to gain by the emergence of a gold premium. Domesticlack of confidence, justified by precedent, was the first cause of Argentina’s withdrawal from the goldstandard. The second was the lack of any institutional mechanism for coping with a balance-of-paymentsdeficit by short-term capital movements and in the longer run by credit contraction.20 This illustrates aprevious theme: a creditor (lending) country can always—easily—bring relief to its balance of payments bylending less abroad; a debtor (borrowing) country will find it hard or even impossible to bring relief byborrowing more—the more so if it has a history of currency depreciation and is thus ‘suspect’internationally. Yet to some extent similar economic conditions prevailed in other primary producingcountries and exchange stability was preserved before 1914. More, indeed, depended on social, political,and moral attitudes to exchange stability, on the structure of society and the political system, and ontradition.

After 1885 exchange-rate stability was forgotten in Argentina until 1898–9, although income movementsstill provided the main balance-of-payments adjustment force.21 The years 1896–9 saw a steady fall in thegold premium as export values rose, both on account of rising production and rising world primary productprices, so that the distribution of incomes moved against the export-producing and landed interests and infavour of those whose incomes were relatively stable in terms of paper currency, the fall in the goldpremium being sufficient to outweigh the rising export prices. Accordingly, the former used theirconsiderable political influence to prevent any further appreciation of the paper peso by enacting in 1899 ameasure to stabilize the exchange of paper notes for gold at the rate of 44 gold pesos to 100 paper, or 227.27paper to 100 gold. Furthermore, these interests claimed that stability would encourage trade and stimulateforeign investment in Argentina, which would become ‘respectable’ again with fixed exchange rates, inaddition preventing too drastic a price-cost deflation which would have occurred otherwise, and whichmight have inhibited the growth of the economy. Nevertheless, their basic reason for rejoining the goldstandard was to prevent any further adverse shift in the distribution of income.

This measure which was put into operation in 1900 provided that the note circulation could only increaseif gold were deposited at an exchange bureau, or Caja, which in exchange would issue notes at the fixedlegal


Table 9.1 Argentina 1900–14

(1) (2) (3) (4) (5) (6) (7)

Exports Imports Londonissues forArgentina

Gold imports(net)

Actual noteissue

Potentialnote issue

(5) as % of(6)

million gold pesos million paper pesos

1900 155 113 7 6.7 295 310 95

1901 168 114 25 0.1 295 310 95

1902 179 103 16 5.5 296 317 93

1903 221 131 26 24.5 380 379 100

1904 264 187 21 23.3 408 423 96

1905 323 205 61 31.7 498 480 104

1906 292 270 63 16.6 527 506 104

1907 296 286 72 20.4 532 545 98

1908 366 273 80 28.6 581 598 97

1909 397 303 109 65.7 685 740 92

1910 373 352 115 35.4 716 817 88

1911 325 367 84 9.8 723 838 87

1912 480 385 101 35.9 800 920 87

1913 484 421 60 3.6 823 938 88

1914 349 272 76 −13.3 803 966 83

Sources: (1) and (2), Extracto Estadistico (1916), p. 3; (3) Author’s computations based on The Economist ‘NewIssues’ sections; (4) Extracto Estadistico, (1916), p. 203; (5) and (6) ibid., p. 297.

Notes: £1=5 gold pesos.Import values are somewhat unreliable, as they resulted from the valuation of volumes at Argentine official customs

values, which were changed infrequently or suddenly with no close relation to variations in world prices ofthe goods in some cases. If Argentine imports from UK are compared with UK exports to Argentina, bothseries display similar fluctuations and are of comparable absolute magnitude. Export values are more reliable,except that frozen meat exports were valued at constant ‘prices’ 1900–14, so that official figures are a littletoo low in total, but by not more than 5 per cent. Despite these difficulties the official figures do give a goodindication of what happened.

Potential note issue is the issue which would have resulted if all gold imports (less quantity of gold absorbed by theConversion Fund of the Bank of the Nation) had been deposited at the Caja in exchange for notes.

ratio. Conversely, the bureau was legally bound to give out gold in exchange for paper currency. SinceArgentina was not a gold producer, international gold movements formed the main determinant of the noteissue (see table 9.1), and secondly of the level of bank deposits and bank credit, as banks’ cash reservesvaried. The institutional framework thus allowed gold exports, if obtained from the Caja or from bankingreserves, to contract the quantity of money; gold imports, unless hoarded privately, to expand the moneysupply (except in so far as the Bank of the Nation tried to offset such forces temporarily). Thus far, then, allwas in accordance with the ‘rules of the game’, but the considerable institutional and structural differencesfrom Britain noted earlier still persisted.

After 1900 Argentina remained a member of the gold standard club until 1914, in these yearsexperiencing considerable growth of production, population, railway length, exports and imports, withinwhich context her successful gold standard adjustments must be studied. Export values, which expanded at


an annual trend rate of 7.5 per cent (her export prices, it is important to note, rose at 3 per cent per annum),provided the main force promoting rising incomes and rising profitability of investment projects. After 1904rising foreign investment and immigration were helpful not merely in bringing income increases as thefunds were spent, but also in expanding the capacity of the economy to produce exportables. Thesemovements are illustrated in table 9.1 where also the uneven response of total imports to these forces maybe noted. However, if imports are split into consumption imports and investment imports (following theofficial subdivision), this may be explained by the resurgence of foreign borrowings. For, on the one hand,rising export proceeds brought rising incomes and hence rising consumption good imports, together withextra imports of investment goods if extra domestic investment was induced; on the other hand, theincreased foreign-investment funds were partly spent directly abroad on investment goods; whilst theremainder was transferred to Argentina to finance local spending, thus increasing incomes and imports ofconsumption goods. The influence of this increased foreign investment after 1904 is most noticeable in thebehaviour of imports expressed as a percentage of exports as depicted in figure 9.1. This ratio fell until 1903as debt-service charges and profits remitted were claiming an extra share of export proceeds; thereafter theratio increased until 1911 under the influence of rising foreign investment, declining after 1911 as foreignborrowings fell. When the ratios of consumption imports and investment imports as a percentage of exportsare considered, it is noticeable how sharply the latter was affected by rising foreign investment, whilst theformer was mildly affected, for foreign funds transferred to Argentina then were of relatively minorimportance as a generator of incomes as compared with export proceeds.

The expansion of foreign-currency receipts, despite the increases in payments on account of growingimport purchases and foreign debt-service payments, brought a net influx of gold into the economy in allyears except 1914, which was either deposited in the Caja (or Conversion Bureau) in exchange for notes oradded to the banking system’s reserves. International movements of gold thus enabled the note issue toexpand at an annual trend rate of 8.8 per cent, whilst bank deposits rose both because of increases in cashreserves and because of declines in the cash/deposit ratios as the economy prospered. On the other hand, inthe second half of 1913 and in 1914 when gold was exported on a considerable scale, the note circulationfell and likewise bank deposits, both because of falling cash reserves and because of rising cash/depositratios. Thus the monetary system served to intensify the influence of rising foreign-currency receipts onincomes by increased liquidity and credit, whilst it aggravated the influence of falling foreign-currencyreceipts, since liquidity and credit were curtailed.

The balance-of-payments adjustment mechanism for Argentina under gold standard conditions may beoutlined as follows. Foreign-currency receipts, it is asserted, provided the principal cause both offluctuations in domestic incomes and of changes in Argentina’s balance-of-payments position, whilstforeign-currency payments were a dependent variable. Fluctuations in receipts were large enough to swampany contrary ‘autonomous’ changes in domestic expenditure—indeed it is to be expected that domesticallyfinanced investment would be affected directly by earlier variations in foreign-currency receipts.22 Thefluctuations in export prices which were world-determined as noted earlier and must be treated as anindependent variable, were quite unconnected with gold movements into or out of Argentina, as thedissimilar behaviour of grain export prices as compared with wool and meat prices clearly indicated (seetable 9.2). Indeed, the scope here for gold movements to influence prices was severely restricted to theprices of home-produced non-exportables, land, and real estate, for import prices were fixed by the foreignsellers. Low elasticities of substitution (a) between imports and non-exportables and (b) betweenexportables and non-exportables rendered slight the influences of any relative price changes, whichinternational gold movements had brought about via changes in domestic credit policy, on domesticpurchases.23 Nevertheless, these changes would help to promote adjustment, assuming given terms of trade.


Balance-of-payments adjustment to the main disequilibrating force (variations in foreign-currencyreceipts) was achieved predominantly by corresponding changes in incomes and in imports (more especiallyconsumption goods, since investment good purchases were linked to for

Table 9.2 Export and import price index numbers for Argentina 1900–14

(1) (2) (3) (4) (5)

Total exports Grain exports Pastoral exports Total imports Terms of trade (1)÷ (4)

1900 100 100 100 100 100

1901 99 108 85 87 114

1902 107 114 97 86 123

1903 97 94 102 88 110

Figure 9.1 Three-year moving averages


(1) (2) (3) (4) (5)

Total exports Grain exports Pastoral exports Total imports Terms of trade (1)÷ (4)

1904 103 103 103 88 117

1905 116 114 118 90 129

1906 121 114 132 96 126

1907 125 123 129 100 125

1908 120 133 101 92 131

1909 138 150 119 89 155

1910 142 148 133 87 146

1911 137 143 128 100 137

1912 137 142 130 108 127

1913 139 135 144 108 129

1914 142 136 152 n.a. n.a.

Source: Ford (1955).

eign borrowing) and in invisibles, e.g., dividends and remittances would be higher in years of high activity;of these, imported consumption goods were the principal pliable item. Finally, movements in incomesaffected slightly the domestic consumption of exportable products and hence the supply of exports in anequilibrating fashion. Furthermore, this process was aided by the response of the banking system. Forchanges in imports lagged behind changes in foreign-currency receipts so that a rise in export values permonth, for example, brought a balance-of-payments surplus and import of gold, which gradually declinedper month as imports rose, disappearing if the marginal propensity to save was zero. Thus income changeswere reinforced by the liquidity and credit changes resulting from the gold movements—indeed the noteissue and circulation would move in a close direct relationship with the level of incomes.

For completeness, it is necessary to consider the responses of this system to changes in importsunconnected with foreign-currency receipt changes. Suppose imports increased on account of a shift intastes, or a rise in domestic investment, whilst foreign-currency receipts remained steady, so that gold wasexported and the money supply contracted. In both cases the liquidity check would operate to reducespending and incomes, whilst in the former case, if more were spent from a given income on imports, lesswould be available to spend on home goods and/thus incomes would be reduced further. Any such changein imports, then—apart from the assertion that it would be of small magnitude in the Argentine setting—contained its own adjustment mechanism to damp its influence down.

It is difficult to verify these mechanisms because of insufficient data (e.g., the lack of full estimates ofcapital influx) and because of the comparative unreliability of some available material (e.g., export andimport valuations). Nevertheless, if first differences of annual figures of exports and London issues forArgentina are compared with actual gold movements (both international and internal), it will be found that:

1 International gold movements were associated with these first differences of the main constituents offoreign-currency receipts—the curves of the net import of gold and the first differences of exports, forexample, show broadly similar movements.

2 The supply of gold from private hoards varied closely with export first differences—the net supplybeing greater when exports rose. (Net supply from private holdings=excess of absorption of Caja andbanking system over net gold imports for any time period. A negative value indicates an increase in privategold holdings.)


3 The supply of gold from banking reserves (or alternatively the absorption of gold by banks) was not atall closely associated with export first differences, or in other words changes in banks’ visible holdings werevery much a residual, balancing the potential supply of gold for note issue and the actual demand for notes.

These movements are illustrated in figure 9.2, and in table 9.3 with correlation coefficients.The potential supply of gold per time period available for changing the note issue thus depended largely

on the behaviour of foreign-currency

Table 9.3 Correlation coefficients

(a) Net gold imports and export firstdifferences (15 observations)

+0.54

Figure 9.2 First differences and gold movements


(b) Net gold imports and export plusLondon issues first differences (15observations)

+0.61

(c) Net gold supply from private holdingsand export first differences (11observations)

+0.85

(d) Net gold supply from bankingreserves and export first differences(11 observations)

−0.53

Note: The t-test indicates that in cases (a) and (b) correlation coefficients above 0.59 are significant at the 1 per centlevel and above 0.45 at the 5 per cent level; in cases (c) and (d) those above 0.68 at the 1 per cent level andabove 0.52 at the 5 per cent level.

receipts and the confidence (or otherwise) which this generated in private holders of gold so thatfluctuations in the potential note issue would be closely associated with fluctuations in foreign-currencyreceipts and national income. Furthermore, the predominant demand for notes in Argentina was fortransactions purposes, so that the actual issue would also be closely linked to movements in national income.Accordingly, it would be reasonable to expect average note issue per year (average. of figures for end-March, June, September, December) to be a reliable indicator of the behaviour of national income. If valuesfor consumption imports and average note issue are compared for 1900–14, a correlation coefficient of +0.95 results, whilst the correlation coefficients for deviations from linear and exponential trends are +0.79 and+0.83 respectively. (With thirteen degrees of freedom the t-test indicates that coefficients above +0.76 aresignificant at the 0.1 per cent level.) Although these coefficients, it is suspected, would have been higher ifArgentine official customs valuations had not varied so much in the years 1904–6, nevertheless the resultsdo demonstrate the close connection between national income and consumption import purchases.

Argentina had rejoined the gold standard in 1900 largely to stop a shift in income distribution which wasunpleasant to the politically dominant landed and export-producing groups, and had employed a systemwhereby international gold movements affected the note issue. The adjustment mechanism to changes in thebalance of payments (reflected in gold movements) consisted of consequent income movements, whichwere reinforced by the monetary system, and of their effects on import purchases and remittances. Inaddition, in years when foreign investment was changing, direct purchases of investment goods abroadchanged likewise. Short-term international capital movements, which were so responsive to Bank Rate andmarket rate of discount changes in the British case, were largely absent.

Why was this system successful in preserving exchange-rate stability until 1914, whilst the system of1884 collapsed within a year? First, the system never met with the serious test of a sharp efflux of gold asforeign-currency receipts declined relatively to payments until the second half of 1913 and afterwards,24

whilst it was in the interests of the export-producing and landed groups to see that the system worked. After1902 there were no runs on gold reserves for private speculative reasons until early 1914. Indeed the Cajasystem, which contemporary commentators dubbed ‘fair weather’, met ‘fair weather’ conditions until late1913. For foreign-currency receipts grew steadily as export values and foreign investment rose, and heremust be emphasized the importance of rising export prices, which enhanced the value of the growing outputof these products, cut the ‘produce’ value of fixed interest debt, and increased the attractiveness ofArgentina to foreign investors. So rapid was the growth of the economy that any temporary setbacks toexport values through poor harvests, or to foreign loans, meant a slackening in the rate of growth ofincomes, consumption, and imports rather than any sustained downturn.


The difficulties which a primary producing country, heavily dependent on export sales for its prosperity,may experience in maintaining exchange stability, are well known, and in Argentina were aggravated by herinternational debtor status; a debtor has service charges, which for Argentina had a large core fixed in goldor sterling and formed a considerable item in her foreign-currency payments.25 Further, in times of worlddepression or adversity there was a tendency for short-term funds to move back to creditor countries and awayfrom debtors, irrespective of the latter’s interest rates—a movement which affected Argentina especiallykeenly because of the lack of international confidence in the peso. Thus no short-term capital movementsserved to soften the export of gold, which had to be staunched speedily by income movements if exchangestability was to be preserved.

Again, Argentina illustrates the weakness for an ‘export’ economy where the quantity of money isdetermined by international gold and foreign-currency movements, in that booms and slumps generallyarose because of changes in foreign-currency receipts which also brought imports or exports of gold as wellso that the initial income movements were exaggerated by changes in liquidity. It was thus difficult toaccumulate a sufficient gold reserve during a boom to cope with bad times—apart from the itching palm ofsome early Argentine governments for idle gold! Furthermore, the fact that in years of depression the initialcontractive effects were intensified by monetary factors tended to alienate support for the gold standardsystem. This is quite different from the cases of some other countries where booms and slumps originatedfor domestic reasons and met with a fairly stable monetary supply so that rising or falling interest ratestended to mitigate these initial income movements. Here booms were associated with adverse balances ofpayments, slumps with favourable balances so that the monetary policy dictated by gold movements tendedto lessen (not enhance) income fluctuations and the gold standard system was more acceptable than in theformer case.

These economic difficulties which might have sufficed in themselves to render adherence to the goldstandard impossible, were supplemented by political and social factors which in the last resort proveddecisive. The domestic convertibility of notes for gold, which was the prime object for Britain and certainother economies and from which the international gold standard sprang, was not such a point of honour andmorality. Other primary producers, such as Australia and New Zealand, maintained exchange-rate stability,which is explained by different administrative and political systems with different social structures, and bytheir banking systems being based on London. However, in Argentina, aided by the particular economic andpolitical structure, the landed and export-producing oligarchy willingly abandoned or adopted the goldstandard system whenever it was to their benefit and profit.

*Much of the discussion above has been concerned with short-run conditions and adjustment of each

economy, and frequently each has been considered in isolation. In this section certain long-run features andinterconnections between these economies will be noted briefly. Although the successful working of thissterling system—for that was really what the pre1914 gold standard amounted to—did depend on thesupreme international confidence in the pound, the particular institutional structure of international bankingand finance which was centred in London, the acknowledged dominance of the Bank of England,nevertheless particular economic relationships played an important role in ensuring that, for Britain at least,maladjustments in the flows of international payments and receipts were never too large or too persistent toswamp the institutional structure.

When the stability of this sterling system as a whole is considered, it is important, first, to note thatBritain provided a steadily growing market for the increasing supplies of primary products from thedeveloping countries. Despite considerable fluctuations in British foreign investment and exports, the broadtrend of activity showed steady growth, for (as noted earlier) in periods when the former variables were high


so far as underlying trend values are concerned, domestic British investment was low. and vice versa. so thatone offset somewhat variations in the other. Nevertheless, an element of instability for the primaryproducers especially remained because they had little influence over the world prices of their produce andhence the terms of trade of primary products for manufactures. Indeed, a sharp fall in their export pricesseriously affected their well-being and could strain their adherence to stable exchange rates, especially iftheir foreign debt-service charges had a large element of fixed-interest payments. However, in general, thegrowing, assured, market afforded to these producers by Britain enabled them to earn sufficiently increasedsupplies of sterling to meet their debt-service charges and to increase their imports and well-being.26

Important, secondly, was the interlinking of the lending and borrowing countries and the responsivenessof trade flows to capital movements—especially for Britain. For British foreign investment (e.g., by meansof new overseas issues on the London Stock Exchange) at times increased considerably and was transferred(in real terms eventually) without any serious strain on the balance of payments because of the influence ofex ante overseas investment directly and indirectly on export sales, whilst any temporary strain, as the netproceeds from borrowing were transferred out of Britain in the form of gold or sterling bills, was speedilyeased by the reflux of British funds employed abroad. Furthermore, when British investment abroaddeclined, exports likewise declined, reducing the British current-account surplus, so that sterling did notbecome scarcer. Here the long-run tendency for the peaks in home investment to occur with troughs inforeign investment so that domestic activity was more stable in its growth than exports, was of especialimportance in bringing about a rising demand for imports, and helping the primary producers whoseproduction of exportables was expanding as investment projects financed from the last wave of foreigninvestment from Britain matured.

The following crude pattern of economic relationships between Britain and the primary producers whowere on the periphery of the gold standard system suggests itself. British loans to overseas countriesincreased their purchases from Britain and their debt-service payments so that on the one hand they did notpermanently gain much gold, nor did Britain lose much gold through the loan-transfer transactions.Secondly, when the investment projects matured, their production of exportables was expanded, for whichthere was a ready market in Britain. The primary producers’ exports rose so that they could pay their foreigndebt-service charges and dividend remittances, and purchase more imports, whilst because of their growingprosperity British exports expanded. Such trade flows thus contributed to the economic growth of bothpartners and their mutual welfare, besides providing a basis for stability.

This idyllic scheme, although perhaps realized pretty well in the long run, did give rise to particular short-run difficulties—especially for the primary producers. For, in the upsurge of foreign lending if the sumstransferred were temporarily in excess of Britain’s current-account surplus and caused the loss of gold,Bank Rate adjustments and the reflux of British short-term funds brought speedy relief. On the other hand,the decline in foreign lending often occurred before the investment projects had been completed and hadexpanded exportable production so that balance-of-payments difficulties faced the borrowers because of theadded burden of extra debt-service charges on an as yet unexpanded production—especially if the ratio ofdebt-service charges/exports had risen sharply.27 In this setting of crisis (and possible British fears aboutthese countries’ exchange stability or possible default on interest payments) it is dubious whether the short-term pool of credit in London helped such economies to the extent that has sometimes been suggested.Rather they were left to adjust themselves through income movements, which affected their importpurchases, whether they remained on the gold standard or perforce embraced a system of flexible exchangerates.

Again, variations in the prices of primary products and in the terms of trade tended to affect these economiesmore sharply than Britain, for they were more heavily dependent on international trade in most cases.28


Indeed, in the short run sharp fluctuations impeded or enhanced development, and brought in the formercase balance-of-payments strain. For example, falling primary product prices between 1889 and 1896prevented Argentine export values from rising, despite a great expansion in the volume of exportableproduction which foreign investment had facilitated; after 1896, however, rising prices enhanced exportvalues, facilitated greatly her return to the gold standard in 1900 and her subsequent adherence until 1914.

In conclusion, mention must be made of one important role played by gold discoveries—they served tokeep the whole gold standard system liquid enough to support a growing volume of internationaltransactions and to provide the necessary underpinning of sterling.

Notes

1 Whale (1937).2 Whale (1937).3 These may be reworked in the case of an import of gold.4 If more exportables were consumed at home at the expense of domestic savings, no contractionary income

movements would result immediately—the equilibrating forces would be limited to the effects of the loss of goldon money supply, etc.. as in the case of 1. If other countries employ the proceeds of extra loans from the countryin question to buy extra goods from her (i.e., increased exports) then the loss of gold and the decline in incomeswill both be arrested, perhaps after a time-lag.

5 This rise will be moderated or even eliminated in the cases where activity falls.6 Assuming that the demand elasticities are big enough.7 See Sayers (1936) for an interesting and revealing discussion.8 Cf. Schumpeter (1939), vol. II, p. 673. This credit ‘which was currently turned into cash to be presently

reinvested almost anywhere within the gold area, responded to the Bank’s slightest move very much morepromptly than foreign-owned balances would have done, facilitated great capital transactions, supported foreignbusiness, mitigated domestic stringencies. Because of its presence, tightening the open market—raising open-market rates—not only regulated, but by drawing gold, eased situations.’

9 For example, in 1913 the monetary stringencies in London led to Argentine railway companies postponing newissues until the more propitious conditions of early 1914.

10 For example, Clare (1931), p. 75, referring to the Bank of England and its measures to prevent the export of gold:‘If 5 per cent fails to arrest the export, other measures may have to be resorted to. The Reichsbank, in such acase, would possibly give exporters to understand that they must be prepared to incur the consequences of itsdispleasure if the withdrawals were persisted in. This the Bank (sc. of England) cannot do; nor can it put apremium on gold—like the Bank of France did before the law of June 1928 made France a free gold country.’Compare also Goschen (1905), pp. 117–19, and Withers (1916), pp. 6, 38 (especially).

The Bank of England’s own evidence before the US National Monetary Commission of 1910 is instructive: Q:‘How do you account for the fact that at times a higher bank rate in England fails to attract gold from thecontinent, when lower rates prevail there?’ A: ‘Because there is no gold market on the continent so free as theLondon Market and the continental markets frequently do not release gold for export until the rate in London hasreached a figure which threatens disturbance to their own financial position’ (US National MonetaryCommission, 1910, p. 27).

11 The author’s study of Argentine experience between 1880 and 1914 has convinced him of this rather arrogantclaim.

12 See Goschen (1905), p. 121, referring to domestic gold circulation, ‘You cannot “tap” this immense mass ofsovereigns when you most want gold, except so far as there might be a slight flow towards the centre in times ofvery dear money. Even if the rate of interest were 8 per cent the bulk of the people would not carry less goldabout then than they did before.’

13 If increased import purchases were made at the expense of savings, then the analysis of case (a) is appropriate.


14 It is the net proceeds of a loan transferred abroad in this form which affect the exchanges, not the total proceedsof a loan, for some portion of this never left Britain, being spent directly on capital goods for export to theborrower.

15 It is true that rising export sales would increase British imports as activity increased, but this point needs to bediscussed with reference to long-term trends.

16 For a fuller discussion, see Ford (1958–9), pp. 302–8.17 In the years of high capital export the ratio of net home investment plus realized foreign investment to national

income is high (e.g., in 1872, 15.0 per cent and in 1913, 15.2 per cent, whilst in years of low capital export it islow (e.g., in 1882, 10.4 per cent and in 1896, 9.7 per cent). With a balanced budget, these figures indicate thebehaviour of the ratio of ex post net domestic savings to national income, and may provide some guide to ex antesavings. The estimates are taken from Lenfant (1951), and years of high activity have been chosen for thepurposes of this comparison.

18 See White (1933). Again, if the French had not lent abroad, they might well have disrupted the system byamassing most of the world’s gold.

19 Indeed, in the short run the supply of exportables and the supply of exports (for domestic consumption ofexportables was inelastic with respect to price and income) were both highly inelastic, whereas the supply ofimports may be taken as perfectly elastic.

20 It may be doubted whether the loss of external reserves was allowed to affect the domestic credit position in thesame way as in Australia or New Zealand, where sterling balances formed both the banking system’s cashreserves and the country’s foreign-exchange reserves. Given fairly stable cash/deposit ratios, any loss in foreignexchange would bring about a deflationary banking policy with speed. In the 1870s also, Argentine governmentspreferred to authorize extra issues of paper and subsequent depreciation to the rigours of credit contractionneeded to preserve specie payments in that crisis.

21 For a full treatment of this episode, see Ford (1958).22 For example, domestic investment in the meat trade, whether in freezing plants, in ranches, or in stock, was

conditioned by the behaviour of the export markets for meat. Domestic building operations also were affected bychanges in the balance of payments: on the one hand the prosperity which rising foreign-currency receiptsbrought induced extra building, whilst the increase in the quantity of money in Argentina, following thisexpansion of receipts and the net import of gold, provided extra resources for the banking system to lend andlowered interest rates.

23 It should be noted that rising foreign-currency receipts (besides bringing a gold influx) would also increase thedomestic demand for non-exportables via rising incomes, and so promote some slight equilibrating changes inpurchases if their prices rose. It would indeed be difficult to say whether such prices rose because of such risingincomes, or because of the expansionary effects of the gold influx.

24 Because of lags in the transmission of income effects, foreign-currency payments, although growing, would tendto lag behind growing foreign-currency receipts over a period of years so that each year a favourable balance ofpayments and net import of gold resulted. Once the growth in foreign-currency receipts slackened or ceased,payments would catch up, bringing a less favourable or even adverse balance of payments. Indeed any decreasein receipts would bring a sharp adverse movement in the balance of payments until payments reversed theirprevious trend and declined.

25 In the period 1911–14 service charges amounted to some 35 per cent of export values.26 This is a long-run view, which applies once the investment projects (financed from abroad) have matured and

expanded the production of exportables. In the short run for primary producers balance-of-payments difficulties,which strained adherence to the gold standard, did occur because frequently the debt-service charges wereimmediate (e.g., fixed-interest securities) and bore heavily on an unexpanded output and supply of foreignexchange. Such difficulties did not arise for Britain.

27 Fixed-interest securities, rather than equities, were a feature of British foreign lending, so that service chargeswere immediate. So long as lending continued, these could be met without much strain, but when lendingdeclined, crises often arose. See Ford (1956).


28 Furthermore, even if primary product prices did move against Britain, she did benefit somewhat, for debtorscould pay their service charges and dividend remittances were greater, whilst purchases of British exports rose.

References

Clare, George (1931), A Money Market Primer, 3rd edn, London, E.Wilson.Extracto Estadistico de la Republica Argentina correspondiente al ano, 1915 (1916), Buenos Aires.Ford, A.G. (1955), ‘Export Price Indices for the Argentine Republic, 1881–1914’, Inter-American Economic Affairs

(Autumn).——(1956), ‘Argentina and the Baring Crisis of 1890’, Oxford Economic Papers.——(1958), ‘Flexible Exchange Rates and Argentina, 1885–1900’, Oxford Economic Papers.——(1958–9), ‘The Transfer of British Foreign Lending, 1870–1914’, Economic History Review.Goschen, G.J.G. (1905), Essays and Addresses on Economic Questions, London, Edward Arnold.Lenfant, J.H. (1951), ‘Great Britain’s Capital Formation, 1865–1914’, Economica, New Ser. 18.Sayers, R.S. (1936), Bank of England Operations 1890–1914, London, P.S.King.Schumpeter, Joseph A. (1939), Business Cycles, New York, McGraw-Hill.US National Monetary Commission (1910), Interviews on the Banking Systems of England, Scotland, France,

Germany, Switzerland and Italy, Senate Doc. No. 405, Washington, DC, US Government Printing Office.Whale, P.Barrett (1937), ‘The Working of the Pre-War Gold Standard’, Economica (February).White, H.D. (1933), The French International Accounts 1880–1913, Cambridge, Mass., Harvard University Press.Withers, Hartley (1916), Money Changing, 2nd edn, London, John Murray.


10The gold standard since Alec Ford

Barry Eichengreen*

Alec Ford’s The Gold Standard 1880–1914: Britain and Argentina is one of a handful of classics on thegold standard written in the twentieth century. The book made three contributions. First, it elaborated a newmodel of the balance-of-payments adjustment mechanism. Analysing the experience of Britain, Ford arguedthat adjustment worked through different channels than the price-specie flow mechanism emphasized byHume or the interest-rate-induced capital flows emphasized by Whale.1 Ford’s Keynesian modelhighlighted the tendency of gold outflows to raise interest rates, lower domestic demand, and restoreexternal balance through the reduction of output, employment and import demand.2 Relative pricemovements, or changes in interest rates induced by central banks playing by the rules of the game, might aidadjustment but their role was subsidiary.

The second contribution of the book was to contrast the very different nature of gold standard experiencein general, and of the adjustment mechanism in particular, in less-developed primary-producing countries.Analysing the experience of Argentina, Ford argued that the record of stability under the gold standard wasless satisfactory at the periphery than the centre. As at the centre, at the periphery adjustment workedthrough changes in income and demand. An inflow of long-term capital, for example, tended to stimulatedemand, increasing imports and thereby tending to restore balance to the external accounts. But at theperiphery exceptionally large fluctuations in income were required. Given the underdeveloped state ofdomestic financial markets and even the absence of a central bank, there was little scope for interest ratechanges to induce accommodating short-term capital flows. The commodity prices facing primary-producing countries were dictated by world commodity markets. Hence, the burden of adjustment fellsquarely on changes in demand, often brought about by fluctuations in employment.

Ford’s third contribution was to indicate how countries of both types fit together in an equilibrium system.He showed how the operation of markets throughout the international economy facilitated adjustment tobalance-of-payments disturbances. An increase in long-term foreign lending by London, for example,automatically induced an increase in the recipient country’s demand for commodity imports from Britain.The fall in British output and employment that otherwise would be caused by the shift from domestic toforeign investment was attenuated by the export boom. But Ford also emphasized the extent to which theoperation of these markets depended on ‘the social and political environments, as well as economic, of thecountries concerned’.3 Specifically, the maintenance of international equilibrium depended on the fortuitouslyrapid growth of the world economy, which subdued domestic devaluation lobbies. It depended on the

* S.N.Broadberry and N.F.R.Crafts (eds), Britain in the International Economy 1870–1939, Cambridge: CambridgeUniversity Press, pp. 49–79. Copyright © 1992 Cambridge University Press. Reprinted with the permission ofCambridge University Press.

propensity of capital-importing nations to use foreign funds to finance investment rather than consumption.It depended on the fact that the leading capital-exporting nation was also an exporter of capital goods.

Ford’s model, with its prominent Keynesian features, was a product of its time. The subsequent quarter ofa century has seen a radical change of fashion in macroeconomics, and a recasting of the literature on the goldstandard. The rise of monetarism has produced monetarist models of the gold standard. The rise of efficient-markets models of the macroeconomy has led to efficient-markets models of the gold standard. Theliterature on rules versus discretion and on the credibility and time consistency of macroeconomic policyhas led recent authors to view the gold standard in this light.

In this chapter I offer a synthesis of the literature on how the gold standard worked. The model turns outto be very much in the spirit of the work of Alec Ford.

Notes on the working of the gold standard before 19144

In The Gold Standard 1880–1914, Ford analysed time series data for Britain and Argentina, withoutestimating a formal structural model of their balances of payments. Using time-series plots, he describedfirst differences of exports, imports, foreign lending, net gold movements and the like. He then calculatedcorrelation coefficients for pairs of these variables, and used them to inform his story of how the goldstandard worked.

In this section, I proceed in similar fashion. Rather than specifying and estimating a structural model, Ibegin with significantly less structured data analysis, which considers correlations among the relevantvariables in the time domain. The technique is a straightforward extension of Ford’s methodology—namely, vector autoregression. I regress each element of a vector of endogenous variables on lagged valuesof itself and lagged values of the other variables. The estimated coefficients can be used to summarize thecorrelations between a variable of interest and lagged values of the other variables. Impulse-responsefunctions, in which the error term in one of these equations is perturbed and the dynamic response of thesystem is traced out, can be used to analyse the consequences of various disturbances.

Vector autoregression reveals little about economic or historical causality. It is unlikely to tell us muchabout the ultimate causes of balance-of-payments problems under the gold standard.5 It simply summarizesthe relevant correlations in a digestible form. Simulations using the estimated coefficients may help us tonarrow the range of plausible models consistent with the facts, although there may be more than one modelwhich satisfies this consistency criterion. Ultimately, other economic or historical evidence will be requiredto arrive at a satisfying interpretation.

Britain’s experience is only one part of the larger gold standard story, as Ford’s work makes clear. But ananalysis of international adjustment by Britain, the centre country of the international gold standard, is alogical starting point for any such study. Here I limit the statistical analysis to data for Britain.

The six variables upon which I concentrate are the volume of British exports, the volume of Britishimports, Britain’s international terms of trade (export prices relative to import prices), the Bank ofEngland’s discount rate, gold reserves, and the volume of British foreign lending. This is the minimal list ofvariables necessary to capture the most important disturbances to the British balance-of-payments and themost important channels through which balance-of-payments equilibrium was restored. The foreign lendingvariable is new issues for overseas borrowers (calls) on the London capital market, from Simon (1968),transformed to constant prices using the GNP deflator. Gold reserves (gold in the Issue Department of the Bankof England) are drawn from Mitchell and Deane (1962). The Bank of England’s discount rate is drawn fromPalgrave (1903), as supplemented by the Economist Magazine. The remaining variables are drawn fromFeinstein’s (1972) national income accounts. All variables are for the period 1871–1913.


Statistical analysis

Figures 10.1 to 10.5 juxtapose the gold flows against the other five variables. Figure 10.1 shows the well-known tendency of the Bank of England to adjust Bank Rate in response to changes in its gold reserve.Generally, as the reserve rises Bank Rate is reduced, although there are exceptions to the rule. Figure 10.2juxtaposes the gold reserve and the real value of new issues on behalf of overseas borrowers. Here, too,there is some evidence of the expected inverse correlation. Generally speaking, as the volume of overseaslending rises, the gold reserve tends to fall.

Interpretation of the remaining three figures is less straightforward. Figures 10.3 and 10.4 display thechange in the volume of British imports and exports along with the gold reserve. (Since import and exportvolumes trend strongly upward over the period, first differences are displayed in figures 10.3 and 10.4.)Although there is some rather weak evidence in figure 10.3 of positive comovements between imports andgold reserves (as if the accumulation of reserves allowed an expansion of money supply which fueled thedemand for foreign goods), a notable feature of figure 10.3 is that the upward shift in the gold reserve in the1890s was not accompanied by a comparable shift in the growth of imports. Figure 10.4 makes a similarimpression: export volumes and gold reserves move together (with notable exceptions like the 1907 panic),but once again there is little noticeable upward shift in export growth in the 1890s.

Finally, figure 10.5 juxtaposes the terms of trade against the Bank of England’s gold reserves. There issome evidence of sympathetic movements in the two series, albeit with long and variable lags. In the 1870sand 1880s, movements in gold seem to lag behind the terms of trade, while in the 1890s and 1900s, theopposite, if anything, is true.

Table 10.1 shows correlation coefficients for the six variables, all expressed in level form. The strongestcorrelation is between exports and imports. The important question is whether fluctuations in onecomponent of the trade balance responded to fluctuations in the other, or whether both respondedcontemporaneously to movements in other variables. Both exports and imports appear to have been tightlylinked to the level of the gold reserve. There is evidence of a strong contemporaneous correlation between

Figure 10.1 Bank rate and gold reserve, 1872–1914

THE GOLD STANDARD SINCE ALEC FORD 137

foreign lending and exports, as noted previously by Ford, raising anew the question of the direction ofcausality.

Table 10.2 reports the results of regressing each of these variables on three own lags and three lags ofeach of the other variables.6 The sample is 1874–1913 to allow for lags. The figures reported are theconfidence

Figure 10.2 New issues and gold reserve, 1872–1914

Figure 10.3 Change in import volume and gold reserve, 1872–1914


Table 10.1 Correlation coefficients

New issues Exports Imports Terms of trade Gold reserve

Bank rate 0.291 0.295 0.224 0.315 −0.114

New issues — 0.701 0.535 0.084 0.293

Exports — — 0.961 0.332 0.723

Figure 10.4 Change in export volume and gold reserve, 1872–1914

Figure 10.5 Terms of trade and gold reserve, 1872–1914


New issues Exports Imports Terms of trade Gold reserve

Imports — — — 0.462 0.801

Terms of trade — — — — 0.432

Note: Sample period for all variables is 1871–1914.

intervals implied by F-statistics testing the joint significance of lagged values of an explanatory variable.The null hypothesis is that lagged values of a variable are jointly unrelated to current values of thedependent variable. An entry of 0.05, for example, indicates that the null hypothesis of no association canbe rejected at the 95 per cent confidence level.

The first column summarizes influences on the Bank of England’s discount rate. By far the mostimportant determinant of changes in Bank Rate (other than its own lags) is changes in gold in the IssueDepartment, plausibly enough given the Directors’ concern with changes in gold reserves. There is someevidence that Bank Rate was raised in response to increases in the volume of imports and foreign lending. Theobvious

Table 10.2 Statistical significance of independent variables (confidence levels at which F-statistics indicate nullhypothesis can be rejected)

Dependent variable

Independent variable Bank rate New issues Exports Imports Terms of trade Gold reserve

Bank rate 0.049 0.810 0.334 0.403 0.281 0.766

New issues 0.051 0.019 0.143 0.242 0.714 0.518

Exports 0.448 0.114 0.006 0.913 0.584 0.482

Imports 0.049 0.043 0.098 0.044 0.272 0.456

Terms of trade 0.078 0.554 0.340 0.241 0.015 0.021

Gold reserve 0.002 0.616 0.712 0.781 0.161 0.008

interpretation is that these variables were viewed as leading indicators of balance of payments trends andhence of future gold flows.

The second column, which summarizes influences on the volume of overseas lending, paints a differentpicture. Changes in Bank Rate, in gold flows, and in the terms of trade display little association withmovements in new foreign issues. Only lagged changes in British imports (and, to a lesser extent, exports)have much tendency to ‘produce’ a change in new foreign lending. It is difficult to determine, on the basis ofthese correlations, whether it is more appropriate to regard fluctuations in new foreign lending as inducedby swings in other components of the balance of payments or as relatively autonomous, as suggested byFord.7

The strongest influence on British exports (other than their own lagged values) is British imports. Thetwo variables covary positively. Apparently, business cycle movements which raised British importssubsequently induced a rise in British exports which served to moderate the deterioration in the balance oftrade. One can imagine a two-country model in which the rise in British imports stimulated economicactivity abroad through the export multiplier, raising foreign incomes and stimulating demands overseas forBritish exports. There is also evidence of a less definitive nature that a rise in British foreign lending wasfollowed by a rise in commodity exports. Ford suggested that British funds lent to regions of recentEuropean settlement were disproportionately devoted to investment projects, creating a demand for


imported capital goods and stimulating British exports. There is some support in the table for what Fordcalled the ‘sensitivity’ of British exports ‘directly and indirectly’ to British overseas lending, but theevidence is far from overwhelming.

Imports appear to have been little affected by changes in Bank Rate or other variables. If monetary policywas effective in influencing the evolution of the balance of payments, it appears to have operated mainlyby influencing other components of the external accounts. Similarly, the terms of trade do not appear tohave been directly responsive to the Bank of England’s discount rate. There is little evidence in table 10.2of the Triffin Effect’—the tendency of a rise in Bank Rate to induce liquidation of commodity stocks andstrengthen Britain’s terms of trade, although the impact of changes in Bank Rate on the terms of trade ismore pronounced than its impact on foreign lending, exports, imports or gold flows.9 Indeed, the dominantimpression that emerges from the first row of the table is that Bank Rate had only a weak impact on theBritish balance of payments. This is consistent with Sayers’ (1936) conclusion that, for much of the period,the Bank of England was still struggling, with mixed success, to render its discount rate effective, and withFord’s conclusion that the power and influence of Bank of England monetary policy has been exaggerated.

The final column of table 10.2 shows that lagged values of most of the variables under consideration hadonly a weak direct impact on the Bank of England’s gold reserve. (Those variables still could have affectedthe reserve indirectly, a possibility considered below.) However, lagged values of the terms of trade doexhibit an association with the Bank of England’s reserve, which may give pause to those inclined todismiss the relevance of the price-specie flow mechanism.

The above are partial equilibrium inferences. To focus attention on general equilibrium repercussions,Figures 10.6 to 10.11 show the response of the system to perturbations to the disturbance terms in three ofthese equations. (The equations are first transformed to moving-average form, as is standard in the literature.)The resulting system is then perturbed, in turn, by a one standard deviation shock to the disturbances to theforeign issue, export, and gold equations.10 I interpret the three disturbances as temporary shocks to thecapital account, the current account, and to confidence.

Consider first the response, depicted in figures 10.6 and 10.7, to a temporary increase in new foreignlending. The autonomous weakening of the balance of payments produces an immediate gold outflow. The

Figure 10.6 Impulse response to shock to foreign lending


Bank of England’s discount rate is raised, presumably in an effort to damp the loss of reserves. The volumeof new foreign issues falls back toward pre-disturbance levels, as if higher British interest rates increase theattractiveness of domestic investment. Indeed, the volume of new foreign issues declines temporarily belowits steady state level, due one supposes to higher domestic interest rates.

The response of the current account of the balance of payments is less straightforward. Exports riserelative to imports with the initial weakening of the capital account (figure 10.7). The rise in exports isassociated initially with some deterioration in Britain’s terms of trade, as if the relative price of Britishgoods had to decline in order to stimulate their increased absorption overseas. But by the second year following

Figure 10.7 Impulse response to shock to foreign lending

Figure 10.8 Impulse response to shock to exports


the shock, the terms of trade have recovered fully. From years three to seven, they are above their steady statelevel, as if the surge in foreign lending, once it had time to work its way through the system, shiftedrightward the foreign demand curve for British goods.

Figures 10.6 and 10.7 may say something about the geographical source of shocks to British foreignlending. If shifts from domestic to foreign investment had been produced by increases in the marginalefficiency of capital overseas, this should have provoked a rise in import demand overseas at the same timeas it attracted financial capital from Britain. The rise in foreign demand should then have strengthenedBritain’s terms of trade. If, in contrast, shifts from domestic to foreign investment had been produced

Figure 10.9 Impulse response to shock to exports

Figure 10.10 Impulse response to shock to gold


mainly by inward shifts in the marginal efficiency of capital schedule at home, this should have provoked adecline in domestic demand and a deterioration in Britain’s terms of trade sufficient for an increased shareof domestic output to be willingly absorbed abroad. There is support for this interpretation in the fact thatBritain’s terms of trade deteriorate on impact, although their subsequent tendency to strengthen issubversive of that conclusion.

A possible resolution of the paradox may be to argue that foreign shocks in fact drove the fluctuation offoreign lending over the cycle, but that foreign households and firms could not finance their increasednotional demand for imports from Britain due to a binding balance-of-payments constraint. They had towait for the response by British investors and the receipt of long-term capital inflows before increasing theirpurchases of British goods. Figures 10.6 and 10.7 support this interpretation. Thus, the domestic demand forBritish goods declined initially, as financial capital was shifted from home to overseas investment inresponse to the productivity shock abroad. Over time, overseas investors begin to devote their increasedsterling balances to purchases of British goods, driving up British export prices. Thus, following their initialdeterioration, Britain’s terms of trade quickly strengthen. Exports rise steadily. It indeed appears that in theshort run an increase in British foreign lending provoked or facilitated little rightward shift in the demandschedule for British exports, but that over time British lending began to translate into a rise in foreignexpenditure, including expenditure on imports. Eventually, these effects damp out, as exports, foreignlending and the terms of trade return to their steady state levels. But over the intermediate run, from two tosix years following the shock, the terms of trade strengthen relative to their steady state level.

Figures 10.8 and 10.9 show the response to a temporary increase in exports. The autonomous rise in exportrevenues strengthens the balance of payments and leads initially to an accumulation of gold reserves. Butthe gold inflow is quickly attenuated. Imports rise in response to the export boom. Long-term foreignlending also rises (figure 10.9). Presumably, the rise in demand overseas for the products of British industryincreased British incomes, provoking the rise in import demand.

That foreign lending rises rather than falls following the autonomous increase in exports is helpful fordistinguishing between two competing interpretations of the export shocks experienced by the British

Figure 10.11 Impulse response to shock to gold


economy. If exports had tended to fluctuate mainly in response to productivity shocks at home, the logicalconsequence of a positive shock to export capacity would have been to render domestic investment moreattractive relative to foreign investment, and to produce a decline in new foreign issues. If, in contrast,exports tended to fluctuate primarily because of external shocks to incomes overseas, a positive shock toexports would have rendered foreign investment more attractive relative to domestic investment, and onewould have expected foreign lending to rise. It is the second case that is observed. The notion that thefluctuation of British export markets was driven mainly by disturbances abroad is supported by the initialimprovement in the terms of trade following the positive export shock. If fluctuations in capacity at homerather than in demand abroad had been the primary source of British export cycles, one would expect to seethe relative price of British exports decline (the terms of trade deteriorate) following a positive export shock,whereas they improve initially. The explanation for the subsequent deterioration in the terms of trade is lessevident.

Figures 10.10 and 10.11 show the response to a temporary, one standard deviation fall in gold reserves. Byconstruction, the initial gold outflow is not associated with a rise in imports, a fall in exports, or an increasein long-term lending abroad. Thus, the disturbance can be interpreted as a purely financial shock—say anautonomous short-term capital outflow. In response to the loss of reserves, the Bank of England raises itsdiscount rate. Reserves recover quickly. It would appear that the principal channel through whichadjustment to short-term capital outflows took place was the response of the Bank of England and itscapacity to attract capital flows. Interestingly, there is little evidence of an improvement in the trade balancefollowing the loss of gold reserves. To the contrary, the fall in export volume which takes places in responseto the loss of reserves and the rise in Bank Rate swamps the fall in import volume. The behaviour of exportsis attributable, presumably, to the decline in economic activity induced by the financial market disturbanceand the higher interest rates it provokes. The decline in export volumes is offset, however, by someimprovement in the terms of trade. The two effects roughly cancel, yielding little net change in the currentaccount of the balance of payments.

Interpretation

The metaphor for the balance-of-payments adjustment mechanism that the statistical analysis suggests is aslim man in a winter storm. Like a slim man with little flesh on his bones, the Bank of England had only aslim gold reserve surrounding a vulnerable gold standard frame. To survive in a winter storm, a slim manmust dress in layers of clothing to insulate himself from the cold. The Bank of England similarly possessedseveral layers of insulation to protect itself from the elements. I describe the nature of this insulation byconsidering export, foreign lending and confidence shocks in turn.

The first layer of insulation from export fluctuations was provided by parallel movements in imports. Aforeign expansion which raised the volume of British exports induced parallel movements in importsthrough two channels: first, the terms-of-trade improvement provoked by the rise in the overseas demandfor British exports reduced the relative price of British imports; second, the rise in export demandstimulated domestic production in Britain, increasing the need for intermediate imports and, by raising realincomes, further augmenting the demand for imported customer goods. Thus, even when trade flows werethe source of the external imbalance, they also provided the first layer of insulation.

But induced changes in imports financed only a fraction of autonomous fluctuations in exports. Thesecond layer of insulation was provided by changes in foreign deposits and security holdings. Countrieswhich made up Britain’s principal overseas markets tended to hold sterling balances in London. Instead ofdemanding that the Bank of England convert into gold any sterling balances they acquired from export sales,


often they invested in securities in London. In effect, a British trade deficit automatically generated a short-term capital inflow which helped to relieve the pressure on the external accounts.

The third layer of insulation was provided by the banking system through the mechanism emphasized byWhale. Insofar as the autonomous rise in export demand stimulated industrial activity, it increased domesticdemands for money and credit. Interest rates were driven up, and short-term capital was attracted fromabroad. The increase in circulating media was provided by the banking system, partly in response to itsreceipt of short-term deposits from abroad.

The fourth and final layer of insulation was provided by the Bank of England. The Bank could raise itsdiscount rate in response to gold losses to attract capital inflows. Characterizing the Bank’s role in this wayserves to remind that, while the balance-of-payments adjustment process was far from automatic, neither didit hinge exclusively on discretionary management by the Bank of England.

The sources of insulation that came into play in response to swings in foreign lending were essentially thesame, although their relative importance differed. When new issues for an overseas borrower were floated,the proceeds would be deposited to the borrower’s London account. In the first instance, then, new foreignlending led to no change in Britain’s balance of payments position. Only over time, as those deposits weredrawn down, would the British balance of payments weaken and might the Bank of England begin to losegold. In the short run, therefore, foreign deposits were Britain’s first layer of insulation and the onlyinsulation required.

As the borrower subsequently drew down its deposits to finance purchases of imports, British exportswould begin to rise. The importance of this mechanism should not be exaggerated. Typically, a borrowerlike Canada or Australia would use only a portion of any increase in foreign funds to import commoditiesand equipment. Only a portion of any such purchases would come from Britain. It is possible, of course,that the demand for British exports was stimulated indirectly. If Canada purchased imports from the UnitedStates, the stimulus to US incomes might lead to increased American imports of British goods. But thiscircular mechanism was subject to leakages which attenuated its operation and built lags into the responseof British commodity exports to prior foreign lending.

There was, none the less, at least some response of commodity exports to prior foreign lending. At thesame time, since investment had shifted from the domestic to the foreign market, economic activity at homewould decelerate, generating a sympathetic fall in British imports. Thus, a strengthening of the tradebalance was Britain’s second layer of insulation.

The final layer of insulation was again provided by the Bank of England. The Bank could raise itsdiscount rate and induce an inflow of short-term capital to partially offset the outflow of long-term funds.

The response to an autonomous outflow of gold not associated with a fall in exports or a rise in foreignlending—in other words, a shock to confidence—was very different. There was no reason for foreigndeposits to rise or for the trade balance to strengthen. Thus, Britain’s two outer layers of insulation werestripped away. The burden of adjustment was placed squarely on the Bank of England.

It is not obvious how the Bank of England so successfully shouldered this burden. Part of the explanationis that the British authorities attached clear priority to defence of the gold standard. If the nationsimultaneously experienced gold losses and a cyclical downturn, or gold losses and financial panic, therewas no question that the authorities attached priority to defence of the monetary standard, even if thisimplied an intensification of domestic difficulties. They had demonstrated their commitment to the goldstandard in 1847, in 1866, and on numerous other occasions. The impact of monetary policy on domesticactivity, while vaguely understood, had not been clearly articulated, as it was by Keynes and others in the1920s. Prior to the extension of the franchise and the rise of the Parliamentary Labour Party, there could belittle effective pressure to adapt monetary policy toward employment targets.


To use modern terminology, the Bank of England had acquired a reputation for action that rendered itscommitment to the gold standard fully credible. There was no need for domestic or foreign depositors to runon the Bank as a way of testing that commitment. To the contrary, in times of difficulty, short-term capitalwould tend to flow toward Britain in anticipation of official intervention. If sterling fell toward the goldexport point, speculators purchased it in anticipation of official actions designed to strengthen the exchangerate. Intervention would be rendered largely redundant.

The Bank’s gold reserve was small in comparison with Britain’s external obligations. It is not obvious,therefore, how it succeeded in defending its reserve and maintaining convertibility in instances like 1890and 1907 when the shock to confidence and consequent gold flows were large relative to the Bank’sreserve. Part of the explanation is that by 1890 the official commitment to the gold standard was not merelynational but international. The Bank of England could rely on assistance from abroad, notably from theBank of France. The resources available for sterling’s defence were not limited to those of the Bank ofEngland. In 1890, in response to the Baring Crisis, the Bank of England obtained, with the aid ofRothschilds, a loan of £2 million of gold from the Bank of France. It secured £1.5 million in gold from theRussian government.11 The mere announcement that these funds had been made available proved sufficientto stem the drain from the Bank of England. There was no need for much of the French gold to be ferriedacross the Channel. Anticipating that concerted international action to defend the sterling parity would beforthcoming, speculators reversed the direction of capital flows so as to render that action unnecessary.

The 1907 crisis provides an even more telling illustration of the point. Bank failures led to a shift out ofdeposits and into gold in the United States, and to a flow of specie from Britain to America. In response tothe loss of gold reserves, the Bank of England first borrowed on the market. Next it raised Bank Rate in aneffort to attract gold from third countries, and restricted discounts to short-dated paper only. Finally itobtained support from abroad. But with memory of 1890 still fresh, it was not even necessary for foreignsupport to be actively solicited. The Bank of France purchased sterling bills, presumably on its own initiative.Apparently aware that the Bank of France was intervening on behalf of its British counterpart, speculatorsreversed course, repurchasing the sterling assets they had liquidated previously. Again, the response of themarket minimized the need for official intervention. The credibility of the commitment to the sterlingparity, which extended beyond British shores, did much to relieve the pressure on the Bank of England.

Conclusion

In this chapter I have provided a ‘Ford-like’ model of how the gold standard worked. The model suggeststhat the gold standard’s smooth operation—smooth from the perspective of the centre—depended on aparticular constellation of market forces. Many of those relationships are the very ones first emphasized byFord more than a quarter of a century ago. My model differs by attaching somewhat less weight than didFord to the linkage running from British capital exports to capital-good imports by the recipient countriesand to exports of capital goods by Britain. It attaches more weight than did Ford to monetary managementby the Bank of England. But in emphasizing the adjustment of real variables to monetary impulses as wellas adjustments in the other direction, and in its retention of certain Keynesian features, the model is verymuch in the spirit of the one developed by Ford.

The final element of my explanation for the stability of the British gold standard is the credibility of theofficial commitment to gold. Policymakers in Britain were unwavering in their commitment to goldconvertiblility. To the extent to which there existed other goals of policy, these were accorded lowerpriority. Knowing that policy-makers would intervene in defence of the gold standard, markets responded inthe same direction in anticipation of official action. Hence, the need for actual intervention was minimized.


Credibility and stability were by-products of a reputation for willingness to take the necessary action. But acentral element of my argument is that credibility derived from the fact that the commitment to the goldstandard was international. Central banks like the Bank of England could rely on foreign assistance in timesof exceptional stress. Again, the need for actual assistance was minimized because the commitment to offerit was fully credible. The markets anticipated the actions of the central bankers, basing their anticipations onthe policy-makers’ track records. Their anticipations and consequent actions rendered official interventionlargely redundant.

In fact, this point was understood well before the recent literature on credibility and reputation. ‘[D]omestic confidence in convertibility was nourished on its past success’, wrote Ford in 1962. ‘[F]or somecountries the maintenance of specie payments was never endangered by domestic speculative runs on gold,while for others a less successful history meant that the additional threat of a domestic speculative drain wasever present…’12

Notes

1 See Hume (1752) and Whale (1937).2 Precursors who also emphasized the output and employment effects of international reserve flows include Angell

(1926), Ohlin (1929) and White (1933). Hints of the approach can be discerned even earlier, in the debatebetween Taussig (1917) and Wicksell (1918). None of these authors elaborated a fully-articulated Keynesianmodel, obviously. But Ford’s model owed much to Meade’s (1951) seminal work on income determination in theopen economy.

3 Ford (1962), p. 189.4 The title of this section is lifted directly from Ford (1960).5 I have utilized vector autoregression in another context, where I present my views of its applicability to questions

of causality and justify my preferred interpretation of such results. See Eichengreen (1983).6 A constant term and time trend also were included but are not reported in the table.7 Ford (1962), p. 190.8 Ford (1962), p. 190.9 Note the emphasis of the word ‘directly’ in the preceding sentence. Gold flows also appear to have had some

impact on the subsequent evolution of the terms of trade. Hence it may be premature to conclude in favour of oragainst this hypothesis on the basis of the coefficients on Bank Rate alone. It could be that Bank Rate affected theterms of trade indirectly (by attracting gold flows, which altered the terms of trade). This hypothesis is pursued inpp. 198–200 below.

10 Given the order in which the variables are entered, the Choleski Factorisation chooses Bank Rate as the ‘mostexogenous’ variable. That is, the error term in the moving average representation of its equation is uncorrelated withthe other error terms. It is followed, in order, by new issues, exports, imports, the terms of trade, and the goldreserve (the ‘most endogenous’ variable). I experimented with other orthoganalisations and found them to havelittle impact on the results. For the diagram, I have multiplied the results for Bank rate by ten and divided thosefor imports and exports by ten. Otherwise, movements in variables other than imports and exports (especiallyBank Rate) would not be apparent to the naked eye.

11 Sayers (1936), p. 103.12 Ford (1962), p. 189.

References

Angell, James W. (1926), The Theory of International Prices, Cambridge, Mass., Harvard University Press.


Eichengreen, Barry (1983), ‘The Causes of British Business Cycles, 1833–1913’, Journal of European EconomicHistory, 12, 145–61.

Feinstein, Charles (1972), National Income, Expenditure and Output of the United Kingdom, 1855–1965, Cambridge:Cambridge University Press.

Ford, A.G. (1960), ‘Notes on the Working of the Gold Standard Before 1914’, Oxford Economic Papers, 12, 52–76.Ford, A.G. (1962), The Gold Standard 1880–1914: Britain and Argentina, Oxford, Clarendon Press.Hume, David (1752), ‘On the Balance of Trade’, in Essays, Moral, Political and Literary, vol. I, 1898 edition, London,

Longmans, Green.Meade, J.E. (1951), The Theory of International Economic Policy, Oxford, Oxford University Press.Mitchell, B.R. and P.Deane (1962), Abstract of British Historical Statistics, Cambridge, Cambridge University Press.Ohlin, Bertil (1929), ‘The Reparation Problem: A Discussion’, Economic Journal, 39, 172–3.Palgrave, R.H. (1903), Bank Rate and the Money Market, London, J.Murray.Sayers, Richard S. (1936), Bank of England Operations, 1890–1914, London, P.S. King and Son.Simon, M. (1968), ‘The Pattern of New British Portfolio Investment, 1865– 1914’, in A.R.Hall (ed.), The Export of Capital

1870–1914, London, Methuen, pp. 15–44.Taussig, Frank (1917), ‘International Trade Under Depreciated Paper’, Quarterly Journal of Economics, 31, 330–403.Whale, P.B. (1937), ‘ZThe Working of the Prewar Gold Standard’, Economica, 11, 18–32.White, H.D. (1933), The French International Accounts, 1880–1913, Cambridge, Mass., Harvard University Press.Wicksell, Knut (1918), ‘International Freights and Prices’, Quarterly Journal of Economics, 32, 401–10.


11The dynamics of international monetary systems: international

and domestic factors in the rise, reign, and demise of theclassical gold standard*

Jeffry A.Frieden

An international monetary system of fixed nominal rates is at one and the same time very simple and verycomplex. Such a system is rudimentary in that all participants simply agree to a set of stable values of theircurrencies. The prime example of a fixed-rate system is the classical gold standard, which was a pillar of theworld economy from about 1870 until 1914. That such a system is complex in many ways is indicated bythe abject failure of continued attempts to reestablish the gold standard during the interwar years, and by theinvolved and extended negotiations over a limited fixed-rate system among the members of the EuropeanCommunity since 1973.

The purpose of this essay is to examine the dynamics of international monetary systems, specificallysystems of fixed nominal exchange rates. The broad question is how such systems can arise and becomestable over time. More specifically, I focus on explaining how and why the classical gold standard arose,and how and why it was as stable as it was before World War One. This explanation provides the tools tounderstand why subsequent attempts at establishing fixed-rate systems have met with less stability andsuccess.

A fixed-rate international monetary system: description, definitions, and analysis

An international monetary system can be compared to a domestic monetary standard on the three traditionalroles of money: unit of account, medium of exchange, and store of value. The international monetaryregime establishes a way to equate currency values for the purposes of measuring or accounting for relativeprices, either by way of a fixed rate among currencies or against a commodity such as gold, or by way ofmarket-based floating rates. For international payments, monetary systems use as a medium of exchangeeither national currencies—usually a limited number of key currencies—or such a common tender asprecious metals. And some mix of national currencies and common tender are held for investment (store ofvalue) purposes.

All of these purposes can theoretically be served by atomistically derived and maintained markets inwhich national currencies are traded freely. However, historically most international monetary systems haveincluded some function for explicit government policy regarding the exchange rate. One such set of policiesis that which constitutes a fixed-rate international monetary system. In a fixed-rate system, currencies aretied to each other at established parities, and governments commit themselves not to alter these parities. Themost famous such system was the classical gold standard of the late nineteenth and early twentieth century,

* Reprinted from Jack Snyder and Robert Jervis (eds), Coping with Complexity in the International System, Boulder,Colorado: Westview Press, 1992, pp. 137–162. Copyright © 1992 Westview Press. Reprinted by permission ofWestview Press.

in which the currencies of the world’s leading economies were fixed to gold by their national governmentsat some legal rate. Typically (with some exaggeration) each national monetary authority on the goldstandard stood ready to exchange gold for its currency, or its currency for gold, at the established rate andon demand.

Traditional discussions of the international monetary system focus on the efficiency effects of a particularsystem, that is, its effects on global welfare (trade, payments, economic growth). In less grandiosediscussions the implications of exchange-rate arrangements on national welfare or on specific groups withinnations are considered. These discussions are primarily normative and policy-oriented. From an analyticalperspective, we can ask two parallel sets of questions along these lines, one at the global and the other at thenational level.

The first set of analytical questions is concerned with how at the international level an exchange-rateregime is adopted and sustained. In posing such questions, we abstract from global welfare considerations toask how it is that nation states are able to cooperate in establishing and maintaining a global standard. Inthis regard, international monetary coordination is an exercise in collective action, with the normal freerider and informational problems and criteria for overcoming them.

Indeed, there is some reason to regard an international monetary regime as a public good. From a globalstandpoint, as with domestic money, any standard is almost certainly better than none at all. Some generallyagreed rules of the game—even if only to allow free trading in national currencies—are needed to providethe stability necessary for international trade and payments. In the context of a fixed-rate regime, andinasmuch as international exchange-rate stability is something of a public good, there is an incentive(especially for small countries) to defect. A country can, for example, reap specific benefits by devaluing itscurrency. This makes its exports more competitive but does not challenge the stability of the system. Of course,if enough countries act similarly, the regime will collapse.

The experience of international monetary relations suggests one set of dynamics, at the internationallevel, that can help create and reinforce a regime such as the gold standard. The more countries participatein such a system, the greater the incentives to any one country to affiliate with it. This is becauseparticipation in the system gives greater access to trade and investment with other members. All else equal,firms and individuals are more likely to trade with, invest in, and borrow from countries whose currencyvalues are more predictable. So such a system can exhibit a synergistic feedback mechanism or virtuouscircle: the more countries are members of the system, the more attractive is membership. Of course, thecircle can be vicious as well: once the size of the regime begins to decline, the system can collapse rapidlyas countries defect.

In this context, the crucial question is how the synergistic process gets started and reinforced.Experiences from other such processes suggest the importance of a focal point around which actors canconverge—in the international monetary realm, perhaps a major trading and investing nation that can leadothers toward a mutually beneficial agreement on international monetary norms.1

Although interstate considerations are important, in the final analysis exchange-rate regimes are the resultof a series of national choices. Analysts thus need to examine a second set of questions, those that concernthe determinants of national policies toward the international monetary system. How, in other words, docountries come to make and sustain a commitment to an international currency standard? Determinants ofsuch a choice include both ‘national interest’ calculations of the optimal national policy and more politicalconsiderations based on the role of interest groups within national societies.

The national welfare implications of different monetary regimes have especially to do with the degree towhich they allow national policymakers to sustain or restore internal and external macroeconomicbalance.2 Perhaps the best way to examine this is by way of the contrast between fixed and floating rates.

INTERNATIONAL MONETARY SYSTEMS 151

Fixed rates provide for stability and predictability, while floating rates allow national policymakers moreindependence to respond to country-specific economic conditions. This is especially true in a world inwhich capital can move freely from country to country.

Where capital is not mobile across borders, there is no contradiction between fixed rates and nationalmonetary autonomy, as the monetary authorities can affect domestic macroeconomic conditions by alteringthe interest rate. However, where capital is mobile, interest rates are by definition set on world markets; theexchange rate is the only tool the monetary authorities have for affecting macroeconomic conditions. Thissets up a trade-off between the benefits of exchange-rate stability and the benefits of national policyautonomy. This trade-off can be (and typically is) evaluated on a national-interest basis in terms of suchcharacteristics of the country as how open to trade the economy is, how vulnerable it is to unique shocks,and so on. For example, consideration of such aggregate national-level characteristics leads to theconclusion that small open economies dependent on the export of a few commodities whose prices tend tofluctuate widely are almost certainly better off with floating than fixed rates. In this context, the choice ofregime is affected by structural characteristics of the national economy.

This suggests a second set of dynamics in the evolution of international monetary regimes. As a stableregime grows in importance and extent, and more countries associate with it, the level of foreign trade andpayments for each member grows as well. In other words, not only does the existence of the system attractmore members but the existence of a stable system leads these members to trade, invest, and borrow more.As this happens the countries become more open on current and capital accounts, and their economiesbecome more integrated and less vulnerable to unique shocks. These trends in turn give the countriesstronger national-welfare reasons to maintain their commitment to the regime. So another feedbackmechanism or virtuous circle can operate: the growth of a stable international monetary regime increases itsmembers’ international trade and investment and, therefore, their interest in ensuring regime stability.

National welfare considerations are important, but domestic distributional considerations are also centralto the choice of exchange-rate regimes. In explaining any national policy, we must carefully delineate theways in which national welfare considerations translate into pressures on policymakers and therefore intoinfluences on outcomes. We know all too well that the fact that a policy is welfare-improving from thestandpoint of society as a whole does not guarantee its adoption. The process is mediated through theinterests of domestic groups and the effects of domestic political institutions.

In fact, different exchange-rate policies have different effects on domestic socioeconomic interest groups.For those heavily engaged in international trade and payments, the stability and predictability of a fixed rateis eminently desirable. However, in a financially integrated world (as today and before 1914), a fixed rateeliminates the possibility for independent national monetary policy (as discussed above). This may matterlittle to those whose economic horizons are global (international banks, multinational corporations, majorexporters), but it is a real sacrifice for those tied to the domestic market. For this reason, we expectinternationally oriented economic actors to favor fixed rates, and domestically oriented economic actors tofavor floating or adjustable rates. By the same token, those that favor a devaluation (essentially producers ofimport or export-competing tradable goods) tend to oppose a fixed-rate system that prohibits devaluations.

National choices on whether and how to associate with an international monetary regime are theconsequence of domestic bargaining among interested groups in national society and within nationalpolitical institutions. The stronger those who favor a fixed rate, the more likely it will be adopted. Hereagain, a virtuous-circle feedback mechanism can be at work. The more encompassing the internationalmonetary regime, the greater the interest of internationally oriented firms in national societies inencouraging their governments to associate with the regime. Once a government becomes a member of theregime, the proportion of the economy that is oriented to foreign trade and payments is likely to grow, and


with it the influence of that sector’s policy preferences—especially to remain committed to the system. Ofcourse, here as elsewhere the mechanism can operate in reverse: as an international monetary systemcollapses it pulls the political rug out from under its supporters in national political economies by reducingthe economic influence of internationally oriented groups and lessening their commitment to the system.

In summary, two important analytical questions dominate the discussion of any international monetaryregime. First, what are the ‘national interests’ in play and what are their domestic political determinants?Second, how do independent nation states interact in their policies toward the international monetary arena?In answering both questions I have identified important dynamic effects. At the subnational level, the moreextensive and stable the international monetary regime, the greater the incentives for internationally orientedgroups within any individual country to pressure their government to affiliate. At the national level, thelarger the international monetary regime, the higher the level of national trade and payments of eachmember country and the more linked the markets among countries, and thus the stronger the nationalwelfare argument for each country to affiliate with the regime. At the interstate level, the greater the numberof countries that are members of the regime, the greater the incentive for additional countries to affiliate.

This is really one set of dynamics—a feedback mechanism that includes both strategic interaction amongstates and politics within them—but it can usefully be divided between its international and domesticcomponents. Internationally, we should observe convergence of ever greater numbers of states around aninternational monetary focal point as such a focal point becomes more and more attractive and credible.Domestically, we should observe more domestic support for affiliation with this focal point (regime) as itgrows in extent and credibility.3 In the next section I trace these interrelated processes through the history ofthe classical gold standard.

The rise of the gold standard: international dynamics

It is unclear when the classical gold standard began, for the world’s major countries went on gold atdifferent times. For example, from the 1820s onward the United Kingdom, France, and the United Stateswere all on a specie standard quite similar to the subsequent monometallic gold standard.4 By 1880,however, every major trading and financial nation on earth had tied its currency legally to gold at a fixedrate. This situation was to prevail—indeed, more and more countries were to join the regime—until 1914,when World War One brought the system down. During these decades, the advanced industrial countries ofthe world, and many developing countries as well, grew at unprecedented rates. International trade andpayments increased continually until the world economy was probably more integrated than it has ever beenbefore or since. In purely economic terms, the gold standard era was something of a truly golden age.

The near automaticity assumed to inhere to the gold standard was expressed eloquently, and not withoutsome nostalgia, by John Maynard Keynes in 1920:

The inhabitant of London could order by telephone, sipping his morning tea in bed, the variousproducts of the whole earth, in such quantity as he might see fit, and reasonably expect their earlydelivery upon his door-step; he could at the same moment and by the same means adventure hiswealth in the natural resources and new enterprises of any quarter of the world, and share, withoutexertion or even trouble, in their prospective fruits and advantages; or he could decide to couple thesecurity of his fortunes with the good faith of the townspeople of any substantial municipality in anycontinent that fancy or information might recom mend. He could secure forthwith, if he wished it,cheap and comfortable means of transit to any country or climate without passport or other formality,could despatch his servant to the neighboring office of a bank for such supply of the precious metals as


might seem convenient, and could then proceed abroad to foreign quarters, without knowledge oftheir religion, language, or customs, bearing coined wealth upon his person, and would considerhimself greatly aggrieved and much surprised at the least interference. But, most important of all, heregarded this state of affairs as normal, certain, and permanent.5

The classical gold standard was a remarkably uncomplicated mechanism. To be ‘on gold’, a country simplyfixed a legal value in national currency at which the monetary authorities (typically the mint or the centralbank) would buy or sell gold. This effectively established a fixed legal rate of exchange between gold andthe currency, and thus between all other gold-standard currencies and the national currency. A number ofsubsidiary ‘rules of the game’, not formal but widely understood, were designed to ensure that thegovernment would be able to guarantee free convertibility of the currency into gold.6

The economic implications of the gold standard were also quite simple and were understood in latemedieval times. If a country ran a persistent trade deficit, gold would flow out and the money supply wouldcontract.7 This would drive domestic prices down relative to world prices, thus increasing exports and reducingimports—and bringing trade back into balance. The process ran in reverse for countries with persistentsurpluses.

To repeat the points made more generally above, a credible commitment to gold provided economicagents with a marvelously predictable exchange rate, but it also greatly restricted the ability of nationalgovernments to affect national monetary conditions. This trade-off was also widely recognized, especiallyin the common practice of all governments of going off gold during major crises, such as wars, in whichpolicy autonomy was unarguably more important than exchange-rate stability.

The classical gold standard’s prehistory began in 1717, when Britain’s master of the mint, Sir IsaacNewton, set the price of an ounce of gold at £3 17s 10 l/2d.8 Silver remained coin, and the country waslegally bimetallic, but at this rate gold was overvalued at the mint, so silver gradually disappeared fromcirculation.9 Silver was demonetized in 1774 in recognition of the fact that only gold circulated in theUnited Kingdom.

In 1797, in the midst of the Napoleonic Wars, the British government suspended gold convertibility. Thecountry remained off gold until 1821, and this ‘paper pound’ period led to one of the most famous debatesin the history of economics, the Bullionist Controversy. Inconvertibility had been accompanied by inflationin Britain, and discussion of the causes of inflation raged throughout the suspension. Of the position thatprevailed in the parliamentary Bullion Committee report of 1810, David Laidler has written, ‘No otherdiscussion of economic policy issues prepared by working politicians has had so sound an intellectual basisand has stood the test of time so well’.10 The report presaged much of modern monetary theory, pointingout that inconvertibility allowed the monetary authorities to increase the money supply at will and thuscreated the potential for inflation.’ 11 It similarly argued for a commitment to gold as a safeguard againstsuch a danger. In other words, it recognized the trade-off between the ability of the government to affectmonetary conditions and commitment to a fixed exchange rate.

At much the same time as Britain was going back to gold, France was settling into a stable bimetallicstandard. The gold-silver rate was, in other words, set so that neither metal was overvalued relative to theother. In these conditions, bimetallism was consonant with the gold standard. The United States, aftershifting the rates a bit in the search for balance, settled in the 1830s on a rate that essentially drove silverout of circulation and put the country on gold. Almost all other countries were on a monometallic silverstandard.

Conditions were disturbed somewhat by California and Australia gold discoveries, which increasedannual world gold production from about $36 million in the 1840s to about $119 million in the 1850s.12


Countries solely on silver or gold were safe, but those on a bimetallic standard found gold increasinglyovervalued at the previous mint price—the market price was dropping while the legal price remained thesame. Most bimetallic countries, which tended to favor silver over gold, readjusted the gold-silver rate toavoid a de facto shift to gold; some demonetized gold.

In some ways, this experience can be taken as the starting point of our dynamic international story. Bythe 1850s, two things were clear. First, the United Kingdom was irrevocably on the gold standard. Neitherchanging gold market conditions nor intermittent panics would shake the Bank of England’s commitment tothe general principles of the gold standard.13 Second, the United Kingdom was the world’s financial andcommercial center. London had come to dominate international finance, shipping, and trade; the country asa whole was far and away the world’s most important economy; and the pound sterling was becoming thevehicle currency for most international trade and payments.14 This process was especially clear in thefinancial realm: after averaging £5.5 million a year in the 1830s and 1840s, British net foreign investmentrose to £20 million a year in the 1850s and £37.2 million a year in the 1860s.15

The centrality of the British economy to the global economy, and unshakable British commitment to gold,exerted a gradually increasing pull on the rest of the world. The areas of recent settlement both in and out ofthe British Empire, which were tightly integrated into British trade and payments in this era, were earlyadherents to gold. They relied on London for almost all their foreign finance, and foreign finance was quiteimportant to their economies (especially to the construction of a transportation infrastructure). Their all-important commodity exports also found their major market in London.

The rationale was straightforward. British importers and investors were more likely to sign contracts withcountries whose currencies had stable values against sterling. The simplest way to signal such predictabilitywas to go onto gold. British traders and financiers could (and often did) specify payment in sterling, butcurrency instability in the foreign country added an unnecessary risk to business transactions. From thestandpoint of the foreign country’s economic agents, of course, local currency movements were especiallyunsettling if contracts were in sterling: an unexpected fluctuation of the local currency could bankrupt asterling debtor. For both reasons, those drawn into the British commercial and financial orbit had strongreasons to gravitate toward gold as well.

Germany was perhaps the single most important such case. From 1837 on the process of unification wenthand in hand with movement toward currency union, largely around silver. However, internationallyoriented financial and commercial interests after 1860 pressed continually for conversion to gold. This wasfacilitated after the country’s victory over France in 1871, which led to a substantial French indemnitypayment in gold. The resulting increase in the government’s gold reserves allowed it to go onto gold withlittle difficulty between 1871 and 1873.

The French resisted longer. In 1865 France, Belgium, Switzerland, and Italy created the Latin MonetaryUnion, which attempted to establish a stable bimetallic standard with realistic gold-silver rates. Aninternational monetary conference convened at French initiative in 1867 attempted to arrive atinternationally coordinated bimetallism, with no success. Indeed, only the French were enthusiastic aboutbimetallism per se; the other members of the union preferred going onto gold.

In the Latin Monetary Union as in other countries wavering between gold and bimetallism, the issue wasforced after 1873, as silver dropped in price after major discoveries in the United States. Bimetallic countrieswere faced with either having to buy silver at an overvalued rate, thus driving gold out of monetary use, orengaging in another round of recalibrations of the gold-silver rate. The latter course seemed unstable, assilver continued to flood onto the market. The former course meant forgoing the use of gold as money at atime in which gold was more and more important to international trade and payments. Virtually all


countries chose gold over silver, commitment to the international regime over domestic policyindependence.

By 1878 the members of the Latin Monetary Union had declared for gold. Meanwhile another internationalmonetary conference, called at US initiative to try to salvage bimetallism, was a fiasco. Britain, Germany,Belgium, and Switzerland refused to go along, and while a few countries were sympathetic to the Americanposition they did not represent a critical mass.16 For Italy the issue was a bit contrived. In the face of chronicfiscal imbalances, the country went to inconvertibility in 1866 and did not go back to gold until 1884, onlyto go off again in 1894.

Italy was the exception. As Germany went to gold between 1871 and 1873, Sweden, Norway, andDenmark followed suit. With Britain and Germany accounting for the vast majority of their trade andpayments, their way was clear. The three Scandinavian countries also formed a Scandinavian MonetaryUnion, which largely consisted of provisions to allow each country’s currency to circulate freely in all threenations. By 1875 the Netherlands was de facto on gold, and in that year the United States (which hadsuspended convertibility during the Civil War) announced that it was going back to gold by 1879. In 1877Finland, politically dependent upon Russia but economically similar to the Scandinavian countries, linked togold.

Austria-Hungary and Russia had been on depreciated paper currencies since 1848 and 1839 respectively.As silver prices dropped, the paper currencies actually rose above their legal silver prices in the 1870s. In1879 Austria-Hungary announced its intent to go to gold, which it achieved in 1892. In 1885 Russia went tobimetallism but was forced off silver and onto gold alone between 1893 and 1895. Even Japan went frombimetallism to gold in 1871, only to be forced off during a financial crisis in 1878. Fifteen years later thegovernment declared its intention to go back to gold, and the process was assisted by a military victory overChina in 1895 and the subsequent payment of a £39 million gold indemnity.

As this account indicates, the process exhibited much of the synergistic feedback expected. Theincentives to go to a monometallic gold standard increased especially after international financial flows outof London became very large in the 1850s and 1860s. The issue was forced by the drop in the market priceof silver over the course of the 1870s. German accession to the gold standard in 1871–3 pulled Scandinavia,the Netherlands, and Finland along by 1877. Japan went to gold in 1871 and the US in 1875, while Belgium,France, and Switzerland were officially monometallic by 1878. By this point, a precipitate seven-year rushtoward gold had placed virtually every major trading and financial nation on earth on the gold standard.This would appear to be clear evidence of the upward spiral or feedback effect at the international level, inwhich each additional member (especially larger members) served in a dynamic way to attract furthermembers of the regime until the gold standard was essentially universal among major nations.

The rise and reign of gold: domestic-international dynamics

The international dynamic described above relied upon a more nuanced domestic dynamic. Countries’choices to go on or off gold were made in the context of often bitter debates among groups in society thathad vested interests for or against the fixed-rate standard. The contours of these distributional divisionswere outlined above; in this section I discuss how international trends and national policies interacted tospeed the rise and strengthen the reign of the classical gold standard.

Even in Britain, the gold standard was controversial at a crucial turning point. The debate overresumption after the end of the Napoleonic Wars was not just about economic theory; it involved economicinterests. Those tied to international trade and investment, such as Dutch-Portuguese Jewish merchant andfinancier David Ricardo, were favorable to early resumption of specie convertibility at the preexisting rate.


Those producing tradable manufactures, especially in and around Birmingham, were opposed. They wanteda depreciated currency, not the appreciated one resumption implied, even if this required staying on papermoney.17 City of London (financial and commercial) interests prevailed, an outcome that accords withrecent arguments about the political power of the city against manufacturing.18

In Germany, the politics of gold was more complex. The powerful rye farming interests, the Junkers,were strong supporters of silver, which implied a relatively depreciated currency. A weak currency wouldraise the amount of German currency received per pound of rye exported, and the domestic price ofimported grain. The Junkers were dismayed after 1871, when the German government sold off its monetarysilver in order to go to a monometallic gold standard. Silver sales drew money out of circulation and tendedto appreciate the gold mark, thus making German rye less competitive on world markets. Debates over silverraged in tandem with debates over trade protection.

Both issues were decided when, in 1879, Bismarck shifted toward protectionism. The Junkers hadpreviously been indifferent to agricultural protection (as exporters) and hostile to industrial protection.Bismarck essentially cut a deal with Junker soft-money supporters. In return for their support for high levelsof industrial and agricultural tariffs, Bismarck halted the selling of monetary silver, thus arresting the realappreciation of the mark. This sop to the silver interests was sufficient to win them to the side of protection,and it was mild enough not to threaten the German commitment to gold.19 The Scandinavian countries weredominated by economic actors closely tied to the import-export trade and foreign finance; they stronglysupported gold and were quickly triumphant.20

In some cases the domestic-international dynamic was most clearly operative when the gold standardcame under domestic political challenge. In these circumstances, political conflict over gold tended todivide those strongly tied to international commercial and financial activities, who had an interest inaffiliation with the system, from those who either were domestically oriented or wanted a depreciatedexchange rate, who had an interest either in paper currency or in devaluation (usually both). In the parlanceof the day, these two positions were commonly known as ‘hard money’ and ‘soft money’, respectively.21

By far the most important and striking example of the domestic-international dynamic was the UnitedStates. Although international trade and payments affected only small portions of the American economy inthe late nineteenth century, groups tied to the foreign sector were powerful. American financial marketswere in fact closely linked with those abroad, especially in London.22

Support for hard money came from Northeastern traders, bankers, and investors, and from most export-oriented manufacturers. Soft money, devaluation, and going off gold was preferred by farmers andmanufacturers from the interior, whose markets were domestic and who worried primarily about the lowdomestic prices of their products. The division persisted throughout decades of conflict, which wasexacerbated when international conditions made the preferences of either of the groups more intense—suchas when the growth of British trade and finance increased the incentives for the internationalist groups to tiethemselves to London, or when falling farm prices increased the desire of agricultural groups for adevaluation.

During the Civil War the United States went off gold as prices more than doubled under a paper currency(‘greenback’) standard. After the Civil War, the Treasury shrank the money supply to appreciate the dollarand move toward resumption of gold convertibility at the prewar rate. This real appreciation put severepressure on tradables producers, especially manufacturers. The Greenback movement first developed as aresponse among the iron and steel manufacturers of Pennsylvania. They were the country’s leadingprotectionists and recognized that devaluation cum reflation would reverse the relative price decline. Thiscould only be accomplished if the country stayed off gold. The railroadmen concurred, as did investmentbankers and others tied to these industries. Soft-money advocates worried little about the international


credibility gold might bring, for they bought and sold next to nothing abroad. For them the world economywas a threat from which to be protected. As a Chicago merchant put it, ‘these gentlemen on the seaboardbase all their calculations on gold, to bring them to par with foreign countries, leaving us in the West to takecare of ourselves’.23

Around 1873 two important groups were drawn into the soft-money camp. Farmers were originallyindifferent, for farm prices held up quite well in the first years after the Civil War. However, the Panic of1873 initiated a secular decline in farm prices. Like manufacturers, farmers recognized that devaluationwould raise the domestic price of imported farm products and would raise farm prices relative to the pricesof such nontradables as transportation and financial services.

The second important group brought into play after 1873 was silver miners. In 1873 silver was removedfrom circulation (‘de-monetized’). As the price of silver declined relative to gold, miners and Greenbackersdevised a common program to meet both their needs. If silver was ‘remonetized’ at the old 16:1 rate againstgold, the government would be forced to buy silver at well above the market rate. This would act as asubsidy to the miners; it would inject money into the economy as the government bought up silver; and itwould force the country off gold and onto a de facto (and depreciated) silver standard. A powerful allianceof Midwestern manufacturers, farmers, associated nontradables producers, and miners opposed the return togold under the banners of greenback issue and free silver.24

Supporters of gold were concentrated in the Northeast among the financial and commercial communitieswith strong ties to European trade and payments. As the New York Chamber of Commerce put it incomplaining about the risk attached to a floating exchange rate: ‘Prudent men will not willingly embarktheir money or their merchandise in ventures to distant markets…with the possibility of a fall [in gold] eretheir return can be brought to market.’25

This fundamental disagreement caused bitter political battles. In April 1874 Congress passed an inflationbill by a wide margin. The vote on the bill, which mandated expansion of the supply of paper money,illustrated the overlap of economic and regional differences. Over 95 per cent of northeastern congressmenopposed the bill, while over three quarters of congressmen from the agrarian South and the agrarian andindustrial West (including Pennsylvania) supported it. Northeastern hard-money interests immediatelymounted a furious campaign to overturn the bill. President Ulysses Grant came through with a veto, but thiscost the Republicans the congressional elections of 1874.

After the Republican electoral debacle, Grant and Republican Party leaders attempted a display of partyunity to salvage their chances for the 1876 presidential elections. In January 1875 they convinced lame duckRepublicans to vote for the Resumption Act, mandating a return to gold on 1 January 1879.

In 1876 Republican Rutherford B.Hayes, a supporter of hard money, was elected. The new Congressremained dominated by soft-money interests, and one of its first acts in February 1877 was to pass themoderately silverite Bland-Allison Act; President Hayes’s veto was easily overridden. However, in late1877 an attempt to repeal the Resumption Act was barely defeated: it passed the House and failed by one votein the Senate. Hayes’s Treasury secretary, John Sherman, had in fact worked with Republicans andDemocracts alike to find a compromise and had determined that Bland-Allison was the price of defeatingrepeal of the Resumption Act. Even so, Hayes was forced to wield the blunt instrument of patronage in orderto gather enough votes to save the gold standard.26

In the meantime, some soft-money supporters became disgusted with the two major parties. Theyfounded the Greenback Party, which stood strongly for devaluation and a flexible exchange rate. The partyran Peter Cooper for president in 1876, with little success, but did better in the 1878 congressionalelections. However, by then Hayes and Sherman had traded for or bought enough votes in Congress to saveresumption, and the country went back to gold at the beginning of 1879.


American commitment to its fixed rate against gold was relatively unquestioned during the 1880s.However, silver sentiment erupted amid the agricultural depression that began in 1888. Farm prices droppedprecipitously, and—unlike in manufacturing—productivity advances were not sufficient to counteract thistrend. Reflation and devaluation under the silverite banner would have mitigated the farm crisis, andfarmers were well aware of this. The silver miners, for obvious reasons, continued to support silvermonetization.

The most striking reflection of agrarian interests was the rise of Populism. In 1890 the Farmers’ Alliancemovement scored major electoral successes in the western states. In 1892 the People’s (Populist) Party wasformed by the southern and western alliances, along with labor groups; in that year the Populists got over amillion votes for president and sent hundreds of legislators to state houses and Congress.

Devaluation cum inflation ranked at the top of the Populists’ demands. They called for a paper money-silver standard, with the dollar fluctuating against gold. The Treasury would have been directed to regulatethe money supply to avoid deflation. Gold clauses, tying contracts to the value of gold as a hedge againstdevaluation, would have been made illegal.27

In the opposing corner, northeastern commercial and financial interests remained at the core of the hard-money camp. The bankers’ position had if anything hardened. World trade and payments were at their highpoint, and huge amounts of European money were flowing into the United States through New York.Indeed, many on Wall Street had come to hope that New York would soon be an international financialcenter, for which commitment to gold was a prerequisite.

Manufacturers were more receptive to hard-money arguments than they had been in the 1870s for threereasons. First, declining prices of manufactured products were more than compensated for by rapidproductivity increases, so few manufacturers felt substantially disadvantaged by the real appreciation.Second, by the 1890s larger portions of American industry were internationally oriented: manufacturedexports had expanded and foreign direct investment was increasing.28 Third, the manufacturers’ interest inthe money question had become secondary to their concern to defend tariff protection, which was underattack from agricultural interests.

Republican Benjamin Harrison beat eastern Democrat and gold supporter Grover Cleveland in the 1888election by promising support for silver. Harrison made good on his promise with the 1890 Sherman SilverPurchase Act. This doubled the amount of silver purchased by the Treasury under the Bland-Allison Act.The bill was too mild to satisfy antigold interests, and it was coupled with the prohibitive McKinley Tariffof 1890, which generated agrarian opposition. The result was that the Republicans lost the House in the1890 midterm elections, then lost both chambers and the presidency to Cleveland in 1892.

The Democrats had run on a silverite platform, but Cleveland was known as a gold supporter—a ‘GoldDemocrat’ in contemporary parlance. In 1893 the country was hit by a severe panic, which the financialcommunity blamed on uncertainty about commitment to the gold standard. Despite his party’s platform,President Cleveland pushed Congress to repeal the Sherman Act. Cleveland allied with gold Republicansagainst the majority of his own party and, as Grant and Hayes before him, used patronage to bludgeon keyDemocrats into submission. This repudiation of soft money was responsible for the Democrats losing the1894 midterm elections. In turn, the gold conservatives lost the battle for control of the Democratic Party tofree silver supporters.

The 1896 election was fought largely over the gold standard. Democrats and Populists jointly fieldedWilliam Jennings Bryan, who ran against the ‘cross of gold’ upon which, Bryan thundered, the country wasbeing crucified. The Republicans, in response, cobbled together a hard moneyhigh tariff coalition withpresidential candidate William McKinley as the link. He had impeccable protectionist credentials, havingdesigned the tariff of 1890; despite his long-standing support for silver, he switched to gold in 1896. This


made the McKinley candidacy a peculiar coalition of hard-money eastern trading and financial interests andhigh-tariff Midwestern manufacturers. McKinley’s striking reversal on gold lost him the western states’ silverRepublicans. However, once the election became a referendum on money, the Republicans became theconduit for millions of dollars in eastern business contributions to ensure the victory of gold.29

The narrow defeat of Bryan effectively sealed the fate of silver, and in any event antigold sentimentdampened over the next few years as economic conditions improved. In 1900 Congress passed the GoldStandard Act, Bryan was defeated a second time, and the country’s commitment to the gold standard wasfirm.

The stories told until now might be read to indicate that the domestic order simply, albeit often slowlyand painfully, validated national-welfare arguments for a decision to link to gold. This would be a mistake,as the decision was not always and everywhere taken. One fascinating example is that of Argentina.30 Thepampean country was economically and politically dominated by wheat farmers and cattle ranchers. In the1870s a secular decline in farm prices began. In this context, farm producers clamored for a currencydepreciation to increase their peso returns from declining sterling prices for wheat and beef. The dominanttradables producers, in other words, wanted a paper currency that could depreciate and compensate them (inpeso terms) for the decline in the world prices of their goods. This they achieved in 1885, when Argentinawent off gold and onto inconvertible paper money.

However, Argentina went back to gold in 1900 and stayed on until World War One. The reason for this isnot hard to discover. In 1897 world prices of beef and wheat began to rise. This meant an increased inflowof foreign currency into Argentina, and a decline in its price relative to paper pesos—in other words, anappreciation of the paper peso. In these circumstances, farmers were receiving fewer pesos for every poundof beef or wheat exported. In response, they demanded that the exchange rate be fixed against gold at adepreciated rate—to avoid the real appreciation that would make their export earnings less valuable in pesoterms. This they achieved in 1900. In other words, Argentine policy toward gold was purely and entirely afunction of the interests of the beef and wheat producers: whatever exchange-rate policy would maximizetheir earnings was adopted—floating rates and depreciation as farm prices fell, a low fixed rate as farmprices rose.

The American episodes, along with the parallel Argentine experience, illustrate the dynamic interactionof international economic trends and the domestic politics of gold. As farm and silver prices dropped, farmersand silver producers pushed for the government to go off gold and onto a depreciated silver standard; inArgentina the demand was for a depreciated paper currency. However, internationally oriented trading andfinancial groups pressed for maintenance of the gold commitment in order to ensure their fullest possibleparticipation in world trade and payments. For reasons that are far beyond the scope of this paper, in theUnited States gold won, while in Argentina gold lost until the interests of export-based farmers and ranchersshifted in the late 1890s.

The historical record, then, shows how the growth of the solidity and scope of the classical gold standardreinforced the gold sentiments of internationally oriented economic actors within national policy debates.However, increased international competition (especially the secular decline in farm prices) drove manyothers, primarily export-competing farmers, to desire a devaluation against gold. This sort of feedbackmechanism is an important intermediate step in the growth of the gold standard. Indeed, the mechanism bywhich this feedback operated often ran through groups in domestic societies who saw their interests closelylinked to national commitments to gold.


Conclusions

An international monetary system requires implicit or explicit agreement among member states about thecharacteristics and requirements of membership. This agreement in turn appears, both analytically andhistorically, to be more likely to ensue if two sets of dynamics are operative. First, at the international level,governments cohere best when there is a clear focal point around which their policies can converge. Thisprocess of cohesion is interactive inasmuch as the greater the number of countries that converge around thefocal point, the stronger the focal point. Second, at the domestic level, governments are able to commit toconvergence more successfully when they can muster domestic political support to sustain national policiesrequired by the system in question.

Under the classical gold standard, both such dynamics were present. The United Kingdom so dominatedworld trade and investment that the incentives for other countries to gravitate toward Britain and its gold-backed currency were great, and they grew ever greater as more countries so gravitated. The extraordinarygrowth of world trade and payments in the nineteenth century greatly increased the size, wealth, andpolitical power of those groups within national societies that stood to gain from being part of the goldstandard. The greater the influence of these groups, the more likely their governments were to link to gold;the more governments linked to gold, the more rapid the growth of world trade and investment; the more theworld economy grew, the stronger the incentives for groups in yet other countries became to press for theirgovernments to go onto the gold standard; and so on.

A linked international and domestic dynamic drew more and more of the world onto the classical goldstandard; the absence of this dynamic doomed the interwar gold standard to failure.

Notes

The author acknowledges support from the Social Science Research Council’s Program in Foreign PolicyStudies and from the German Marshall Fund and comments from Barry Eichengreen, Giulio Gallarotti, LisaMartin, Allan Rousso, and Tami Stukey.

1 For a related discussion of such considerations in general, see the articles in Stephen Krasner, Ed. InternationalRegimes (Ithaca: Cornell University Press, 1983).

2 For more on this topic see my ‘Invested interests: The politics of national economic policies in a world of globalfinance’, International Organization 45, No. 4 (Autumn 1991), pp. 425–451.

3 The two dimensions discussed here are similar to those that serve to orient Barry Eichengreen, Golden Fetters (NewYork: Oxford University Press, 1992). Eichengreen focuses on the twin pillars of cooperation and credibility.Cooperation, the collaborative efforts of national monetary authorities, is comparable to the inter-statecomponent of my discussion. Credibility, the reliability of national commitments to cooperative ventures, isrelated to the domestic aspects of my discussion.

4 The principal divergence from ‘classical’ gold rules was that both the United States and France were legallybimetallic. As will be discussed further on, this was not a problem until the gold-silver rate began to shift,especially in the 1870s. In other words, inasmuch as the gold-silver rate was stable, being on a bimetallicstandard (as many countries were or came to be over the course of the early nineteenth century) was tantamountto being on gold. In this sense we can push the origins of the international gold standard back to the 1820s.

5 John M. Keynes, The Economic Consequences of the Peace (New York: Harcourt, Brace and Howe, 1920), pp.11–12.

6 The best collection on the mechanism is The Gold Standard in Theory and History Ed. Barry Eichengreen(London: Methuen, 1985).


7 This ignores all of the fascinating complexities of the mechanism, such as slippage between gold reserves andcurrency issue and the role of capital movements. These are not important for our purposes. That at some levelthe rules operated as constraints is all that matters here. For the debates see Eichengreen, The Gold Standard; andA Retrospective on the Classical Gold Standard, 1821– 1931 Ed. Michael Bordo and Anna Schwartz (Chicago:University of Chicago Press, 1984).

8 This history is drawn especially from Marcello de Cecco, Money and Empire: The International Gold Standard,1890–1914 (Oxford: Basil Blackwell, 1974); R.G.Hawtrey, The Gold Standard in Theory and Practice (London:Longmans, Green, and Company, 1927); and Jacques Merten, La naissance et le développement de l’étalon-or1696–1922 (Louvain: Editions Warny, 1944). Other sources include the essays in Eichengreen, The GoldStandard and in Bordo and Schwartz, A Retrospective; Arthur Bloomfield, Monetary Policy Under theInternational Gold Standard (New York: Federal Reserve Bank of New York, 1959); and Charles Kindleberger,A Financial History of Western Europe (London: George Allen and Unwin, 1984), pp. 55–70. The facts arerelatively uncontroversial, and specific citations will only be provided for specific quotations or somewhat morecontroversial data. For a perspective on the events that focuses especially on gold as a commitment mechanism,see Michael Bordo and Finn Kydland, ‘The Gold Standard as a Rule’ (mimeo, 1991).

9 To clarify the process for those unfamiliar with it, where the mint price of a precious metal (the price thegovernment offered) was below the market price, private agents would not bring the metal to the mint for coiningbut would instead sell it at the higher market price. Where the mint price was above the market price (asNewton’s rate accomplished), gold would be brought in for coining. On a bimetallic standard, if the mint price ofone metal is overvalued relative to the other, the undervalued metal will disappear from circulation. Both metalscan be maintained in circulation if their mint prices are kept in line with market prices, which requiresadjustments in the event of major discoveries of one or the other metal. This relationship was important in thelatter part of the nineteenth century, discussed further on.

10 In The New Palgrave, under ‘Bullionist Controversy’. The classic study of the process and its aftermath is FrankW.Fetter, Development of British Monetary Orthodoxy 1797–1875 (Cambridge: Harvard University Press,1965).

11 On the actual experience, see Ian Duffy, ‘The Discount Policy of the Bank of England During the Suspension ofCash Payments, 1797–1821’, Economic History Review 35 (February 1982), pp. 67–82.

12 Calculated from David Martin, The Impact of Mid-Nineteenth Century Gold Depreciation Upon WesternMonetary Standards’, Journal of European Economic History 6 (Winter 1977), p. 643.

13 Ironically, this perception was probably reinforced by the Bank’s issue of uncovered banknotes during the Panicof 1847, as this indicated that temporary quasi-suspension in times of dire crisis would not threaten long-termadherence to the rules of the game. Bordo and Kydland, ‘The Gold Standard as a Rule’, discuss this use of acontingent rule. On this see also Rudiger Dornbusch and Jacob Frankel, ‘The Gold Standard and the Bank of theEngland in the Crisis of 1847’, in Bordo and Schwartz, Eds., A Retrospective, pp. 233–64.

14 On the theory of which see Paul Krugman, ‘Vehicle Currencies and the Structure of International Exchange’,Journal of Money, Credit, and Banking 12, No. 3 (August 1980), pp. 513–26. The rise of a vehicle currency isclosely related to the development of an international monetary regime, of course.

15 Michael Edelstein, Overseas Investment in the Age of High Imperialism (New York: Columbia University Press,1982), p. 21.

16 On these international episodes see Charles Kindleberger, ‘International Monetary Reform in the NineteenthCentury’, in Richard Cooper, Peter Kenen, Jorge Braga de Macedo, and Jacques van Ypersele, Eds. TheInternational Monetary System Under Flexible Exchange Rates (Cambridge: Ballinger, 1982), pp. 203–16.

17 Charles Kindleberger, ‘British Financial Reconstruction, 1815–22 and 1918– 25’, in Charles Kindleberger andGuido di Tella Eds., Economics in the Long View (London: Macmillan, 1982), pp. 105–120. See also Fetter, pp.64–95.

18 See especially P.J.Cain and A.G.Hopkins, ‘The Political Economy of British Expansion Overseas, 1750–1915’,Economic History Review 33, No. 4 (November 1980), pp. 463–90.


19 On the German case see Mertens, pp. 142–7; and Paul McGouldrick, ‘Operations of the German Central Bankand the Rules of the Game, 1879–1913’, in Bordo and Schwartz, Eds., A Retrospective, pp. 311–49.

20 On which see Lars Jonung, ‘Swedish Experience under the Classical Gold Standard, 1873–1914’, in Bordo andSchwartz, A Retrospective, pp. 361–99.

21 It should be noted that the dichotomous division masks nuances. It is possible to support a fixed but depreciatedrate, for example. However, inasmuch as devaluation implied going off gold, preferences over the level of theexchange rate tended to coalesce with preferences over its flexibility. Soft-money advocates typically wantedboth a flexible rate against gold (paper or silver currency) and a low (depreciated) rate; hard-money advocatestypically wanted both a fixed rate and a high (appreciated) rate. For more on this see my ‘Invested Interests’.

22 For a couple of representative studies, see Larry Neal, ‘Integration of International Capital Markets: QuantitativeEvidence from the Eighteenth to Twentieth Centuries’, Journal of Economic History 45, No. 2 (June 1985), pp.219–26; and Lawrence Officer, ‘The Efficiency of the Dollar-Sterling Gold Standard, 1890–1908’, Journal ofPolitical Economy 94, No. 3 (1986), pp. 1038–73.

23 Cited in Irwin Unger, The Greenback Era: A Social and Political History of American Finance, 1865–1879(Princeton: Princeton University Press, 1964), p. 157. See also James Kindahl, ‘Economic Factors in SpecieResumption in the United States, 1865–1879’, Journal of Political Economy 69 (February 1961), pp. 30–48.

24 It is interesting in this context to note that Junker silver interests in Germany at the time wanted the governmentto stop selling off its monetary silver, while American silver interests wanted the government to buy up silver formonetary purposes. The results of course would have been parallel.

25 Cited in Unger, The Greenback Era, p. 151.26 Unger, The Greenback Era, p. 371. Against the argument that resumption of gold payments at the pre-Civil war

parity was economically or morally inevitable, we can cite a source above suspicion for its monetaryresponsibility: ‘Our own judgement in retrospect is that, given that a gold standard was to be reestablished, itwould have been preferable to have resumed at a parity that gave a dollar-pound exchange rate somewherebetween the pre-Civil War rate and the rate at the end of the war’ (Milton Friedman and Anna Schwartz, AMonetary History of the United States, 1867–1960 (Princeton: Princeton University Press, 1963, p. 82n). Thiswould put Friedman and Schwartz somewhere between the moderate greenbackers and the strong silverites.

27 The classic study is John Hicks, The Populist Revolt (Minneapolis: University of Minnesota Press, 1931).28 See David A.Lake, Power, Protection, and Free Trade (Ithaca: Cornell University Press, 1988), pp. 91–118, for a

survey of American trade policy in this period.29 How many millions is not clear; the formal audit showed $3.5 million, with $3 million from New York; the actual

figure could have been twice or three times this. Herbert D.Croly, Marcus Alonzo Hanna (New York:Macmillan, 1912), p. 220. For details of these two episodes—Republican commitment to gold and reliance oncorporate contributions—see pp. 192–204 and pp. 209– 27.

30 The classic study of this is the masterful A.G.Ford, The Gold Standard 1880– 1914: Britain and Argentina (Oxford:Clarendon Press, 1962), especially pp. 81–169.


Part III

The interwar gold exchange standard

Introduction

The gold standard was reconstructed at considerable cost and effort after World War I. The selection fromthe First Interim Report of Britain’s Cunliffe Committee provides a definitive statement of the prevailingview of the system’s indispensability. After little more than a decade, it was recognized, however, that theinterwar system was not functioning as anticipated. The extract from the Report of the British MacmillanCommittee describes these problems as seen by informed contemporaries. Heavily influenced by Keynesand written on the eve of the 1931 financial crisis, among the destabilizing factors it emphasizes are wageand price rigidities, actions of central banks to offset the impact of international gold flows, and thetendency to accumulate foreign exchange reserves. The third selection, written by Ragnar Nurkse for theLeague of Nations, remains even today the most influential account of the collapse of the monetary system.

12First interim report

Cunliffe Committee on Currency and Foreign Exchanges after the War*

To the Lords Commissioners of His Majesty's Treasury and the Minister ofReconstruction

Introduction

My Lords and Sir,1 We have the honour to present herewith an interim report on certain of the matters referred to us in

January last. In this report we attempt to indicate the broad lines on which we think the serious currencydifficulties which will confront this country at the end of the war should be dealt with. The difficultieswhich will arise in connection with the foreign exchanges will be no less grave, but we do not think that anyrecommendations as to the emergency expedients which may have to be adopted in the period immediatelyfollowing the conclusion of peace can usefully be made until the end of the war is clearly in sight and amore definite opinion can be formed as to the conditions which will then prevail. We propose also to deal ina later report with questions affecting the constitution and management of the Bank of England, and withthe applicability of the recommendations contained in this report to Scotland and Ireland, in regard to whichwe have not yet taken evidence. We have therefore confined our inquiry for the present to the broadprinciples upon which the currency should be regulated. We have had the advantage of consultation with theBank of England, and have taken oral evidence from various banking and financial experts, representativesof certain chambers of commerce and others who have particularly interested themselves in these matters.We have also had written evidence from certain other representatives of commerce and industry. Ourconclusions upon the subjects dealt with in this report are unanimous, and we cannot too strongly emphasizeour opinion that the application, at the earliest possible date, of the main principles on which they are basedis of vital necessity to the financial stability and well-being of the country. Nothing can contribute more to aspeedy recovery from the effects of the war, and to the rehabilitation of the foreign exchanges, than the re-establishment of the currency upon a sound basis. Indeed, a sound system of currency will, as is shown inparagraphs 4 and 5, in itself secure equilibrium in those exchanges, and render unnecessary the continuedresort to the emergency expedients to which we have referred. We should add that in our inquiry we havehad in view the conditions which are likely to prevail during the ten years immediately following the end ofthe war, and we think that the whole subject should be again reviewed not later than the end of that period.

* From Cd. 9182, London, HMSO, 1918, pp. 3–7, 11–12, abridged.

The currency system before the war2 Under the Bank Charter Act of 1844, apart from the fiduciary issue of the Bank of England and the

notes of Scottish and Irish Banks of Issue (which were not actually legal tender), the currency in circulationand in Bank reserves consisted before the war entirely of gold and subsidiary coin or of notes representinggold. Gold was freely coined by the Mint without any charge. There were no restrictions upon the import ofgold. Sovereigns were freely given by the Bank in exchange for notes at par value, and there were no obstaclesto the export of gold. Apart from the presentation for minting of gold already in use in the arts (which undernormal conditions did not take place) there was no means whereby the legal tender currency could beincreased except the importation of gold from abroad to form the basis of an increase in the note issue of theBank of England or to be presented to the Mint for coinage, and no means whereby it could be diminished(apart from the normal demand for the arts, amounting to about £2,000,000 a year, which was only partlytaken out of the currency supply) except the export of bullion or sovereigns.

3 Since the passing of the Act of 1844 there has been a great development of the cheque system. Theessence of that system is that purchasing power is largely in the form of bank deposits operated upon bycheque, legal tender money being required only for the purpose of the reserves held by the banks againstthose deposits and for actual public circulation in connection with the payment of wages and retailtransactions. The provisions of the Act of 1844 as applied to that system have operated both to correctunfavourable exchanges and to check undue expansions of credit.

4 When the exchanges were favourable, gold flowed freely into this country and an increase of legaltender money accompanied the development of trade. When the balance of trade was unfavourable and theexchanges were adverse, it became profitable to export gold. The wouldbe exporter bought his gold fromthe Bank of England and paid for it by a cheque on his account. The Bank obtained the gold from the IssueDepartment in exchange for notes taken out of its banking reserve, with the result that its liabilities todepositors and its banking reserve were reduced by an equal amount, and the ratio of reserve to liabilitiesconsequently fell. If the process was repeated sufficiently often to reduce the ratio in a degree considereddangerous, the Bank raised its rate of discount. The raising of the discount rate had the immediate effect ofretaining money here which would otherwise have been remitted abroad and of attracting remittances fromabroad to take advantage of the higher rate, thus checking the outflow of gold and even reversing the stream.

5 If the adverse condition of the exchanges was due not merely to seasonal fluctuations, but tocircumstances tending to create a permanently adverse trade balance, it is obvious that the procedure abovedescribed would not have been sufficient. It would have resulted in the creation of a volume of short-datedindebtedness to foreign countries which would have been in the end disastrous to our credit and the positionof London as the financial centre of the world. But the raising of the Bank’s discount rate and the stepstaken to make it effective in the market necessarily led to a general rise of interest rates and a restriction ofcredit. New enterprises were therefore postponed and the demand for constructional materials and othercapital goods was lessened. The consequent slackening of employment also diminished the demand forconsumable goods, while holders of stocks of commodities carried largely with borrowed money, beingconfronted with an increase of interest charges, if not with actual difficulty in renewing loans, and with theprospect of falling prices, tended to press their goods on a weak market. The result was a decline in generalprices in the home market which, by checking imports and stimulating exports, corrected the adverse tradebalance which was the primary cause of the difficulty.

6 When, apart from a foreign drain of gold, credit at home threatened to become unduly expanded, theold currency system tended to restrain the expansion and to prevent the consequent rise in domestic priceswhich ultimately causes such a drain. The expansion of credit, by forcing up prices, involves an increased

CUNLIFFE COMMITTEE REPORT 167

demand for legal tender currency both from the banks in order to maintain their normal proportion of cashto liabilities and from the general public for the payment of wages and for retail transactions. In this casealso the demand for such currency fell upon the reserve of the Bank of England, and the Bank wasthereupon obliged to raise its rate of discount in order to prevent the fall in the proportion of that reserve toits liabilities. The same chain of consequences as we have just described followed and speculative tradeactivity was similarly restrained. There was therefore an automatic machinery by which the volume ofpurchasing power in this country was continuously adjusted to world prices of commodities in general.Domestic prices were automatically regulated so as to prevent excessive imports; and the creation ofbanking credit was so controlled that banking could be safely permitted a freedom from state interferencewhich would not have been possible under a less rigid currency system.

7 Under these arrangements this country was provided with a complete and effective gold standard. Theessence of such a standard is that notes must always stand at absolute parity with gold coins of equivalentface value, and that both notes and gold coins stand at absolute parity with gold bullion. When theseconditions are fulfilled, the foreign-exchange rates with all countries possessing an effective gold standard aremaintained at or within the gold specie points.

Changes which have affected the gold standard during the war8 It will be observed that the fall in a number of the foreign exchanges below the old export specie points

which has taken place since the early part of 19151 is not by itself a proof that the gold standard has brokendown or ceased to be effective. During the present war the depredations of enemy submarines, highfreights, and the refusal of the government to extend state insurance to gold cargoes have greatly increasedthe cost of sending gold abroad. The actual export specie point has, therefore, moved a long way from itsold position. In view of our enormous demands for imports, coupled with the check on our exports due tothe war, it was natural that our exchanges with neutrals should move towards the export specie point.Consequently, the fall in the export specie point would by itself account for a large fall in our exchangerates. Such a fall must have taken place in the circumstances, even though all the conditions of an effectivegold standard have been fully maintained.

9 The course of the war has, however, brought influences into play in consequence of which the goldstandard has ceased to be effective. In view of the crisis which arose upon the outbreak of war it wasconsidered necessary, not merely to authorize the suspension of the Act of 1844, but also to empower theTreasury to issue currency notes for one pound and for ten shillings as legal tender throughout the UnitedKingdom. Under the powers given by the Currency and Bank Notes Act 1914, the Treasury undertook toissue such notes through the Bank of England to bankers, as and when required, up to a maximum limit notexceeding for any bank 20 per cent of its liabilities on current and deposit accounts. The amount of notesissued to each bank was to be treated as an advance bearing interest at the current Bank Rate.

10 It is not likely that the internal demand for legal tender currency which was anticipated at thebeginning of August 1914 would by itself have necessitated extensive recourse to these provisions. But thecredits created by the Bank of England in favour of its depositors under the arrangements by which theBank undertook to discount approved bills of exchange and other measures taken about the same time forthe protection of credit caused a large increase in the deposits of the Bank. Further, the need of thegovernment for funds wherewith to finance the war in excess of the amounts raised by taxation and by loansfrom the public has made necessary the creation of credits in their favour with the Bank of England. Thus,the total amount of the Bank’s deposits increased from, approximately, £56,000,000 in July 1914 to £273,000,000 on 28 July 1915 and, though a considerable reduction has since been effected, they now (15August) stand as high as £171,870,000. The balances created by these operations passing by means of


payments to contractors and others to the joint stock banks have formed the foundation of a great growth oftheir deposits which have also been swelled by the creation of credits in connection with the subscriptions tothe various war loans.2 Under the operation of these causes the total deposits of the banks of the UnitedKingdom (other than the Bank of England) increased from £1,070,681,000 on 31 December 1913, to £1,742,902,000 on 31 December 1917.

11 The greatly increased volume of bank deposits, representing a corresponding increase of purchasingpower and, therefore, leading in conjunction with other causes to a great rise of prices, has brought about acorresponding demand for legal tender currency which could not have been satisfied under the stringentprovisions of the Act of 1844. Contractors are obliged to draw cheques against their accounts in order todischarge their wages bill—itself enhanced on account of the rise of prices. It is to provide this currencythat the continually growing issues of currency notes have been made. The Banks instead of obtaining notesby way of advance under the arrangements described in paragraph 9 were able to pay for them outright bythe transfer of the amount from their balances at the Bank of England to the credit of the currency noteaccount and the circulation of the notes continued to increase. The government subsequently, by substitutingtheir own securities for the cash balance so transferred to their credit, borrow that balance. In effect, thebanks are in a position at will to convert their balances at the Bank of England enhanced in the mannerindicated above into legal tender currency without causing notes to be drawn, as they would have beenunder the prewar system, from the banking reserve of the Bank of England, and compelling the Bank toapply the normal safeguards against excessive expansion of credit. Fresh legal tender currency is thuscontinually being issued, not, as formerly, against gold, but against government securities. Plainly, given thenecessity for the creation of bank credits in favour of the government for the purpose of financing warexpenditure, these issues could not be avoided. If they had not been made, the banks would have beenunable to obtain legal tender with which to meet cheques drawn for cash on their customers’ accounts. Theunlimited issue of currency notes in exchange for credits at the Bank of England is at once a consequenceand an essential condition of the methods which the Government have found necessary to adopt in order tomeet their war expenditure.

12 The effect of these causes upon the amount of legal tender money (other than subsidiary coin) in bankreserves and in circulation in the United Kingdom are shown in the following paragraph.

13 The amounts on 30 June 1914, may be estimated as follows:

Fiduciary Issue of the Bank of England £18,450,000Bank of England notes issued against gold coin or bullion £38,476,000Estimated amount of gold coin held by banks (excluding gold coin held in the IssueDepartment of the Bank of England) and in public circulation

£123,000,000

Grand Total £179,926,000

The corresponding figures of 10 July 1918, as nearly as they can be estimated, were:

Fiduciary Issue of the Bank of England £18,450,000Currency notes not covered by gold £230,412,000Total Fiduciary Issues3 £248,862,000Bank of England notes issued against coin and bullion £65,368,000Currency notes covered by gold £28,500,000Estimated amount of gold coin held by banks (excluding gold coin held by IssueDepartment of Bank of England), say

£40,000,000


Grand total £382,730.000

There is also a certain amount of gold coin still in the hands of the public which ought to be added to thelast-mentioned figure, but the amount is unknown.

14 As Bank of England notes and currency notes are both payable at the Bank of England in gold coin ondemand this large issue of new notes, associated, as it is, with abnormally high prices and unfavourableexchanges, must have led under normal conditions to a rapid depletion, threatening ultimately the completeexhaustion, of the Bank’s gold holdings. Consequently, unless the Bank had been prepared to see all its golddrained away, the discount rate must have been raised to a much higher level, the creation of banking credit(including that required by the government) would have been checked, prices would have fallen and a largeportion of the surplus notes must have come back for cancellation. In this way an effective gold standardwould have been maintained in spite of the heavy issue of notes. But during the war conditions have notbeen normal. The public are content to employ currency notes for internal purposes, and, notwithstandingadverse exchanges, war conditions interpose effective practical obstacles against the export of gold.Moreover, the legal prohibition of the melting of gold coin, and the fact that the importation of gold bullionis reserved to the Bank of England, and that dealings in it are limited have severed the link which formerlyexisted between the values of coin and of uncoined gold. It is not possible to judge to what extent legaltender currency may in fact be depreciated in terms of bullion. But it is practically certain that there hasbeen some depreciation, and to this extent therefore the gold standard has ceased to be effective.

Restoration of conditions necessary to the maintenance of the gold standard recommended15 We shall not attempt now to lay down the precise measures that should be adopted to deal with the

situation immediately after the war. These will depend upon a variety of conditions which cannot beforeseen, in particular the general movements of world prices and the currency policy adopted by othercountries. But it will be clear that the conditions necessary to the maintenance of an effective gold standardin this country no longer exist, and it is imperative that they should be restored without delay. After the warour gold holdings will no longer be protected by the submarine danger, and it will not be possibleindefinitely to continue to support the exchanges with foreign countries by borrowing abroad. Unless themachinery which long experience has shown to be the only effective remedy for an adverse balance of tradeand an undue growth of credit is once more brought into play, there will be very grave danger of a creditexpansion in this country and a foreign drain of gold which might jeopardize the convertibility of our noteissue and the international trade position of the country. The uncertainty of the monetary situation willhandicap our industry, our position as an international financial centre will suffer and our generalcommercial status in the eyes of the world will be lowered. We are glad to find that there was no differenceof opinion among the witnesses who appeared before us as to the vital importance of these matters.

Cessation of government borrowings16 If a sound monetary position is to be re-established and the gold standard to be effectively maintained,

it is in our judgement essential that government borrowings should cease at the earliest possible momentafter the war. A large part of the credit expansion arises, as we have shown, from the fact that theexpenditure of the government during the war has exceeded the amounts which they have been able to raiseby taxation or by loans from the actual savings of the people. They have been obliged therefore to obtainmoney through the creation of credits by the Bank of England and by the joint stock banks, with the resultthat the growth of purchasing power has exceeded that of purchasable goods and services. As we havealready shown, the continuous issue of uncovered currency notes is inevitable in such circumstances. This


credit expansion (which is necessarily accompanied by an evergrowing foreign indebtedness) cannotcontinue after the war without seriously threatening our gold reserves and, indeed, our national solvency.

17 A primary condition of the restoration of a sound credit position is the repayment of a large portion ofthe enormous amount of government securities now held by the banks. It is essential that as soon as possiblethe state should not only live within its income but should begin to reduce its indebtedness. We accordinglyrecommend that at the earliest possible moment an adequate sinking fund should be provided out ofrevenue, so that there may be a regular annual reduction of capital liabilities, more especially those whichconstitute the floating debt. We should remark that it is of the utmost importance that such repayment ofdebt should not be offset by fresh borrowings for capital expenditure. We are aware that immediately afterthe war there will be strong pressure for capital expenditure by the state in many forms for reconstructionpurposes. But it is essential to the restoration of an effective gold standard that the money for suchexpenditure should not be provided by the creation of new credit, and that, in so far as such expenditure isundertaken at all, it should be under taken with great caution. The necessity of providing for ourindispensable supplies of food and raw materials from abroad and for arrears of repairs to manufacturingplant and the transport system at home will limit the savings available for new capital expenditure for aconsiderable period. This caution is particularly applicable to far-reaching programmes of housing andother development schemes.

The shortage of real capital must be made good by genuine savings. It cannot be met by the creation offresh purchasing power in the form of bank advances to the government or to manufacturers under governmentguarantee or otherwise, and any resort to such expedients can only aggravate the evil and retard, possiblyfor generations, the recovery of the country from the losses sustained during the war.

Use of Bank of England discount rate18 Under an effective gold standard all export demands for gold must be freely met. A further essential

condition of the restoration and maintenance of such a standard is therefore that some machinery shall existto check foreign drains when they threaten to deplete the gold reserves. The recognized machinery for thispurpose is the Bank of England discount rate. Whenever before the war the Bank’s reserves were beingdepleted, the rate of discount was raised. This, as we have already explained, by reacting upon the rates formoney generally, acted as a check which operated in two ways. On the one hand, raised money rates tendeddirectly to attract gold to this country or to keep here gold that might have left. On the other hand, bylessening the demands for loans for business purposes, they tended to check expenditure and so to lowerprices in this country, with the result that imports were discouraged and exports encouraged, and theexchanges thereby turned in our favour. Unless this twofold check is kept in working order the wholecurrency system will be imperilled. To maintain the connection between a gold drain and a rise in the rateof discount is essential to the safety of the reserves. When the exchanges are adverse and gold is being drawnaway, it is essential that the rate of discount in this country should be raised relatively to the rates ruling inother countries. Whether this will actually be necessary immediately after the war depends on whether pricesin this country are then substantially higher than gold prices throughout the world. It seems probable that atpresent they are on the whole higher, but, if credit expansion elsewhere continues to be rapid, it is possiblethat this may eventually not be so.

Continuance of differential rates for home and foreign money not recommended19 It has been argued before us that during the period of reconstruction and perhaps for many years

afterwards it will be possible and desirable, even though the exchanges are adverse, to keep money forhome industry substantially cheaper in this country than it is abroad and yet retain an effective gold standardby continuing the present practice of differentiating between home money and foreign money. It is held that


relatively low rates should be offered for home money and charged on domestic loans, while gold is at thesame time prevented from going abroad by the offer of high rates for foreign money. In our judgement, sosoon as the present obstacles in the way of international intercourse are removed, any attempt to maintain thisdifferentiation must break down because it would be impracticable to prevent people from borrowing at thelow home rate and contriving in one way or another to re-lend at the high foreign rate. This could only beprevented, if at all, by the maintenance of such stringent restrictions upon the freedom of investment afterthe war as would, in our opinion, be most detrimental to the financial and industrial recovery of this country.Even, however, if differentiation, as a postwar policy, were practicable, it would not, in our judgement, bedesirable. For the low home rate, by fostering large loans and so keeping up prices would continue toencourage imports and discourage exports; so that, even though the high rate offered for foreign moneyprevented gold from being drawn abroad, it would only do this at the cost of piling up an ever-growing debtfrom Englishmen to foreigners. It would be necessary at the same time to continue to pay for our essentialimports of raw materials by borrowing in the United States and elsewhere, instead of by increasing ourexports, thus imposing further burdens of foreign debt. This process could not continue indefinitely, and mustsooner or later lead to a collapse. We are, therefore, of opinion that the need for making money dear in theface of adverse exchanges cannot, and should not, be evaded by resort to differential rates.

Legal limitation of note issue necessary20 The foregoing argument has a close connection with the general question of the legal control of the

note issue. It has been urged in some quarters that in order to make possible the provision of a liberal supplyof money at low rates during the period of reconstruction further new currency notes should be created, withthe object of enabling banks to make large loans to industry without the risk of finding themselves short ofcash to meet the requirements of the public for legal tender money. It is plain that a policy of this kindis incompatible with the maintenance of an effective gold standard. If it is adopted there will be no checkupon the outflow of gold. Adverse exchanges will not be corrected either directly or indirectly through amodification in the general level of commodity prices in this country. On the contrary, as the issue of extranotes stimulates the conditions which tend to produce an advance of prices, they will become steadily moreand more adverse. Hence the processes making for the withdrawal of our gold will continue and nocounteracting force will be set in motion. In the result the gold standard will be threatened with destructionthrough the loss of all our gold.

21 The device of making money cheap by the continued issue of new notes is thus altogetherincompatible with the maintenance of a gold standard. Such a policy can only lead in the end to aninconvertible paper currency and a collapse of the foreign exchanges, with consequences to the wholecommercial fabric of the country which we will not attempt to describe. This result may be postponed for atime by restrictions on the export of gold and by borrowing abroad. But the continuance of such a policyafter the war can only render the remedial measures which would ultimately be required inevitably morepainful and protracted. No doubt it would be possible for the Bank of England, with the help of the jointstock banks, without any legal restriction on the note issue, to keep the rate of discount sufficiently high tocheck loans, keep down prices, and stop the demand for further notes. But it is very undesirable to place thewhole responsibility upon the discretion of the banks, subject as they will be to very great pressure in amatter of this kind. If they know that they can get notes freely, the temptation to adopt a lax loan policy willbe very great. In order, therefore, to ensure that this is not done, and the gold standard thereby endangered,it is, in our judgement, imperative that the issue of fiduciary notes shall be, as soon as practicable, oncemore limited by law, and that the present arrangements under which deposits at the Bank of England may beexchanged for legal tender currency without affecting the reserve of the Banking Department shall be


terminated at the earliest possible moment. Additional demands for legal tender currency otherwise than inexchange for gold should be met from the reserves of the Bank of England and not by the Treasury, so thatthe necessary checks upon an undue issue may be brought regularly into play. Subject to the transitionalarrangements as regards currency notes which we later propose, and to any special arrangements in regardto Scotland and Ireland which we may have to propose when we come to deal with the questions affectingthose parts of the United Kingdom, we recommend that the note issue (except as regards existing privateissues) should be entirely in the hands of the Bank of England; the notes should be payable in gold inLondon only, and should be legal tender throughout the United Kingdom.[…]

Summary of conclusions47 Our main conclusions may be briefly summarized as follows: Before the war the country possessed a

complete and effective gold standard. The provisions of the Bank Act 1844, operated automatically tocorrect unfavourable exchanges and to check undue expansions of credit (paras 2–7).

During the war the conditions necessary to the maintenance of that standard have ceased to exist. Themain cause has been the growth of credit due to government borrowing from the Bank of England and otherbanks for war needs. The unlimited issue of currency notes has been both an inevitable consequence and anecessary condition of this growth of credit (paras 8–14).

In our opinion it is imperative that after the war the conditions necessary to the maintenance of aneffective gold standard should be restored without delay. Unless the machinery which long experience hasshown to be the only effective remedy for an adverse balance of trade and an undue growth of credit is oncemore brought into play, there will be grave danger of a progressive credit expansion which will result in aforeign drain of gold menacing the convertibility of our note issue and so jeopardizing the internationaltrade position of the country (para. 15).

The prerequisites for the restoration of an effective gold standard are:

(a) The cessation of government borrowing as soon as possible after the war. We recommend that at theearliest possible moment an adequate sinking fund should be provided out of revenue, so that theremay be a regular annual reduction of capital liabilities, more especially those which constitute thefloating debt (paras 16 and 17).

(b) The recognized machinery, namely the raising and making effective of the Bank of Englanddiscount rate, which before the war operated to check a foreign drain of gold and the speculativeexpansion of credit in this country, must be kept in working order. This necessity cannot, and shouldnot, be evaded by any attempt to continue differential rates for home and foreign money after thewar (paras 18 and 19).

(c) The issue of fiduciary notes should, as soon as practicable, once more be limited by law, and thepresent arrangements under which deposits at the Bank of England may be exchanged for legal tendercurrency without affecting the reserve of the Banking Department should be terminated at theearliest possible moment. Subject to transitional arrangements as regards currency notes and to anyspecial arrangements in regard to Scotland and Ireland which we may have to propose when wecome to deal with the questions affecting those parts of the United Kingdom, we recommend thatthe note issue (except as regards existing private issues) should be entirely in the hands of the Bankof England. The notes should be payable in London only and should be legal tender throughout theUnited Kingdom (paras 20 and 21).

As regards the control of the note issue, we make the following observations:


1 While the obligation to pay both Bank of England notes and currency notes in gold on demand shouldbe maintained, it is not necessary or desirable that there should be any early resumption of the internalcirculation of gold coin (para. 23).

2 While the import of gold should be free from all restrictions, it is convenient that the Bank of Englandshould have cognizance of all gold exports and we recommend that the export of gold coin or bullionshould be subject to the condition that such coin and bullion has been obtained from the Bank for thepurpose. The Bank should be under obligation to supply gold for export in exchange for its notes (para.24).

3 In view of the withdrawal of gold from circulation we recommend that the gold reserves of the countryshould be held by one central institution and that all banks should transfer any gold now held by themto the Bank of England (para 25).

Having carefully considered the various proposals which have been placed before us as regards the basis ofthe fiduciary note issue (paras 26–31), we recommend that the principle of the Bank Charter Act 1844, shouldbe maintained, namely that there should be a fixed fiduciary issue beyond which notes should only be issuedin exchange for gold. The separation of the Issue and Banking Departments of the Bank of England shouldbe maintained, and the Weekly Return should continue to be published in its present form (para. 32).

We recommend, however, that provision for an emergency be made by the continuance in force, subjectto the stringent safeguards recommended in the body of the report, of section 3 of the Currency and BankNotes Act 1914, under which the Bank of England may, with the consent of the Treasury, temporarily issuenotes in excess of the legal limit (para. 33). We advocate the publication by the banks of a monthlystatement in a prescribed form (para. 34).

We have come to the conclusion that it is not practicable to fix any precise figure for the fiduciary noteissue immediately after the war (paras 35–9).

We think it desirable, therefore, to fix the amount which should be aimed at as the central gold reserve,leaving the fiduciary issue to be settled ultimately at such amount as can be kept in circulation withoutcausing the central gold reserve to fall below the amount so fixed. We recommend that the normal minimumof the central gold reserve to be aimed at should be, in the first instance, £150 million. Until this amount hasbeen reached and maintained concurrently with a satisfactory foreign exchange position for at least a year,the policy of cautiously reducing the uncovered note issue should be followed. When reductions have beeneffected, the actual maximum fiduciary circulation in any year should become the legal maximum for thefollowing year, subject only to emergency arrangements previously recommended. When the exchanges areworking normally on the basis of a minimum reserve of £150 million, the position should again be reviewedin the light of the dimensions of the fiduciary issue as it then exists (paras 40–2).

We do not recommend the transfer of the existing currency note issue to the Bank of England until thefuture dimensions of the fiduciary issue have been ascertained. During the transitional period the issueshould remain a government issue, but new notes should be issued, not against government securities, butagainst Bank of England notes, and, furthermore, when opportunity arises for providing cover for existinguncovered notes, Bank of England notes should be used for this purpose also. Demands for new currencywould then fall in the normal way on the Banking Department of the Bank of England (paras 43 and 44).

When the fiduciary portion of the issue has been reduced to an amount which experience shows to beconsistent with the maintenance of a central gold reserve of £150 million, the oustanding currency notes


should be retired and replaced by Bank of England notes of low denomination in accordance with thedetailed procedure which we describe (paras 45 and 46).

We have the honour to be,My Lords and Sir,

Your obedient Servants,(Signed) CUNLIFFE (Chairman)

C.S.ADDISR.E.BECKETT

JOHN BRADBURYG.C.CASSELS

GASPARD FARRERHERBERT C.GIBBS

W.H.N.GOSCHENINCHCAPER.W.JEANSA.C.PIGOU

GEO.F.STEWARTW.WALLACE

Notes

1 In the abnormal circumstances at the outbreak of war the neutral exchanges moved temporarily in our favourowing to the remittance home of liquid balances from foreign countries and the withdrawal of foreign credits.

2 This process has had results of such far-reaching importance that it may be useful to set out in detail the mannerin which it operates. Suppose, for example, that in a given week the government require £10,000,000 over andabove the receipts from taxation and loans from the public. They apply for an advance from the Bank of England,which by a book entry places the amount required to the credit of public deposits in the same way as any otherbanker credits the account of a customer when he grants him temporary accommodation. The amount is then paidout to contractors and other government creditors, and passes, when the cheques are cleared, to the credit of theirbankers in the books of the Bank of England—in other words is transferred from public to ‘other’ deposits, theeffect of the whole transaction thus being to increase by £10,000,000 the purchasing power in the hands of thepublic in the form of deposits in the joint stock banks and the bankers’ cash at the Bank of England by the sameamount. The bankers’ liabilities to depositors having thus increased by £10,000,000 and their cash reserves by anequal amount, their proportion of cash to liabilities (which was normally before the war something under 20 percent) is improved, with the result that they are in a position to make advances to their customers to an amountequal to four or five times the sum added to their cash reserves, or, in the absence of demand for suchaccommodation, to increase their investments by the difference between the cash received and the proportionthey required to hold against the increase of their deposit liabilities. Since the outbreak of war it is the secondprocedure which has in the main been followed, the surplus cash having been used to subscribe for Treasury Billsand other government securities. The money so subscribed has again been spent by the government and returnedin the manner above described to the bankers’ cash balances, the process being repeated again and again untileach £10,000,000 originally advanced by the Bank of England has created new deposits representing newpurchasing power to several times that amount. Before the war these processes, if continued, compelled the Bankof England, as explained in paragraph 6, to raise its rate of discount, but, as indicated below, the unlimited issueof currency notes has now removed this check upon the continued expansion of credit.

3 The notes issued by Scottish and Irish banks which have been made legal tender during the war have not beenincluded in the foregoing figures. Strictly the amount (about £5 million) by which these issues exceed the amount


of gold and currency notes held by those banks should be added to the figures of the present fiduciary issuesgiven above.


13Report

Macmillan Committee on Finance and Industry*

The international gold standard

35 To restore gold to its old position as an international standard of value was the avowed aim of currencypolicy for a period of six or seven years after the cessation of hostilities. This aim was endorsed by twoInternational Conferences, that of Brussels (1920) and Genoa (1922). Since the restoration of the goldstandard the main preoccupation of currency authorities has been with the manner in which it has workedand the extent to which, under the conditions actually experienced, it has been responsible for the presentdegree of international disequilibrium.

36 By an international gold standard is to be understood not identity of the currency arrangements of allthe countries comprising the gold standard group but the possession by all of them of one attribute incommon, namely that the monetary unit (i.e. pound sterling, dollar, franc, mark and so on) should possess agold value prescribed by law, or, rather, a gold value within the limits of the buying and selling price ofgold of the local central bank. In almost all countries gold coin has now been withdrawn from circulationand its place taken by paper representatives of the gold. The gold price of the paper representatives of goldis determined by the limits at which the central banks will give gold for paper, or paper for gold. Thus inGreat Britain an ounce of standard gold, containing eleven parts of gold to one part of alloy, is legallyequivalent to . This is the price at which the Bank of England must sell gold in exchange for its notes: theminimum price at which it must give its own notes in exchange for gold is £3.l7s.9d.

37 The legal determination by each of a group of countries of the gold equivalent of the monetary unit,though an essential, is not the only condition of an effective international gold standard. Something more isnecessary; namely, the right freely to import gold and to tender it in unlimited quantities to the central bank;and the converse right, freely and in unlimited quantities, to draw gold from the central bank and to exportthe gold so obtained. During and immediately after the war there were countries which, though nominally,were not actually, on the gold standard because the rights in question did not exist. Thus in Great Britain,although the holder of Bank of England or of currency notes was legally entitled to obtain gold in exchangefor notes, he was, under war conditions, unable in practice to export gold and was later by statute prohibitedfrom exporting it except under licence. Again, in Sweden, gold importers were not free to tender gold inunlimited quantities to the central bank: the Swedish krona therefore possessed a value superior to theweight of gold which it nominally represented.

* From Cmd. 3897, London, HMSO, 1931, pp. 18–24, 106–14, abridged.

38 The right to demand gold in exchange for local purchasing power, or local purchasing power inexchange for gold, is not, however, the only guarantee of the effective working of the gold standard. It issufficient if the local currency is exchangeable for another currency which is itself convertible into gold. Itis possible indeed to classify currency systems based upon the gold standard by the extent to which therelationship between the local monetary unit and gold is proximate or remote. First there are the systems inwhich the monetary unit is itself a gold coin; then come currency systems in which conversion takes theform of the exchange by the central bank of gold bars into local currency or vice versa; more remote still isthe connection when the conversion right takes the form of an exchange of local purchasing power intoforeign currency which is itself convertible into gold bars or gold coin. Nor is it absolutely necessary thatthe right to convert should be conferred by statute; it suffices if the currency authority never refuses to givegold or the equivalent at a fixed rate for the local currency, and vice versa.

Primary objective of the gold standard

39 The primary object of the international gold standard is to maintain a parity of the foreign exchangeswithin narrow limits; this has the effect of securing a certain measure of correspondence in the levels ofprices ruling all over the gold standard area. Incidentally it provides an objective test of the correspondenceof local currency policy with that of the rest of the world, whilst under it the currency mechanism itselfprovides the resources by which, in the event of temporary disequilibrium in the relation of one currency toanother, equilibrium can be restored. But although so long as an international gold standard exists, thecurrency authority of any gold standard country has always a simple test of the effect of its policy constantlybefore it, the disequilibrium between the local currency in relation to the rest of the world leading to a lossof gold may be due to deep-seated causes which, if not corrected by measures involving a definite change invalues in the country in question, would result in a permanent drain of gold, or alternatively it may be due toa merely temporary deficiency in the balance of payments, not in itself due to deep-seated currencyderangements. A sudden check to exports due to the failure of a crop in an exporting country, or a suddenweakening of markets in an importing country, or pressure on the foreign-exchange market due to a demandfor remittance in consequence of a rush of new long-term issues in the capital market, or the withdrawal ofshort-term balances from a money market in consequence of political unrest in the borrowing or lendingcountry, or indeed, in any part of the world, are instances in point. What is necessary is the possession ofsome one exportable commodity, the value of which will not fall, i.e., for which the short-period price is notsubject to change. Gold is such a commodity; so also is the exchange represented by a draft on a bank of thecreditor country. Gold and foreign-exchange reserves are easily mobilizable assets which are available atmoments of emergency, when perhaps nothing else is available. Thus, under the international gold standard,the currency system itself provides the resources by means of which equilibrium can be restored, so long asthe causes by which disequilibrium has been brought about are only temporary in character, and so long asthe international gold standard is itself maintained.

40 The fact that a state of temporary disequilibrium can be adjusted by flows of gold or foreign exchangedoes not remove from the currency authority the necessity of caution and of interpreting the situation in thelight of its fundamental duty of safeguarding the international standard. What appears at first sight to be acase of purely temporary disequilibrium may prove to be a symptom of a more deep-seated lack ofadjustment. Whilst gold exports need not be followed by more drastic remedial measures if thedisequilibrium is temporary, they are always a warning both to the money market and to the central bankitself that further action may later be necessary. Thus, whenever gold is lost, the central bank is providedwith an ‘automatic’ signal of the emergence of conditions which may make positive action necessary. The

178 MACMILLAN COMMITTEE REPORT

ultimate aim—the restoration of the international value of the currency—is clear, but the action to be taken,and the precise moment at which it should be taken, remain in the sphere of discretion and judgement, in aword with ‘management’. That the sphere of ‘management’ in this sense is wide and responsible is beyonddoubt. For if the wrong action is taken, the state of disequilibrium which the monetary authority designs tocorrect may be accentuated or perpetuated. If, in a state of temporary disequilibrium, the monetary authoritytakes action resulting in a change in the level of values here, a new state of disequilibrium may beunnecessarily created. The sense in which the gold standard can be said to be automatic is thus very limited;it is automatic only as an indicator of the need for action and of the end to be achieved.

Secondary objectives of the gold standard

41 It may be considered a secondary object of the international gold standard to preserve a reasonablestability of international prices. The mere existence of an effective international gold standard does not,however, guarantee stability of prices as a whole either over space or over time. In the nineteenth century,indeed, price movements in different gold standard countries showed a marked tendency to move together,but the absolute level of prices showed a considerable degree of long-period instability. And in the postwarperiod the sympathetic movement of prices over space is subject to a much greater degree of interferencethan was the case before the war, whilst the trend of prices over time has shown a very marked degree ofinstability. These circumstances call for investigation.

42 If prices are to be kept in approximate equality over space, two sets of conditions, which must bepresent simultaneously, are necessary. In the first place, countries which are losing gold must be prepared toact on a policy which will have the effect of lowering prices, and countries which are receiving gold mustbe prepared to act on a policy which will have the effect of raising prices. In the second place, the economicstructure (as distinct from the currency structure) must be sufficiently organic and sufficiently elastic toallow these policies to attain their objective. The first condition concerns central banking policy; the secondlimiting condition is not a question of monetary policy, but of the actual economic conditions in which suchpolicy has to work. In practice, it may easily happen that where the first condition is present, the second isabsent, and vice versa.

43 The nineteenth-century philosophy of the gold standard was based on the assumptions that (a) anincrease or decrease of gold in the vaults of central banks would imply respectively a ‘cheap’ or a ‘dear’money policy, and (b) that a ‘cheap’ or a ‘dear’ money policy would affect the entire price structure and thelevel of money-incomes in the country concerned. But, in the modern postwar world, neither of theseassumptions is invariably valid. The growth of the practice of central banks, by which gold inflows oroutflows are offset by the withdrawal or creation of bank credit—the movement of gold not being allowedto produce any effect on monetary conditions—involves not the absence of policy, but a policy inconsistentwith the rapid adjustment of relative money-incomes and prices. Thus, if at a time when gold flows freely tothe United States the Federal Reserve System offsets gold imports by sales of securities, thereby preventingcredit expansion, the level of American money-incomes and prices will not rise. If, at the same time, thelevels of European money-incomes and prices are already higher than international equilibrium justifies, thewhole burden of re-adjustment will be thrust exclusively on Europe. Again, if the level of British costs is outof line with international conditions, but gold exports are offset by the creation of fresh credit by the Bank ofEngland, the maladjustment will continue. Both the failure of American incomes and prices to rise, and thefailure of British incomes and prices to fall, may be deliberately intended, but in that case a policy ofstabilizing local values has been, implicitly or explicitly, substituted for a policy of maintaining a level ofvalues consistent with international equilibrium.


44 Moreover, the effect of the policy of a central bank is uncertain when the general economic structureis itself rigid, and more particularly so when the object of that policy is to reduce the price level and theincome structure. Pressure can be brought to bear upon the users of credit by a restriction of credit or theraising of Bank Rate, but that pressure cannot be directly brought to bear upon the costs of production. If theeconomic structure is rigid, then the effect may be to depress wholesale prices but to leave other pricesunchanged, and the process of restoring incomes and prices to an equilibrium level through gold flows andassociated bank policy may be interrupted or long delayed or even completely obstructed.

45 The degree to which the price level remains stable over a period of time is again profoundlyinfluenced by policy. If gold were the only form of currency, if there were no alternative uses for gold, andif no stocks of gold were held by central banks, the price level would be directly affected by the habits ofthe community on the one hand and the output of gold on the other. Actually, in the modern world, goldplays in the main only an indirect role in the determination of the price level, because the circulating mediaconsist overwhelmingly of paper money and bank deposits; it is this volume of purchasing power whichdirectly affects the price level and not the amount of gold which may be held in reserve. Gold itself affectsthe price level mainly through the decisions of the holders of gold reserves as to the amount of purchasingpower which they will allow to be outstanding against a given holding of gold. Central banks and currencyauthorities as a whole can increase their total reserves only to the extent of the new gold supplies availableeach year for currency purposes. An acute competition for gold in order to increase reserve ratios tends to areduction of the aggregate amount of purchasing power against which the gold is held, and to a fall in pricesadditional to the fall which would in any case have taken place if, in a period of falling supplies of gold, arigid relationship between gold supplies and additional supplies of purchasing power had been maintained,i.e., if reserve ratios had been left unchanged. In a period of expanding trade, production and population, adecreased gold supply, unless accompanied by a similar movement of reserve ratios, lowers prices. If, inaddition, certain countries take steps to obtain gold solely for the purpose of increasing the ratio of theirgold to their liabilities, the price level must fall still more. It falls more because the countries successful inthe struggle do not allow their additional supplies of gold to affect their prices, and because the countrieswhich are threatened by a loss of gold take steps to resist the loss which have the effect of lowering theirprices.

Conditions necessary for the working of the gold standard

46 The international gold standard is intended to subserve the general idea of stability—as regards therelations between currencies, and price relationships over space and time—but it does not, in and of itself,guarantee that this ideal will be realized. The international gold standard can, under appropriate conditions,enable both exchange stability and a considerable degree of price stability to be attained simultaneously,over a wide area, but, the mere fact that the standard is gold, and that it is international, will not under allconceivable conditions and varieties of policy, automatically bring about these results. In other words, thereare ‘rules of the game’, which, if not observed, will make the standard work with undesirable, rather thanbeneficial, consequences.

47 It is difficult to define in precise terms what is implied by the ‘rules of the game’. The management ofan international standard is an art and not a science, and no one would suggest that it is possible to draw upa formal code of action, admitting of no exceptions and qualifications, adherence to which is obligatory onperil of wrecking the whole structure. Much must necessarily be left to time and circumstance. Central banks,in whose hands the working of the international gold standard has been vested, no doubt largely recognizethat in many respects the purely traditional and empirical rules by which their conduct was guided in prewar


days are no longer adequate as guides of action in present-day conditions, and we consider that thefollowing principles would be generally accepted:

(i) The international gold standard system involves a common agreement as to the ends for which itexists.

(ii) It should be an object of policy to secure that the international gold standard should bring with itstability of prices as well as that it should guarantee stability of exchange.

(iii) Action by individual central banks which, by repercussions on the policy of the others, imperils thestability of the price level should, as far as possible, be avoided.

48 The direct influence of central banks upon the price level is exerted through an alteration of the volumeof purchasing power brought about by changes in Bank Rate or by variations in the volume of assets held bythem. But there is also an indirect influence exerted by central banks to which considerable importancemust be attributed. Changes in the level of Bank Rate affect the price, and therefore the yield, of investmentsecurities. In periods of low Bank Rate the price of fixed-interest-bearing securities tends to rise, and theiryield to fall. The supply of such securities is therefore gradually encouraged because the cost of borrowingis lower whilst at the same time the demand for them is increased once confidence has returned in anydegree. Long-term funds are thereby placed at the disposal of governments and entrepreneurs and theprocess of real capital formation is stimulated, with the result that a rise in general prices is engendered. Itmay even be held that the indirect effect of easier conditions in the money market is more important as anintermediate link in the chain of events which leads from acute depression to the gradual restoration ofconfidence than the direct effect of attempting an expansion of the volume of purchasing power. For acentral bank has no assurance that the net amount of purchasing power in existence will in fact be increasedby the steps which it may take to that end. Its own operations in the market, through purchases of securities,may for instance lead to a repayment to it of loans previously made by way of discounts; the net amount ofpurchasing power would then perhaps remain unchanged. The investment market, and particularly theinternational investment market, thus occupies a place of great importance among the agencies determiningprice changes. […]

The main objectives of the monetary system

The gold standard

The recent working of the gold standard

242 Unfortunately, the anticipations of those who were responsible for our return to the gold standard in1925 have to a large extent not been fulfilled. Whether these anticipations were justified at the time, andwhat other course was practically possible, are questions on which we do not all agree, but which it wouldbe unfruitful now to discuss. The six years since that act of policy have, for the reasons stated below, provedto be of a very abnormal character and the sacrifices which a return to gold at the old parity involved havenot been compensated by the advantages of international price stability which were anticipated.

243 The accomplished fact of the restoration of our currency to the prewar gold parity and itsmaintenance there for a period of six years creates, however, an entirely new situation; and it by no means


follows, even if the view be held, as it is by some of us, that a mistake was made in 1925, that theconsequences can be repaired by a reversal of policy in 1931.

244 Apart from the more general considerations relating to the gold standard which we discuss below,there have been two sets of difficulties in the way of its working to advantage in recent years, one of whichwe may expect, and the other of which we may hope, to be temporary.

245 The first set of difficulties has been caused by the fact that the various gold parities established by thecountries returning to the gold standard did not bear by any means the same relation in each case to theexisting levels of incomes and costs in terms of the national currency. For example, Great Britainestablished a gold parity which meant that her existing level of sterling incomes and costs was relatively toohigh in terms of gold, so that, failing a downward adjustment, those of her industries which are subject toforeign competition were put at an artificial disadvantage. France and Belgium, on the other hand, somewhatlater established a gold parity which, pending an upward adjustment of their wages and other costs in termsof francs, gave an artificial advantage to their export industries. Other countries provide examples of anintermediate character. Thus the distribution of foreign trade, which would correpond to the relativeefficiencies of different countries for different purposes, has been seriously disturbed from the equilibriumposition corresponding to the normal relations between their costs in terms of gold. This, however, has beena consequence of the manner in which the postwar world groped its way to back to gold, rather than of thepermanent characteristics of the gold standard itself when once the equilibrium of relative costs has been re-established, though, even after six years, this is not yet the case.

246 The second set of difficulties has resulted from the international lending power of the creditorcountries being redistributed, favourably to two countries, France and the United States, which have usedthis power only spasmodically, and adversely to the country, Great Britain, which was formerly the leaderin this field and has the most highly developed organization for the purpose. This redistribution of lendingpower has been largely due to the character of the final settlement of the war debts in which this country hasacquiesced. For although Great Britain suffered during the war a diminution of her foreign assets of somehundreds of millions, she has agreed to a postwar settlement by which she has resigned her own net creditorclaims, with the result, that on a balance of transactions, virtually the whole of the large annual sums duefrom Germany accrues to the credit of France and of the United States. This has naturally had the effect ofgreatly increasing the surplus of these two countries, both absolutely and relatively to the surplus of GreatBritain. The diminution in Great Britain’s international surplus, due to her war sacrifices remaininguncompensated by postwar advantages, has, however, been further aggravated recently by the adverseeffect on her visible balance of trade of the first set of difficulties just mentioned, namely, the differingrelationships betwen gold and domestic money-costs on which different countries returned to the goldstandard.

247 This redistribution of lending power need not, however, in itself have interfered with the working ofthe gold standard. The difficulties have arisen through the partial failure of the two recipients, during thelast two or three years, to employ the receipts in the way in which Great Britain had always employed hers,namely, either in the purchase of additional imports or in making additional foreign loans on long-term. Onthe contrary, they have required payment of a large part of their annual surplus either in actual gold or inshort-term liquid claims. This is a contingency which the normal working of the international gold standarddoes not contemplate and for which it does not provide.

248 But this set of difficulties, too, one may hope, though with less confidence, to be a temporaryphenomenon. Should it not prove so, we can scarcely expect the international gold standard to survive in itspresent form. If for any reason, however plausible from its own point of view, a creditor country, aftermaking all the purchases it desires, is unwilling to lend its remaining surplus to the rest of the world, there


can be no solution except the ultimate destruction of the export trade of the country in question through arelative reduction in the gold costs of other countries. If there were no international standard, but eachcountry had its own domestic currency subject to fluctuating exchanges, this solution would come about atonce. For in this event the exchanges of France and the United States, for example, would by now haverisen to so high a level relatively to the rest of the world that their exporters would have been driven out ofbusiness, so that their unlent surplus would have disappeared. According to the classical theory of the goldstandard, the same result should ensue, though more slowly and painfully, as a result of movements of goldinflating costs in those countries and deflating costs elsewhere. But in the modern world, where, on the onehand, inflows of gold are liable to be sterilized and prevented from causing an expansion of credit, whilst onthe other hand the deflation of credit set up elsewhere is prevented by social causes from transmitting itsfull effect to money-wages and other costs, it may be that the whole machine will crack before the reactionback to equilibrium has been brought about.

249 Unfortunately upon these two sets of abnormal difficulties there has supervened, starting in theUnited States, a business slump of a more normal type though of quite unusual dimensions, furtheraggravated, in the opinion of some, by being associated with the necessity for a transition from the high ratesof interest appropriate to the war and postwar period back towards the lower rates which were typical beforethe war.

250 Naturally the total result leads some people to question the desirability of adhering to an internationalstandard.

The question of an international standard

251 This brings us to the question whether adherence to an international standard may involve the paymentof too heavy a price in the shape of domestic instability. Many countries, both today and at former times,have found that such continued adherence involves a strain greater than they can bear. But these aregenerally debtor countries, the trade of which is concentrated on a narrow range of primary products subjectto violent disturbances of prices. If we leave aside the position today, experience does not show that acreditor country with diversified trade is liable to suffer undue domestic strain merely as the result ofadherence to an international standard. We are of opinion, therefore, that we should not be influencedmerely by the exigencies of the moment, if there is reason to believe that there may be importantcountervailing advantages on a longer view. If we need emergency measures to relieve the immediatestrain, we should seek them in some other direction.

252 In the particular case of Great Britain we believe that there are such advantages. One of our mostvaluable sources of income, indeed one of our most important export industries, is the practice ofinternational banking and associated services. Along with our shipping and our staple export industries thishas been for a long period past one of our main sources of wealth. It is by no means clear that the possibleadvantages to our export activities from the fact that a fluctuating exchange would automatically offset therigidity of money-incomes would balance the unquestionable loss to the first named; and we might be apoorer country on balance. It is not necessary, in order to reach this conclusion, to exaggerate the benefitswhich accrue to us from our international financial business. They are not so enormous as to outweigh allother considerations. For example, it is not likely that they have gone even a fraction of the way towardscompensating the losses of wealth through unemployment in recent years. It is not our case that industryshould be sacrificed to finance. It is, rather, that the benefits to industry from a fluctuating exchange wouldbe inadequate to compensate the losses in other directions. For whilst a fluctuating exchange would haveundoubted advantages in certain conditions, it would often be merely substituting one form of instability for


another. It would not be possible for a country so intricately concerned with the outside world as GreatBritain is to escape so simply from the repercussions of instability elsewhere.

253 There is, moreover, a further reason which cannot easily be weighed merely in the balance of ourown direct economic advantage and which weighs more heavily with us than any other. There is, perhaps,no more important object within the field of human technique than that the world as a whole should achievea sound and scientific monetary system. But there can be little or no hope of progress at any early date forthe monetary system of the world as a whole, except as the result of a process of evolution starting from thehistoric gold standard. If, therefore, this country were to cut adrift from the international system with theobject of setting up a local standard with a sole regard to our domestic situation, we should be abandoningthe larger problem—the solution of which is certainly necessary to a satisfactory solution of the purelydomestic problem—just at the moment, maybe, when, if we were able to look a little further forward, thebeginnings of general progress would be becoming visible.

254 We conclude, therefore, that we shall best serve the purpose for which we were set up, and have thegreatest hope of securing a sufficient general agreement to lead to action, if we base our recommendationson the assumption, which we hold justified, that the next phase of monetary policy must consist of awholehearted attempt to make the existing international standard work more satisfactorily. It is possible—though we believe that hard experience will teach them otherwise—that some countries may be unable or mayfail to work an international standard in a satisfactory way. But this is not yet proved, and it would beunwise for us, who have so much to gain by it, to give up the attempt to secure a sound internationalcurrency.

Devaluation

255 It has been represented to us that, without in any way departing from the principle or the practice ofadherence to an international standard, it is desirable for us in the national interest to do now what mighthave been accomplished with much less difficulty in 1925, namely, to revise the gold parity of sterling.Such a step is urged on the ground that, if we diminished the gold equivalent of the pound sterling by 10 percent thereby reducing our gold costs automatically by the same percentage, this would restore to our exportindustries and to the industries which compete with imported goods what they lost by the return to gold at afigure which was inappropriate to the then existing facts, and that it would also have the greater advantageof affecting all sterling costs equally, whether or not they were protected by contract.

256 We have no hesitation in rejecting this course. It is no doubt true that an essential attribute of asovereign state is a power at any time to alter the value of its currency for any reasons deemed to be in thenational interest, and that legally, therefore, there is nothing to prevent the British Government andParliament from taking such a step. The same may be said of a measure writing down all debts, includingthose owed by the state itself, by a prescribed percentage—an expedient which would in fact over aconsiderable field have precisely the same effect. But, while all things may be lawful, all things are notexpedient, and in our opinion the devaluation by any government of a currency standing at its par valuesuddenly and without notice (as must be the case to prevent foreign creditors removing their property) isemphatically one of those things which are not expedient. International trade, commerce and finance arebased on confidence. One of the foundation stones on which that confidence reposes is the general beliefthat all countries will seek to maintain so far as lies in their power the value of their national currency as ithas been fixed by law, and will only give legal recognition to its depreciation when that depreciation hasalready come about de facto. It has frequently been the case—we have numerous examples of recent years—that either through the misfortunes of war, or mistakes of policy, or the collapse or prices, currencies have


fallen so far below par that their restoration would involve either great social injustices or national effortsand sacrifices for which no adequate compensation can be expected. The view may be held that our owncase in 1925 was of this character. The British currency had been depreciated for some years. It wasobvious to the whole world that it was an open question whether its restoration to par was in the nationalinterest and there is no doubt in our minds as to our absolute freedom at that time to fix it, if it suited us, at alower par value corresponding to the then existing exchange. But it would be to adopt an entirely newprinciple, and one which would undoubtedly be an immense shock to the international financial world, ifthe government of the greatest creditor nation were deliberately and by an act of positive policy to announceone morning that it had reduced by law the value of its currency from the par at which it was standing tosome lower value.

257 Moreover, considering the matter from another point of view, in the environment of the presentworld slump the relief to be obtained from a 10 per cent devaluation might prove to be disappointing. It isnot certain that, with world demand at its present low ebb, such a measure would serve by itself to restoreour export trades to their former position or to effect a radical cure of unemployment. On the contrary, inthe atmosphere of crisis and distress which would inevitably surround such an extreme and sensationalmeasure as the devaluation of sterling, we might well find that the state of affairs immediately ensuing onsuch an event would be worse than that which had preceded it.

The prospects of the gold standard

258 The course of events in the last two years has had the effect of forcing a number of countries off thegold standard. But these are all debtor countries; and if matters continue as at present, it will be the debtorcountries of the world, and not a creditor country such as Great Britain, which will be the first to find thestrain unbearable. We consider that the leading creditor countries of the world should consult together toprevent matters from continuing as at present. In order that Great Britain may speak with authority in suchdiscussions, it is essential that her financial strength should be beyond criticism. This largely depends in thenear future on an increase of her surplus available for new foreign lending.

259 In this connection it may be worth while to summarize the result of certain information which wehave collected. For it is of a reassuring character and goes some way to answer certain criticisms whichhave been made, or doubts which have been entertained, regarding one aspect of the financial position ofthis country. In view of the large volume of foreign issues which have been floated in recent years inLondon and of the common belief that, in addition, British investors have made large purchases in theUnited States, partly of foreign bonds initially issued in New York and partly of American securities, whilstat the same time our surplus for new foreign lending has been diminished, it has been surmised that GreatBritain must have been financing some part of her new foreign investment on long term by means of anincrease, and perhaps a dangerous increase, in her short-term liabilities to foreign centres. We haveaccord ingly made it our business to collect for the first time as full a summary as possible both of foreignliquid resources in London and of British acceptances on foreign account. We take this opportunity of thankingthe Bank of England, which has done the detailed work on our behalf, and the numerous financialinstitutions which have willingly co-operated in supplying the information, since they have enabled us tofill what was perhaps the greatest gap in our previous knowledge of the financial commitments of thiscountry. We cannot claim that our figures are complete and they could doubtless be improved with furtherexperience; but we believe that they may cover, in quantity, nearly the whole field. The most important itemabout which we have no information is the total of sterling bills held in their own custody by foreign banks,and it may be that this item is a more fluctuating one than the items for which we have obtained figures. We


were prepared to find that these totals might give some support to the fears expressed above, but in fact theyare reassuring.

260 The figures may be sumarized as follows:

End of year Deposits and sterling bills heldin London on foreign account1

Sterling bills accepted onforeign account

Net liability of London

£ £ £1927 419,000,000 140,000,000 279,000,0001928 503,000,000 201,000,000 302,000,0001929 451,000,000 176,000,000 275,000,0001930 435,000,000 161,000,000 274,000,0001931 (March) 407,000,000 153,000,000 254,000,0001Exclusive of sterling bills held by foreign banks in their own custody.

261 We have not obtained figures prior to 1927. We think it probable that in the period between the returnto the gold standard in 1925 and the end of 1927 London’s net liability was increasing substantially, sincethe French balances abroad were mainly built up during that period. But these figures show that in the lastthree years, so far from there having been a large increase in London’s short-term liabilities, there has beena small decrease in their amount, so far as we have been able to ascertain it.1 The preliminary figures whichwe have obtained for the first quarter of the present year indicate that this tendency has been continued.

262 We have not obtained figures of British balances abroad, i.e., of British short-term claims on foreigncentres (which may be quite considerable in amount), or of various other items which would be required fora complete picture. It would be desirable to obtain these, as we recommend later [in a chapter notreproduced here].

263 It seems probable, therefore, that in spite of the reduction of our surplus, the whole of our netpurchases of foreign securities have been paid for out of our currently accruing surplus on income account.It is possible, we think, that this surplus may be somewhat larger than the usual estimate.

264 As regards the immediate situation, it is also interesting to note that the trade returns, unsatisfactorythough they are, bear out the conclusion that the worst strain of a situation such as the present falls on theraw-material countries. During the first quarter of this year the quantity of our exports fell off by more than30 per cent, whereas the reduction in the quantity of our imports was only 6 per cent. Nevertheless, as aresult of the catastrophic fall of raw-material prices, the visible balance of trade has been actually lessadverse to us than in recent years, the net position in terms of money moving £5 million in our favour, sothat less of our surplus under other heads (i.e. from foreign interest, shipping, etc.), is being required todayto finance our imports than in 1930 or in 1929.

265 The same point can be strikingly illustrated by what has happened in the case of the singlecommodity wheat. At the price prevailing in December 1930, the annual cost of our wheat imports would beabout £30 million less than it was in 1929, and £60 million less than in 1925. It is obvious what a largecontribution this single item represents to the national cost of supporting the present volume ofunemployment. It is a great misfortune both for us and for the raw-material countries that we should have agreat volume of unemployment through their inability to purchase from us as a result of the fall in the priceof their produce. But merely from the point of view of our balance of trade it is not to be overlooked that thelatter fact not only balances the former but may even outweigh it. We conclude that the underlying financial


facts are more favourable than had been supposed, and that Great Britain’s position as a creditor countryremains immensely strong.

Note

1 It is interesting to compare these figures with the comparable items for the United States. (The American returnsalso include other figures reducing their total net liability, which we have not yet been able to collect for GreatBritain.)


14The gold exchange standard

Ragnar Nurkse*

Among the various schemes and proposals which repeatedly crop up in discussions of monetary reform, onethat has enjoyed a wide appeal is the idea of an international currency system with exchange rates stable, asunder the gold standard, but with liquid foreign balances constituting the international means of settlementand the international monetary reserves. Gold, it has been suggested, could be dispensed with in such asystem not only as a means of international payment but also as a standard of value, especially if thecurrency or currencies in which the reserves were held were maintained reasonably stable in terms of goodsand services.

Monetary history has furnished many examples of the exchange standard principle. Indeed, the practicalapplication of this principle must find a place in any account, however condensed, of international monetaryrelations during the interwar period. The gold standard that was ‘restored’ in the 1920s was in the main a goldexchange standard. But an exchange standard need not be a gold exchange standard; the sterling area whichemerged from the currency chaos of the Great Depression in the 1930s is another important example of theexchange reserve system.

Origin and growth

The recommendations of the Genoa Conference

The adoption of a gold exchange standard was officially recommended by the Genoa Conference, which metin the spring of 1922 to consider the problems of financial reconstruction. This recommendation was basedon the view that there existed a shortage of gold, due both to a decline in current supply and to an actual orprospective increase in demand for monetary purposes.

The world output of gold declined by about one-third from 1915 to 1922 (see Appendix I) as a naturalresult of the general rise in prices during and after the war, which entailed a rise in gold-mining costs, andwhich was not accompanied by a corresponding rise in the price of gold.

At the same time, it was feared that a return to the gold standard would lead to a scramble for gold,pushing up the commodity value of gold through competitive deflation.

The Financial Commission of the Genoa Conference, presided over by Sir Robert Horne, the BritishChancellor of the Exchquer, therefore included the following resolution among the ‘Currency Resolutions’of its final report:

* From League of Nations, International Currency Experience, Geneva, League of Nations, 1944, pp. 27–46, abridged.

Resolution 9 These steps [balancing of budgets; adoption of gold as a common standard; fixing ofgold parities; cooperation of central banks, etc.] might by themselves suffice to establish a goldstandard, but its successful maintenance would be materialy promoted…by an internationalconvention to be adopted at a suitable time. The purpose of the convention would be to centralize andcoordinate the demand for gold, and so avoid those wide fluctuations in the purchasing power of goldwhich might otherwise result from the simultaneous and competitive efforts of a number of countriesto secure metallic reserves. The convention should embody some means of economizing the use ofgold by maintaining reserves in the form of foreign balances, such, for example, as the gold exchangestandard or an international clearing system.

In Resolution 11 of the report, various proposals were made to serve as a basis for the conventioncontemplated in Resolution 9. Besides stressing the need for balanced budgets, fixed gold parities, freeexchange markets, etc. Resolution 11 contained the folowing provisions:

1 …The maintenance of the currency at its gold value must be assured by the provision of an adequatereserve of approved assets, not necessarily gold.

2 When progress permits, certain of the participating countries will establish a free market in gold andthus become gold centres.

3 A participating country, in addition to any gold reserve held at home, may maintain in any otherparticipating country reserves of approved assets in the form of bank balances, bills, short-term securities,or other suitable liquid resources.

4 The ordinary practice of a participating country will be to buy and sell exchange on other participatingcountries within a prescribed fraction of parity of exchange for its own currency on demand.

5 The convention will thus be based on a gold exchange standard.The proposed convention failed to materialize; but the influence of the Genoa resolutions was

nevertheless considerable.

Prewar antecedents of the gold exchange standard

Before proceeding to show the practical effect of these recommendations, it may be well to observe that thegold exchange system was by no means invented at Genoa. It had been practised in many cases before1914. One example commonly quoted is the arrangement by which exchange between London andEdinburgh was regulated in the second half of the eighteenth century. Another example is the conventionconcluded in 1885 between the central banks of Denmark, Norway and Sweden.1 Of greater historicalimportance was Russia’s policy, adopted in 1894, by which exchange reserves initially acquired by loanwere held abroad—at first in Berlin and later also in other centres—and the government stood ready to buyand sell bills on the centres in question at fixed rates of exchange. The success of the Russian experimentwas widely noticed, and a short time later Austria-Hungary established a similar system.

Even in countries on the full gold standard, central banks prior to 1914 were in the habit of holding acertain amount of foreign exchange in addition to gold. In 1913 fifteen European central banks2 togetherheld about 12 per cent of their total reserves in the form of foreign exchange. In 1925 the percentage offoreign exchange in the total gold and exchange reserves of twenty-four European central banks, as shownin Appendix II, was 27 per cent; and in 1928 it rose to 42 per cent. Clearly the holding of foreign exchangeby central banks was on the whole much less extensive before 1914 than it became later. It was theexistence of a large private fund of mobile balances—it was the constant flow of equilibrating short-termcapital transfers effected by commercial banks, traders and arbitrageurs in response to small changes in


exchange and interest rates—which prior to 1914 created conditions similar to exchange standardarrangements and reduced the need for gold movements. After the monetary upheavals of the war and earlypostwar years, private short-term capital movements tended frequently to be disequilibrating rather thanequilibrating: a depreciation of the exchange or a rise in discount rates, for example, instead of attractingshort-term balances from abroad, tended sometimes to affect people’s anticipations in such a way as toproduce the opposite result. In these circumstances the provision of the equilibrating capital movementsrequired for the maintenance of exchange stability devolved more largely on the central banks andnecessitated a larger volume of official foreign-exchange holdings.

The examples just given of the pre-1914 gold exchange system were confined to Europe. But it wasoutside Europe that the system played its most important part. India, ever since 1898, has provided theclassic instance of the working of an exchange standard. The ‘Gold Standard Fund’ introduced by theUnited States in the Philippine Islands in 1903 is also a well-known case. Argentina and Japan operatedwhat amounted in practice to a sterling exchange system in the years before 1914. Outside Europe theapplication of the exchange standard principle made little further progress in the 1920s. The spread of thegold exchange standard on the lines recommended by the Genoa Conference was, in the main, a Europeandevelopment.3

Modification of central bank statutes

‘Adoption of the gold exchange standard’ meant, in the first instance, adoption of statutes permitting centralbanks to hold foreign exchange instead of gold in their legal reserves against notes in circulation or againstnotes and sight deposits. The first few years after the Genoa Conference were a period of great activity incentral banking legislation, and the new statutes adopted were strongly influenced by the Genoa resolutions.This was true in particular of the countries that restored their currencies with the assistance of the FinancialCommittee of the League of Nations, such as Austria (1922), Danzig (1923), Hungary (1924), Bulgaria(1926), Estonia (1927) and Greece (1928), where the central banks were authorized to hold the whole oftheir reserves in foreign bills and balances convertible into gold. Italy (1927) was among the countries thattook a similar course. In a number of other countries, however, and especially in the later 1920s, the statutesadopted tended to depart from this model, and required the maintenance of a certain proportion of gold inthe total legal reserve of gold plus foreign exchange. The proportion was fixed at 75 per cent in Germany(1924), 33 per cent in Albania (1925), 75 per cent in Belgium and Poland (1927), and 70 per cent inRomania (1929).

Altogether, the countries whose central banks at some time during the period 1922–31 were entitled to holdtheir legal reserves partly or wholly in foreign exchange form quite a long list, including Albania, Austria,Belgium, Bolivia, Bulgaria, Chile, Colombia, Czechoslovakia, Denmark, Ecuador, Egypt, Estonia, Finland,Germany, Greece, Hungary, Italy, Latvia, Peru, Poland, Portugal, Romania, Spain, Uruguay, USSR,Yugoslavia. In addition, there was a number of countries which practised an exchange standard but whichhad not yet established a central bank during this period: for instance, India, New Zealand, Argentina andVenezuela.

It would be misleading to overemphasize the role of central banks’ legal reserve regulations in theworking of the exchange standard system. Central banks which had the right to hold foreign exchange in thelegal reserve did not always make use of this right and preferred sometimes to hold gold. On the other hand,banks which did not have this right sometimes held large amounts of foreign exchange outside the legalreserve. The central banks of Japan, Australia and Norway, for example, were allowed to count nothing but

190 THE GOLD EXCHANGE STANDARD

gold as legal cover; yet they held considerable foreign balances and were (rightly) regarded as practising theexchange standard system in fact.4

If the legal reserve requirements of central banks had exercised an important influence on domestic creditpolicies, then the legal status of foreign-exchange holdings might have made an important difference. Butthe practical effect of the reserve requirements upon national credit policies in the period under review was,on the whole, not very striking. On the other hand, changes in central banks’ foreign assets mayoccasionally have influenced domestic credit policy even if they formed no part of the legal reserve.

Whatever the legal provisions, some central banks undoubtedly found it at times advantageous, inmeeting temporary fluctuations in the balance of payments, to operate a foreign exchange reserve outsidethe legal cover, especially when it was thought that changes in the assets held as legal cover might attractundue attention and provoke undesired psychological repercussions.

In the preceding paragraphs, the question whether the central bank has the power to acquire foreignexchange at all was not even put; such power was taken for granted. The curious case of France shows thatit cannot be taken for granted. Prior to 7 August 1926, the Bank of France was not entitled to purchase goldor foreign exchange at anything but the old prewar parities. A law of 7 August 1926 empowered the Bank toacquire foreign exchange and gold at market rates and to issue notes against these assets beyond the legalmaximum limit of the note circulation. The depreciation of the franc reached its low point in July 1926, anda de facto stability was maintained from December 1926 onwards. The Bank acquired in fact enormousamounts of foreign exchange. When the legal stabilization of the franc was carried out in June 1928, theBank was not permitted to count this foreign exchange as part of the 35 per cent minimum cover againstnotes and sight liabilities; and the law of 7 August 1926 was repealed, though the Bank was not preventedfrom continuing to hold the exchange acquired up to June 1928. France’s de facto adherence to theinternational exchange standard thus lasted less than two years. The important role played by France in thebreakdown of the system will make it necessary to take up her case in more detail at a later point.

Sources of central banks' exchange holdings

After the legal provisions had been made allowing a bank of issue to purchase foreign assets or to countthem as part of its legal reserve, the next task was to create conditions enabling the bank actually to acquiresuch assets. There were several ways in which this was done.

In the first place, various countries received foreign stabilization or reconstruction loans, mainly from theUnited States and the United Kingdom, the proceeds of which passed at least in part into the hands of thecentral banks. Austria and Hungary, for example, were able to raise such loans (amounting to about $100million and $45 million respectively) under the auspices of the League of Nations. Germany obtained astabilization loan of $200 million under the Dawes Plan, Belgium obtained a loan of $100 million, Polandone of $50 million. Italy received a stabilization credit of $125 million from a group of central banks led bythe Bank of England and the Federal Reserve Bank of New York.

Secondly, efforts were made by some countries to acquire foreign reserves by improving the currentbalance of payments; and in some cases this was attempted through a deflation of costs and prices. Suchdeflation may have tended to improve the trade balance, but its main influence on exchange reservesoperated more frequently through the capital account of the balance of payments. Deflation involved highinterest rates, thus tending to attract funds from abroad. Moreover, foreign capital was attracted at times bythe currency appreciation which frequently accompanied the deflation preceding the legal stabilization.Especially when the monetary authorities made known their intention to stabilize the exchange at a levelhigher than the current market value, or when such an intention was assumed to exist, there was an


incentive for bull speculation in the currency concerned. Italy, for instance, experienced a considerablecapital inflow of this speculative character during the eighteen months preceding the de jure stabilization inDecember 1927, when the lira was steadily appreciating. Similar movements of speculative funds wereobserved in Denmark and Norway in the two or three years of exchange appreciation prior to January 1927and May 1928 respectively, when the two countries legally restored the prewar gold parities of their currencies.

Even apart from such cases of deflation and exchange appreciation, interest differentials during the years1925–8 generally provided a sufficient inducement for large amounts of short- and long-term capital tomove, from New York and London in particular, to the countries adhering to the gold exchange standard.To the extent that the individual borrowers needed funds in domestic currency, it was the banking system inthese countries that came to hold the proceeds of private foreign loans in the form of liquid foreign-exchange reserves.

A further source of foreign exchange, as demonstrated particularly in the case of France, was therepatriation of private domestic capital funds, combined to some extent with an improvement of the currentbalance of payments through undervaluation of the currency. The repatriation of capital to France set in assoon as the exchange depreciation was arrested; it continued after de facto stabilization was achieved; and itwas stimulated by the ‘profit’ created by the devaluation on the conversion of foreign balances intodomestic currency and by the fear that this ‘profit’ might be reduced through an appreciation of thecurrency.

Functioning and breakdown

Fluctuations in foreign-exchange reserves prior to 1931

The operation of the gold exchange system is illustrated in Appendix II showing the year-to-year changes incentral banks’ foreign exchange and gold holdings in twenty-four countries during the period 1924–32. Thelast two years of this period are marked by a wholesale liquidation of foreign-exchange reserves, reflectingthe complete collapse of the gold exchange standard. The liquidation period will be dealt with later. For themoment we may briefly note certain features of the system in the earlier and more ‘normal’ years.

One feature worth noting is the number of countries where the gold stock was kept unchanged, and whereconsequently changes in the balance of payments so far as they affected the central bank showedthemselves exclusively in the foreign-exchange reserve. In Finland, as may be observed from Appendix II,the gold stock remained constant throughout the period 1924–32, while the foreign-exchange reserve, whichwas much larger than the gold stock, underwent wide fluctuations. A constant gold stock and a varyingexchange reserve may also be observed in Lithuania (up to 1930), in Portugal (from 1926 to 1931), and inLatvia and Norway (up to 1931). This fact is particularly remarkable in the case of Norway and Lithuania,where foreign assets were not eligible as cover against note circulation or sight deposit liabilities. Of certainother countries such as Sweden, Bulgaria and Yugoslavia, though their gold reserves did not remainabsolutely unchanged, it is likewise true to say that the fluctuating portion of their international monetaryreserves was made up predominantly of foreign liquid assets.

In the Netherlands and in Greece, as shown in Appendix II, a part of the gold stock was converted intoforeign exchange in 1925 and 1928 respectively. In other cases, however, such substitutions as occurredwere mostly in the opposite direction. As early as 1925, Germany shifted a large part of her reserves fromforeign exchange into gold. Hungary did the same in 1926. This tendency away from the gold exchangesystem was reinforced in 1928, when Italy and Poland converted some of their foreign exchange into gold.Even where such large or sudden shifts did not take place, the proportion of foreign exchange in the total


reserves was in some cases allowed gradually to decline, as in Austria (from 1925), Czechoslovakia (from1927) and Yugoslavia (from 1924 onwards).

In Germany, it will be remembered, the central bank was not legally qualified to hold more than one-fourth of its statutory reserve in the form of foreign exchange. Foreign assets held outside the statutoryreserve accounted for most of the fluctuations in the bank’s foreign-exchange holdings. The share of foreignexchange in the statutory reserve fell indeed far below 25 per cent in the years after 1926. In 1928 theproportion of all foreign assets of the Reichsbank to its total gold and exchange holdings fell to the low levelof 16 per cent, compared with 61 per cent in 1924. This sharp drop was only partly due to the conversion offoreign balances into gold in 1925. When the bank’s total reserve increased in 1926, that increase was takenexclusively in the form of gold; when an outflow occurred in the following year, the reduction affectedexclusively the foreign-exchange reserve. The proportion of foreign exchange declined further in the springof 1929. When after the signature of the Young Plan a temporary resumption of the capital influx occurredin the second half of 1929, the bank took advantage of this to replenish its foreign exchange holdings not onlywithin but also outside the statutory reserve, and so to recover a margin of free play between thefluctuations of the balance of payments and the internal currency supply.

But the size of the international monetary reserve in the hands of the Reichsbank was not only a functionof the balance of payments. It depended largely also on the behaviour of the private money market inGermany and particularly on the behavior of the commercial banks. These banks, as a result of Germany’sforeign borrowing, kept substantial liquid resources abroad as working balances. Variations in theirpreference for foreign balances as against domestic central bank funds—determined by changes in interestdifferentials and exchange rates, by fluctuations in trade activity or by the state of confidence—were animportant factor influencing the movement of the Reichsbank’s exchange reserve and the size of thedomestic credit base. Similar conditions prevailed in other European countries that were borrowing heavilyin New York and elsewhere during the period under review.

The striking increase in Italy’s central exchange reserve in 1926 and 1927 was largely due, as mentionedbefore, to a speculative inflow of funds induced by the upward tendency in the value of the lira. After thestabilization in December 1927, the exchange reserve was steadily reduced. A part of the decline, however,was due to purchases of gold, especially in 1928. In spite of the wide fluctuations observed in individualcountries, the proportion of foreign assets in the total central-bank reserves of twenty-three Europeancountries, excluding France, was surprisingly steady in the six years prior to 1931. As may be seen from thelast line of table 14.1, the ratio varied only slightly between 35 and 40 per cent. It is the inclusion of

Table 14.1 Foreign exchange and gold reserves of European central banks ($ million)

End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

Total (24 countries)

Foreign exchange 845 917 1,159 2,145 2,520 2,292 2,300 1,216 505

Gold 2,281 2,367 2,568 2,903 3,490 3,841 4,316 5,273 5,879

Total 3,126 3,284 3,727 5,048 6,010 6,133 6,616 6,489 6,384

Foreign exchange as % of total 27 28 31 42 42 37 35 19 8

Total excluding

France (23 countries)

Foreign exchange 831 904 1,043 1,295 1,233 1,271 1,273 374 329

Gold 1,571 1,656 1,857 1,949 2,236 2,208 2,217 2,574 2,622

Total 2,402 2,560 2,900 3,244 3,469 3,479 3,480 2,948 2,951


End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

Foreign exchange as % of total 35 35 36 40 36 37 37 13 11

France that produces the wide fluctuations in this ratio as calculated for a total of twenty-four countries.

France on the gold exchange standard

At the end of 1928, France alone accounted for more than half of the total of central-bank exchangeholdings shown in the first line of table 14.1. The role of France in the rise and decline of the gold exchangestandard clearly demands a more detailed consideration.

For several years prior to July 1926, France experienced a flight of capital, which depreciated theexchange far below the level corresponding to domestic costs and prices. There were practically no goldexports, the Bank of France preferring to keep its relatively modest gold reserve intact. It was the surplus inthe current balance of payments which arose as a result of the exchange depreciation that provided themeans by which the real transfer of the flight capital was accomplished.

In July 1926, drastic measures were initiated to balance the budget and to restore confidence. The valueof the franc recovered from less than US cents in July to nearly 4 US cents in December 1926, at whichlevel it was stabilized, at first de facto and later, in June 1928, de jure.

In August 1926, as mentioned earlier, the Bank of France was granted the right to purchase gold andforeign exchange at market rates. During the next four or five years, in order to prevent the franc fromappreciating above the level established in December 1926, the Bank had to purchase enormous amounts ofgold and foreign assets. From August 1926 to the date of the legal stabilization—25 June 1928—the Bank’sforeign-exchange acquisitions, amounting to over 26,000 million francs (or over $1,000 million at thecurrent rate of exchange), by far exceeded its purchases of gold, amounting to about 10,000 million francs.After that date, its gold stock rose steadily up to 1932. Under the monetary law of 25 June 1928, the Bank wasno longer permitted to buy foreign exchange. Just before the legal stabilization, the Bank had bought largeamounts of foreign exchange for forward delivery. These forward contracts matured in the second half of1928, raising the total foreign-exchange holdings of the Bank to 32,800 million francs (or nearly $1,300million) at the end of the year. There was some doubt as to whether, under the law just mentioned, the Bankwas justified in thus increasing its holdings. The amount in question was therefore converted into gold, andthe foreign-exchange reserve fell back to the previous level of 26,000 million francs in June 1929. At thatlevel it remained virtually unchanged for over two years, that is until the second half of 1931, when theBank started the liquidation process which practically wiped out its foreign-exchange reserve in the twofollowing years.

It is necessary to distinguish clearly between the two principal sources of the gold and exchange holdingsacquired by the Bank after August 1926, even though the data permit of no sharp separation in the actualstatistics. There was, in the first place, the ‘repatriation of capital’. The desire of French investors torepatriate their funds was natural, in view of the temporary and abnormal character of the preceding capitalexport. In fact, however, in the then existing condition of the balance of payments the ‘repatriation’ meantessentially that the total of foreign assets held by the French public and banking system remainedunchanged, but that an increasing proportion of this total was transferred from the hands of private banksand capitalists into the ownership of the Bank of France.

The second factor at work tended to increase the total of French foreign assets: it was the ‘undervaluation’of the franc, the function of which, prior to July 1926, had been to effect the real transfer of flight capital


abroad. This undervaluation, though appreciably reduced through the recovery of the franc in the latter halfof 1926, persisted in a substantial degree after December 1926, creating as it were an automatic export ofcapital through a surplus in the current balance of payments. Having lost a large part of her long-termforeign assets through the Russian Revolution, France was not prepared to resume foreign long-terminvestment on any considerable scale, and so the current surplus went in the main simply to increase thecountry’s gold reserves and liquid foreign balances.

A third source may have operated at certain times to swell the gold and exchange holdings of the Bank ofFrance: namely, imports of foreign speculative funds. But these do not, in retrospect, seem to have beennearly as important as they appeared to certain contemporary observers (including the authors of the Bank’sannual reports for 1927 and 1928). Most of the speculative positions built up by foreigners during themonths preceding the legal stabilization were rapidly and easily liquidated after June 1928.5

Foreign speculation played, however, an important part in the reasons given by the Bank for convertingsome of its foreign exchange into gold. In its annual reports for 1927 and 1928, the Bank argued that itspurchases of foreign balances and bills created an abnormal condition of liquidity in the centres where theywere held; that it thus itself provided the funds used for speculative purchases of francs; and that only theconversion of its foreign assets into gold could stop the vicious inflationary circle.

But in so far as the Bank’s foreign-exchange purchases represented ‘repatriation of capital’ theyconstituted in fact simply a redistribution inside France in the ownership of foreign assets. The fact that billsor bank deposits in London, for example, passed from the ownership of a private Frenchman into theownership of the Bank of France cannot have had appreciable effects, if any, on market conditions inLondon. It is difficult to see, more particularly, how it could have increased the financial facilities forexchange speculation in London. True, if the private Frenchman’s foreign assets consisted of long-termsecurities, then—since the only foreign assets the Bank of France was prepared to hold were short-term billsand bank balances—the ‘repatriation’ of French capital would have meant a shift of funds from the long-term to the short-term market in London; but the resources of the London market as a whole could hardlyhave been substantially affected. Besides, the private French assets in London or New York were in factmostly held in a highly liquid form, as they had come there in search of security rather than profit.

To the extent that the Bank of France’s purchases of foreign bills and balances reflected the ‘automatic’capital export resulting from the under-valuation of the franc, there was likewise a change in the ownershipof short-term bills and bank deposits in, say, London; but this time from British ownership into theownership of the Bank of France. How this could have had an inflationary effect in London is difficult tosee. The effect is more likely to have been deflationary, especially in the case of bank deposits passing fromthe active domestic circulation into the inactive holding of the Bank of France.

The French funds that had taken flight and were temporarily held abroad represented an amount of finitemagnitude. Their transfer to the Bank of France—the process of ‘repatriation’—was bound to come to anend sooner or later. There was no such definite limit to the current surplus in the balance of paymentsarising from the undervaluation of the franc. How long this surplus was to continue depended on the speedand strength of the corrective forces tending to restore equilibrium in the balance of payments. Oneequilibrating factor, indeed, was to some extent neutralized. The increase in the gold and exchange reserveof the Bank of France was in part offset by a reduction in the Bank’s domestic assets and an increase in itsdeposit liabilities to the government, so that the growth of the gold and foreign-exchange reserve did notproduce an equivalent expansion in the note circulation and in private sight deposits at the Bank. Theadjustment of a country’s balance of payments, however, does not depend solely on changes in the quantityof domestic currency and credit in accordance with changes in international currency reserves; it alsodepends largely on changes in income and effective demand directly connected with the balance of


payments. Briefly, the low exchange value of the franc tended to divert expenditure of Frenchmen as wellas foreigners from foreign goods and services to French goods and services, and this led to an increase inaggregate income and demand in France tending to bring the international accounts into balance. In fact theFrench surplus in current transactions disappeared after 1930; the franc ceased to be undervalued. Theeffect of the undervaluation on economic conditions had been favourable in France, but unfavourableabroad; the French balance of payments no doubt contributed, though probably to a minor degree, to theforces making for depression in the rest of the world at the turning-point of the business cycle in 1929–30.

After June 1928, as stated before, the Bank of France was no longer entitled to buy foreign exchange. Inits annual reports for 1929 and 1930, written at a time when it was no longer possible to justify the purchaseof gold in place of foreign exchange by a desire to curb speculative and inflationary tendencies abroad, theBank referred to its gold imports as a natural result of the automatic gold standard mechanism. In the latterpart of 1931, the Bank began to liquidate its foreign bills and balances. It should be noted that nearly half ofthe reduction shown in the Bank’s foreign assets during 1931 represents the exchange loss incurred on itssterling assets. In 1932 the conversion of foreign exchange into gold continued at an accelerated pace. In1933 the foreign-exchange reserve was reduced to an insignificant amount; but in that year the gold stockalso suffered a reduction, the first of a long series of reductions.

In its annual reports for 1931 and 1932 the Bank declared that ever since 1928 its desire and intention hadbeen to convert its foreign assets into gold, and that it had refrained from doing so only out of regard for the‘monetary difficulties of other countries’.6 This indeed was the essence of the French position: from theoutset, France was a reluctant member of the gold exchange system; she regarded her partial adhesion to itas an essentially temporary makeshift and longed to give her currency ‘an exclusively metallic foundation’.7

The breakdown of the gold exchange standard

The fate of the gold exchange standard was sealed when France decided in 1928 to take nothing but gold insettlement of the enormous surplus accruing to her from the repatriation of capital and from the currentbalance of payments. The French gold imports certainly aggravated the pressure of deflation in the rest ofthe world and especially in London. In London, the pressure became unbearable in the end, and the goldparity of the pound was abandoned. When in the summer of 1931 French capitalists and commercial banksbecame nervous about their sterling assets and anxious to ‘repatriate’ them, the Bank of France could nottake over these assets and hold them in its exchange reserve; they had to be turned into gold. Even withoutthe law of June 1928 it may be that the Bank, sharing the public’s nervousness, would still have demandedgold. In certain other countries which were also repatriating their sterling balances at that time, centralbanks were legally free to acquire these balances, but in fact preferred gold to sterling. Hence the heavygold losses of the Bank of England in the three months preceding the suspension of the gold standard on 21September 1931.

The breakdown of exchange stability led in turn to a further scramble for gold through the liquidation ofpreviously accumulated foreign-exchange reserves of central banks, since the possibility of losses arisingfrom exchange-rate fluctuations rendered the holding of foreign balances risky. Indeed the withdrawal ofbalances from the United Kingdom after September 1931 was often imposed on central banks by theirstatutes, requiring them to hold their exchange reserves exclusively in gold standard currencies. To remaineligible as legal cover, these balances therefore had to be converted into gold or into a gold standardcurrency. As the dollar was also under pressure, a number of countries—especially in eastern and south-eastern Europe—transferred their reserve balance from London to Paris. Paris, in fact, became for a time a


minor gold exchange centre for such countries as Poland, Czechoslovakia and Bulgaria, especially afterApril 1933 when the dollar, too, ceased to be eligible for reserve purposes.

The crisis of September 1931 was followed by a withdrawal of balances from the United States as well asfrom the United Kingdom. In the two years 1931 and 1932, the monetary gold stock of the United Stateswas reduced from $4,225 to $4,045 million, or by $180 million. In the United Kingdom, the net reductionamounted to $125 million (from $718 to $583 million), making a reduction of $315 million for the twocountries together during these years. Over the same period, however, the combined gold reserves of sixcreditor countries, namely France, Belgium, Italy, the Netherlands, Sweden8 and Switzerland, rose by asmuch as $1,929 million (from $2,943 to $4,872 million). Thus only a small part of the gold absorbed bythese countries came from the two centres where most of their foreign-exchange reserves were held. Wheredid the remainder come from? Though some of it came from current gold production, it is clear that to acertain extent the pressure resulting from the collapse of the gold exchange system was passed on fromLondon and New York to the world’s debtor countries, mainly through the abrupt cessation or even reversalof capital movements and through the fall in prices of primary commodities.9

If we separate the twenty-four countries of Appendix II into two groups—creditors and debtors—we seeat once that the conversion of foreign-exchange reserves into gold was confined to the six creditor states.From the end of 1930 to the end of 1932 their gold reserves, as shown in table 14.2, increased by $1,929million, while their foreign-exchange holdings declined by $1,331 million. The tendency to replace foreignassets by gold in this group of countries had already become apparent in 1928/9.

The eighteen debtor states, as shown in table 14.2, suffered a heavy

Table 14.2 Central banks’ foreign exchange and gold reserves ($ million)

End of: 1928 1929 1930 1931 1932

Total of 6 creditor countries

Foreign exchange 1,878 1,604 1,679 1,024 348

Gold 1,987 2,430 2,943 4,214 4,872

Total 3,865 4,034 4,622 5,238 5,220

Foreign exchange as % of total 49 40 36 20 7

Total of 18 debtor countries

Foreign exchange 642 688 621 192 157

Gold 1,503 1,411 1,373 1,059 1,007

Total 2,145 2,099 1,994 1,251 1,164

Foreign exchange as % of total 30 33 31 15 13

reduction in gold reserves as well as in foreign exchange, though the proportionate decline was considerablygreater in foreign exchange than in gold. They lost $366 million of gold in 1931 and 1932. Debtor countriesoutside Europe lost about $260 million of gold in this period.

While most of the gold reserves lost by the debtor countries were in effect passed on to creditor countriesconverting their dollar and sterling balances into gold, the foreign-exchange reserves lost by the debtorcountries were largely used to repay short-term credits called in by the reserve centres themselves (i.e.,mainly New York and London). Just as the granting of private short-term credits by New York and Londonhad supplied the central banks of debtor countries with a large part of their dollar and sterling reserves in the1920s, so the withdrawal of those credits in the early 1930s tended to wipe out these reserves. It is clear thatwhen debtor countries used up their foreign-exchange reserves for payments to the centres in which the


reserves were held, there occurred an extinction and not simply a transfer of central banks’ internationalcurrency reserves.

Thus the breakdown of the gold exchange standard involved a sharp reduction in the aggregate ofinternational currency reserves not only through the conversion of exchange reserves into gold but alsothrough the absorption of exchange reserves by payments to the reserve centres. The total foreign-exchangereserves of the twenty-four countries included in Appendix II declined by $1,800 million in 1931 and 1932—that is, by an amount far in excess of the $1,000 million of gold which came into central bank reserves inthe world as a whole from current production and other sources in these two years.

Merits and defects of the system

The liquidation of the gold exchange standard and the scramble for gold which it implied—more especiallythe large-scale absorption of gold by the countries that later came to be known as the ‘gold bloc’—gaveadded strength to the forces of deflation throughout the world. It was chiefly to avoid deflation that the goldexchange standard was recommended at Genoa. As events turned out, the deflation was only postponed; theprinciple of ‘gold economy’ was abandoned when it was most needed, at a time, namely, when other factorsmaking for depression were coming into play in any case.

The gold exchange standard was often accused of tending to breed inflation.10 In actual fact, of course, nogeneral rise in prices occurred in the period from 1924 to 1928 when the gold exchange standard was infairly extensive operation; on the contrary, prices of primary commodities showed a falling trend from 1926onwards. Yet the charge was perhaps true in a sense, but it betrayed a misconception of the primary objectof the system. The gold exchange standard was intended to be an anti-deflationary device and therefore inthat sense ‘inflationary’. Without it, the shortage of international currency might have led to a generaldeflation which would have ‘corrected’ the situation through a reduction in the value of internationaltransactions and an increase in the output of new gold. With it, the gold shortage was made good byexchange reserves; gold production could remain lower than it otherwise would have been, and there wasthus an ‘economy of gold’ even in the sense of an economy of productive resources engaged in gold-mining.

One serious weakness of the gold exchange system was the great variability in the degree to whichcentral banks relied on exchange reserves as against gold. To understand the causes of this variability it isnecessary to consider the motives that led individual countries to adopt the gold exchange standard and toadhere to it. The principal motive should have been a realization of the common interest or, in negative terms,the fear of a worldwide deflation. But individual countries were inclined to neglect the externalrepercussions of their actions; and the threat of a world deflation was, to each of them, a remote andineffective sanction.

Countries in need of foreign capital could be induced to observe the exchange standard rules by variousmeans of persuasion and pressure exerted by the lending centres. Especially in the countries where postwarmonetary reconstruction was carried out with outside assistance, the advice given and the desires expressedby the experts and financiers of the lending countries played an important role in the practical working ofthe gold exchange standard.

On the other hand, this factor tended to discredit the system in the eyes of certain creditor and also certaindebtor countries. The holding of foreign balances instead of gold in the central monetary reserve came to beregarded as damaging to the prestige of a great or even a moderately great nation. It is largely for this reasonthat the countries whose balances were in absolute amount the most important—including, for instance,France, Germany, Italy and Poland—did not regard their own use of the gold exchange standard as anythingbut a transitory expedient.11


Moreover, in certain quarters the gold exchange standard was regarded as merely a British ‘fad’,12 oreven as a device invented and sponsored by Great Britain to make it easier for her to return to the prewarparity without the necessary internal adjustments in costs and prices, and to retain her gold reserves. It has beenobserved13 that at the Genoa Conference the exchange standard doctrine was propounded mainly by theBritish delegation.

The only tangible bait offered by the gold exchange standard to its adherents was the interest returnobtainable on foreign-exchange reserves as opposed to gold. The ‘heavy expense of the gold standardsystem’14 in terms of the interest forgone was supposed to render the gold exchange method particularlyattractive to the poorer countries. That may have been so; but the argument did not appeal to the nationalpride of the less poor countries, even though it may have appealed to the accountants of their central banks.In the profit-and-loss returns of central banks interest receipts from foreign bills and balances playedsometimes indeed a very prominent part. Thus in 1929 they provided the Bank of France with 65 per cent ofits gross profits. How drastic a change the liquidation of its exchange holdings entailed in the Bank’s profit-and-loss account may be seen from the following comparison of the position in 1929 with that in 1934 whenreceipts from foreign assets had fallen to 3 per cent of gross profits:

Francs (million): 1929 1934Receipts from foreign assets 1,250 20Receipts from domestic assets, commissions, etc. 684 547Total gross profits: 1,934 567

The 1920s were a period of relatively high interest rates. In the leading financial centres, 4 per cent, 5 percent or even more could be obtained from three-month bank bills and similar short-term investments. In1929 the Bank of France earned an average of 4.6 per cent on its foreign bills and bank balances. Just as thehigh level of money rates in the 1920s may have contributed in some degree to the spread and themaintenance of the gold exchange standard, so the sharp decline that took place in 1930 may have tended toweaken people’s attachment to it. In London and New York rates of discount on bank drafts or acceptancesdropped from over 5 per cent in the summer of 1929 to about 2 per cent in the second half of 1930. In 1931,however, with the onset of the international liquidity crisis, money rates shot up again; yet the inducement ofhigher interest was powerless to stop the general flight from the gold exchange standard.

Interest returns became a wholly secondary consideration when stability of exchange rates andconfidence in such stability broke down. The depreciation of a currency in which foreign-exchange reserveswere held meant that these reserves lost some of their value as means of settlement in relation to othercountries, even though their power to purchase commodities or to discharge financial obligations in thecountry in which they were held was not necessarily impaired by the exchange depreciation. The fear ofexchange losses was in fact an overpowering motive for the liquidation of foreign reserves and rendered theoperation of an exchange standard quite impossible. Besides, the statutes of most central banks, asmentioned before, expressly required that if any exchange standard were practised at all, it should be a goldexchange standard; and the liquidation of foreign-exchange reserves became at least in some measurecompulsory as soon as the currency of a reserve centre ceased to be convertible into gold at a fixed parity.

One criticism which has figured prominently in discussions of the gold exchange standard has still to beconsidered. It has been asserted that movements in foreign-exchange reserves do not operate in the samereciprocal fashion as gold movements. Under the gold bullion standard, according to the traditional view,the country losing gold is forced to deflate and the country gaining gold is induced to inflate, so that theburden of any necessary adjustment is shared and therefore eased. Under the gold exchange standard, it is


argued, a country that is gaining or losing foreign-exchange reserves may well be obliged to effect theappropriate expansion or contraction of credit; but in the country where these reserves are held nothinghappens that would bring into play a reciprocal tendency towards contraction or expansion. This argumentcalls for three brief comments.

In the first place, whether the country where the reserves are held is or is not directly affected by changesin their volume depends to some extent on the form in which they are held. If they were held in the form ofsight deposits with the central bank, such changes would have a strong effect analogous to gold movements,since they would involve additions to or withdrawals from the amount of central-bank funds available fordomestic credit base. If the reserves were held in the form of deposits with commercial banks, as was morecommonly the case, they would involve shifts of deposits from the active domestic circulation to theinactive holding of the foreign central bank, or vice versa; their effects, though in the ‘right’ direction,would obviously be much weaker.15

But even if it were true that changes in foreign holdings made no difference to credit conditions in thereserve centre, the contrast that is supposed to exist between the effects of gold movements on the one handand movements in foreign liquid reserves on the other rests largely on a preconceived opinion as to theeffects that gold movements ought to have, rather than on an empirical study of the effects they did have inthe period under review. There was a strong tendency for countries to insulate their internal money supplyfrom the influences of the balance of payments. More often than not, gold movements were offset or‘neutralized’, not always deliberately by any means, but often ‘automatically’; not always owing to theaction but frequently owing to the inaction of the central bank. The adjustment of the balance ofinternational payments does not depend as closely as has sometimes been thought on changes in thedomestic money supply in the various countries.

Thirdly, the common theoretical charge of a lack of reciprocal adjustment in the gold exchange systemloses its force under certain conditions which happened to prevail in reality. Suppose two countries (A andB) hold their reserves in a third (C). If one of them (A) loses reserves to the other (B), there is obviously fullreciprocal action between the two, just as if gold had moved; provided of course that the ‘rules of the game’are observed and that the movement is not neutralized. It may be objected that this is rather a specialassumption to make. But it is an assumption that corresponds to the facts. The gold exchange system wasnot an agglomeration of countries indiscriminately holding each others’ currencies. There was a distincttendency for reserves to be held in a central nucleus (C), even though the nucleus consisted not of one butof two or three countries. In such a system, full reciprocal adjustment of credit and prices can take placebetween the member countries inter sc.16 It is true that between the member countries on the one hand (A, B)and the centre country on the other (C) the adjustment, on the present assumptions, will tend to be unilateralinstead of reciprocal. (If prices in the member countries are on the whole appreciably lower than in thecentre country, their aggregate foreign reserves and hence their domestic credit base will expand; and viceversa.) But this is as it should be. It is part of the centre country’s responsibility in such a system to keep itslevel of prices and employment reasonably stable; it is for the member countries to keep themselves attunedto that level.

Within the area over which the system operates, and so long as it operates, the centre country need haveno anxiety about its own international liquidity, since its own currency is accepted and used as a meansof international settlement by the countries adhering to the system. In return, it is of vital importance for thecentre country to keep up its demand for imports and/or its foreign lending, so as to maintain the liquidity ofthe member countries. The decline of economic activity in 1929–32 was far greater in the United States thanin the rest of the world, and was reflected in a sharp drop in United States imports which, together with the


cessation of long-term lending and the recall of short-term credits, severely depleted the dollar reserves ofmost of the member countries.

In fact, however, the nucleus of the gold exchange system consisted of more than one country; and thiswas a special source of weakness. With adequate co-operation between the centre countries, it need nothave been serious. Without such co-operation it proved pernicious. In addition to the variability of themember countries’ exchange holdings as compared with gold, reserve funds were liable to move erraticallyfrom one centre to another, from London, for instance, to New York and vice versa, and each centre wassubject to risks and disturbances on that account. Gold had to be immobilized in these centres as coveragainst such transfers, and the desired economy of gold was to that extent nullified. When one of themember countries—France—desired to become a gold centre herself, these difficulties were furtherincreased. By and large, however, it was not until the suspension of the gold parity by the United Kingdomthat transfers of funds in search of security set in on a really devastating scale from one centre to another.17

How the breakdown of exchange stability and the cumulative liquidation of the gold exchange system actedand reacted on one another was indicated earlier in this chapter. Here it may be remarked that, given theexistence of several centre countries, it is exchange stability as between these centres that is of primaryimportance for the working of an exchange standard. When the comparative prospects of the various centrecurrencies become subject to discussion, sudden and disruptive shifts of reserve funds will be difficult toavoid. The problem does not arise in a system with only one centre, and an occasional alteration ofexchange rates between a member country and that single centre is unlikely to have serious consequencesfor the functioning of the system as a whole.

Appendix I Gold supply[in millions of US dollars of old gold parity ($20.67 per fine ounce)]

World outputa Easterndishoardingb

Central reservesc Change in centralreserves

Industrialconsumptiond

1914 448 5,345

1915 472 6,241 + 896

1916 455 6,630 + 389

1917 421 7,147 + 517

1918 384 6,8 16e

1919 358 6,805e

1920 333 7,255e

1921 330 8,044e

1922 320 8,417 156

1923 369 8,651 + 234 153

1924 385 8,976 + 325 146

1925 384 8,997 + 21 152

1926 395 9,233 + 236 142

1927 394 9,593 + 360 124

1928 390 10,057 + 464 119

1929 397 10,336 + 279

1930 432 10,944 + 608 102

1931 461 146 11,323 + 379 63


World outputa Easterndishoardingb

Central reservesc Change in centralreserves

Industrialconsumptiond

1932 498 224 11,933 + 610 36

1933 525 152 11,976 + 43 40

1934 570 152 13,050 + 1,074 39

1935 625 102 12,990 –60 49

1936 688 77 13,600 + 710 46

1937 730 40 14,400 + 800 47

1938 780 34 15,300 + 900 30

Notes:a Source: US Bureau of Mint. The estimates include the USSR.b Source: Bank of International Settlements.c Source: Federal Reserve Bulletin, September 1940. Reported gold reserves of central banks and governments

(excluding in particular certain stabilization funds, such as the British Exchange Equalization Account).d Source: US Bureau of Mint. Gross estimates, including not only the new gold but also the scrap and coin used in the

arts. The net consumption, excluding scrap and coin, has been estimated at an annual average of $100 millionduring the five years 1925–9 (see Interim Report of the Gold Delegation, League of Nations 1930, p. 90).

e Excluding Russia’s reserve, not reported in those years: reported at $666 million in 1917 and at $3 million in 1922.

Appendix II Foreign exchange and gold reserves of central banks[in US $ (thousands)]

Key: A: Foreign exchange. B: gold. C: total. D: foreign exchange as % of total

End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

Austria A 67 79 96 105 102 93 113 20 8

B 2 2 7 12 24 24 30 27 21

C 69 81 103 117 126 117 143 47 29

D 97 98 93 90 81 79 79 43 28

Belgium A 6 6 62 73 79 85 135 — —

B 52 53 86 100 126 163 191 354 361

C 58 59 148 173 205 248 326 354 361

D 10 10 42 42 39 34 41

Bulgaria A 7 4 5 8 20 8 6 2 1

B 8 8 8 9 10 10 11 11 1

C 15 12 13 17 30 18 17 13 12

D 47 33 38 47 67 44 35 15 8

Czecho A 20 36 62 72 74 68 72 31 30

slovakia B 27 27 27 30 34 37 46 49 51

C 47 63 89 102 108 105 118 80 81

D 43 57 70 71 69 65 61 39 37

Danzig A 6 5 8 7 8 8 9 5 3

B — — — — — — — 4 4

C 6 5 8 7 8 8 9 9 7

D 100 100 100 100 100 100 100 56 43


Denmark A 13 24 7 26 31 24 27 4 –1

B 56 56 56 49 46 46 46 39 36

C 69 80 63 75 77 70 73 43 35

D 19 30 11 35 40 34 37 9 —

Estoniaa A 1 1 1 3 7 6 4 4 1

B 3 3 3 3 2 2 2 2 4

C 4 4 4 6 9 8 6 6 5

D 25 25 25 50 78 75 67 67 20

Finland A 14 31 27 32 19 17 24 16 19

B 8 8 8 8 8 8 8 8 8

C 22 39 35 40 27 25 32 24 27

D 64 79 77 80 70 68 75 67 70

End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

France A 14 13 116b 850b 1287 1021 1027 842 176

B 710 711 711 954 1254 1633 2099 2699 3257

C 724 724 827b 1804b 2541 2654 3126 3541 3433

D 2 2 14 47 57 38 33 24 5

Germany A 310 243 230 113 126 194 182 –29 –29

B 181 288 436 444 650 544 528 234 192

C 491 531 666 557 776 738 710 205 163

D 63 46 35 20 16 26 26 — —

Greece A 36 30 32 34 48 32 32 14 13

B 12 13 14 14 7 8 7 11 8

C 48 43 46 48 55 40 39 25 21

D 75 70 70 71 87 80 82 56 62

Hungary A 35 61 43 36 17 14 12 4 3

B 7 10 30 34 35 29 29 18 17

C 42 71 73 70 52 43 41 22 20

D 83 86 59 57 33 33 29 18 15

Italy A 21 64 156 398 317 271 228 114 69

B 221 221 223 239 266 273 279 296 307

C 242 285 379 637 583 544 507 410 376

D 9 22 41 62 54 50 45 28 18

Latvia A 9 6 6 10 15 11 8 3 2

B 5 5 5 5 5 5 5 6 7

C 14 11 11 15 20 16 13 9 9

D 64 55 55 67 75 69 62 33 22

Lithuania A 6 3 4 5 5 8 9 3 2

B 3 3 3 3 3 3 4 5 5

C 9 6 7 8 8 11 13 8 7


D 67 50 57 63 63 73 69 38 29

Netherlands A 45 99 75 67 88 88 99 35 29

B 203 178 166 161 175 180 171 357 415

C 248 277 241 228 263 268 270 392 444

D 18 36 31 29 33 33 37 9 7

Norway A 14 19 24 16 11 18 19 6 8

B 39 39 39 39 39 39 39 42 39

C 58 58 63 55 50 57 58 48 47

D 26 33 38 29 22 32 33 13 17

End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

Poland A 49 1 24 100 80 59 46 24 15

B 20 26 27 58 70 79 63 67 56

C 69 27 51 158 150 138 109 91 71

D 71 4 47 63 53 43 42 26 21

Portugal A 18 18 10 8 16 17 9 21 20

B 11 11 9 9 9 9 9 13 24

C 69 27 51 158 150 138 109 91 71

D 71 4 47 63 53 43 42 26 27

Romania A — — — — — 40 10 2 3

B 48 48 49 51 49 55 56 58 57

C 48 48 49 51 49 95 66 60 60

D — — — — — 42 15 3 5

Spain A 7 6 7 7 18 19 16 54 55

B 489 490 493 502 494 495 471 434 436

C 496 496 500 509 512 514 487 488 491

D 1 1 1 7 4 4 3 11 11

Sweden A 36 54 56 70 58 71 105 13 57

B 64 62 60 62 63 66 65 55 55

C 100 116 116 132 121 137 170 68 112

D 36 47 48 53 48 52 62 19 57

Switzerland A 37 43 43 38 49 68 85 20 17

B 98 90 91 100 103 115 138 453 477

C 135 133 134 138 152 183 223 473 494

D 27 32 32 28 32 37 38 4 3

Yugoslavia A 74 71 65 67 45 52 23 8 4

B 14 15 17 17 18 18 19 31 31

C 88 86 82 84 63 70 42 39 35

D 84 83 79 80 71 74 55 21 11

Total (24 countries) A 845 917 1159 2145 2520 2292 2300 1216 505

B 2281 2367 2568 2903 3490 3841 4316 5273 5879


C 3126 3284 3727 5048 6010 6133 6616 6489 6384

D 27 28 31 42 42 37 35 19 8

End of: 1924 1925 1926 1927 1928 1929 1930 1931 1932

Total excl. France (23) countries A 831 904 1043 1295 1233 1271 1273 374 329

B 1571 1656 1857 1949 2236 2208 2217 2574 2622

C 2402 2560 2900 3244 3469 3479 3480 2948 2951

D 35 35 36 40 36 37 37 13 11a 1924–7: Bank of Estonia and Treasuryb Estimated (largely held under ‘sundry assets’).

Notes

1 Each of these banks was authorized by its statutes to hold balances with the two others, and to count thesebalances as part of the reserve on which the issue of notes was based. The convention of 1885 provided inter aliathat ‘each of these three banks shall open a current account with each of the others; on this account they mayissue cheques payable at sight, even if this involves an overdraft; all sums may be paid in to their respectivecredits. No interest will be charged on credit or debit balances…. Debit balances must be paid up at the request ofthe creditor bank.’ See Janssen (1922), p. 6.

2 Austria-Hungary, Belgium, Bulgaria, Denmark, France, Germany, Greece, Italy, Netherlands, Norway, Romania,Russia, Spain, Sweden, Switzerland.

3 Cf. Brown (1940), p. 748.4 In Japan the foreign-exchange transactions of the authorities were conducted, and the foreign balances held, not

by the bank of issue but by the Yokohama Specie Bank.5 Even after the de facto stabilization, from December 1926 to June 1928, the chance that the franc might in the

end be legally stabilized at a level higher than the current market value was by no means negligible. There weremany politicians and financiers strongly advocating such a course. There were in consequence recurrent waves ofbull speculation in the franc, the last of which—in May 1928—was particularly violent.

6 Cf. Bank of France (1932), p. 9: the Bank ‘deemed it preferable not to increase through its actions the monetarydifficulties of other countries. By conserving the major part of the foreign assets—notably of the pounds sterling—that it held, the Bank has contributed in a very large measure, during the last three years, to maintaining thestability of the British currency.’

7 Bank of France (1933), p. 5.8 Sweden, as a creditor country, is included in this group, even though she was not one of the gold-absorbing

countries; her gold reserve actually declined in 1931. However, the inclusion or exclusion of Sweden does notappreciably affect the totals given in the text.

9 Cf. Brown (1940), pp. 849 ff.10 Apart from numerous French writers, mention may be made of Mlynarski (1929) who sharply attacked the gold

exchange standard on this ground. 11 Cf. Brown (1940), p. 789.12 Cf. Royal Institute for International Affairs (1931), p. 91.13 By Sir Otto Niemeyer (1931), p. 90.14 League of Nations (1932), p. 55.15 Apart from the question of the effects of changes in reserve balances on credit conditions in the reserve centre, it

would have been highly desirable in any case to hold such balances at the central bank. Thus the PreparatoryCommission of Experts (1933, p. 16) recommended that in order to secure ‘a system more centralized andsubject to more effective control…foreign exchange holdings in Central Banks should be invested with or


through the Central Bank of the currency concerned or with the Bank of International Settlements. This is all themore important, because it is, in our opinion, imperative that Central Banks should have a complete knowledgeof all the operations of other Central Banks on their markets.’

16 Cf. Keynes (1930), vol. I, p. 353.17 Cf. Hawtrey (1939), p. 267.

References

Bank of France (1932), Annual Report for 1931.——(1933), Annual Report for 1932.Brown, Jr, William Adams (1940), The International Gold Standard Reinterpreted, 1914–1934, New York, National

Bureau of Economic Reserach.Hawtrey, R.G. (1939), The Gold Standard in Theory and Practice, 1st edn, 1927, London, Longmans Green.Janssen, A.E. (1922), A Note on the Plan for an International Clearing House, Geneva, League of Nations, Provisional

Economic and Financial Committee.Keynes, J.M. (1930), A Treatise on Money, London, Macmillan.League of Nations (1932), Report of the Gold Delegation, Geneva, League of Nations.Mlynarski, F. (1929), Gold and Central Banks, New York, Macmillan.Niemeyer, Sir Otto (1931), ‘How to economize gold’, in Royal Institute for International Affairs, The International

Gold Problem, London, Oxford University Press, pp. 84–94.Preparatory Commission of Experts (1933), Draft Annotated Agenda for the Monetary and Economic Conference,

Geneva, League of Nations.Royal Institute for International Affairs (1931), The International Gold Problem, London, Oxford University Press.


Part IV

Bretton Woods and after

Introduction

The Bretton Woods System established after World War II has been described as a gold-exchange or gold-dollar standard. Milton Gilbert analyzes its operation from this perspective. Published in 1968, Gilbert’sarticle warns of instabilities inherent to the system and predicts its collapse three years before the fact. Theaccompanying piece, by Barry Eichengreen, suggests that the success of Bretton Woods rested on a specialset of factors specific to the immediate post-World War II era and cautions that any attempt to reform theinternational monetary system in its image would prove a failure.

15The gold-dollar system: conditions of equilibrium and the price

of goldMilton Gilbert*

This essay was written in the summer of 1967; its purpose was to explain the nature of the internationalmonetary system and how it functioned up to mid-1967 as a background to the consideration of variouspossible improvements in the system. Hence, it deals with the system as it has been—not as it mightbecome. While I have made drafting changes and clarifications, I have deliberately not extended the paperto cover the events of the past year so as to avoid discussion of matters about which there are differences inofficial views. My objective is to analyse the system and not to enter into the political problems of its futureevolution.

More specifically, the essay aims to distinguish between difficulties arising from inadequate adjustmentpolicies of individual countries and difficulties arising from a disequilibrium1 in the system as a whole,which concerns the relationship between gold and the dollar. The analysis is focused on the persistentdeficit in the balance of payments of the United States and is designed to bring out its underlying andtransient causes.

The gold-dollar system

The present system is usually called the gold exchange standard. As it emerged from Bretton Woods and asit has functioned in the postwar period, however, it is more to the point to call it the gold-dollar system.

It may seem curious that economics textbooks do not provide an analytical model of the system. Indeed,there is not really a formal theory of the gold exchange standard, comparable to the theory of the goldstandard. This may be, partly, because the system has not been static but has been developing under thechanging conditions of the last two decades. However, it is also because the system does not lend itselfeasily to presentation by a simplified model, as does the gold standard, since central banks do not constitutea homogeneous universe and do not act according to a set pattern of economic considerations.

The absence of an accepted theory of the gold-dollar system has made for much confusion in publicdiscussion. Economists and officials do not start with a generally agreed conception of how the systemworks or ought to work, such as they have, say, in dealing with problems of demand management. They donot have a model for the equilibrium of the system or a common view on the respective roles of gold andthe dollar. In these circumstances, very diverse, and often contradictory, proposals have been offered tosolve the problems of the system and the deficit of the United States. In the main, these proposals are eitherunconvincing as prescriptions for establishing equilibrium in the framework of the present system, or they

* From Essays in International Finance, no. 70, Princeton, Princeton University Press, 1968, pp. 1–20, 46–7, abridged.

involve changes so fundamental as to constitute a new system. It is hardly surprising that the politicalauthorities have not been able to find their way out of the maze.

Basic principles and behaviour characteristics of the system

The system rests on a series of basic principles and behaviour characteristics which determine its mode ofoperation. These derive from law, from international agreements, from the policy aims of central banks andgovernments and, in some respects, from technical necessity. It is, of course, an evolving organism, whichwas different thirty years ago and which will, no doubt, be different thirty years hence. The concern here, torepeat, is with the system as it has existed during the past two decades.

Fixed exchange rates

1 Fundamental to the system is the aim of monetary authorities to adhere to fixed rates of exchange.Maintenance of the rate has a high priority with all countries and other objectives are often sacrificed to it.It is apparent that fixed rates have overwhelming support from the business and financial community also.

2 To say that we have a fixed-rate system raises the question of what the rates are fixed to. Under the IMFArticles, a country may declare its par value in terms either of gold or of the dollar of the gold weight andfineness in effect on 1 July, 1944. There is only a minor technical difference between these two standards.

However, the operative standard for most countries is the dollar as such, and central banks in practiceintervene in the market when necessary by buying or selling dollars against their own currencies to keep thedollar exchange rate within agreed limits. The cross rates with other currencies are kept in line by marketarbitrage. There are exceptions, of course, such as the countries of the sterling area, which peg theircurrencies to sterling and rely on the Bank of England to maintain the fixed rate between sterling and thedollar. But, generally, central banks operate directly on the market for dollars vis-à-vis their own currency,and it is the market rate on the dollar that is significant for their international competitive position—irrespective of the legal gold content of the dollar.

3 The exception in the system is the dollar itself, which both in law and in fact is fixed in terms of gold—at $35 an ounce. The United States is not obliged to intervene in the exchange market; it has only to beprepared to buy and sell gold at $35 and can leave it to the intervention of other central banks to maintainfixed rates to the dollar. The United States has intervened in the market at times in recent years, both spotand forward, but the purpose was to avoid losses of gold from temporary movements of funds rather than tokeep rates in line.

Reserves

1 To maintain fixed rates the monetary authorities must hold reserves so that they are in a position to ironout fluctuations in supply and demand in the foreign-exchange market. Reserves consist of liquidinternational assets, readily available for intervening in the market, and are almost entirely confined to goldand to foreign-exchange assets in dollars and sterling.

However, while sterling is important to the international economy as a trading currency, it is active as areserve currency only in settlements between the United Kingdom and sterling-area countries. Hence, it is aregional reserve currency and quite different from the dollar, which is the reserve currency of the system.To simplify matters I will discuss only dollars, thus treating the sterling area (and, likewise, any othermonetary area) as a unit which holds reserves in gold and dollars.


It will be seen, therefore, that the system is a gold-dollar system for two reasons: first, currencies arefixed either to the dollar or to gold and, secondly, the reserves of the system are gold and dollars.

2 Each central bank is free to determine the composition of its reserves as between gold and dollars. Itspolicy in this respect is in its own hands, because it can sell or buy gold against dollars at the US Treasury.If there is any constraint on such exchanges, even psychological or political, the convertibility of the dollarat its fixed relationship to gold comes into question. In practice, there is a wide range among the countries inthe ratio between holdings of gold and dollars, indicating that central banks give different weight to thebenefits to themselves of the two categories of assets.

3 Dollars are held almost entirely in money-market instruments and time deposits, as these are liquid assetsthat earn interest. Gold, on the other hand, produces no revenue. It is important to realize that, if centralbanks could not earn interest on dollars, their reserves would be almost entirely in gold. Some centralbankers have stressed that they are not primarily concerned with interest earnings in determining thecomposition of their reserves. This may be true of marginal changes in reserves; but the point is that, if theUnited States did not permit central banks to invest dollars at interest, they would never have acquired thedollars in the first place; they would have acquired gold.

Hence, the first requirement for a currency to become a reserve currency is that there must be an openmoney market in which foreign central banks can freely invest in short-term paper. In addition, the moneymarket must be capable of absorbing large central-bank transactions, and the convertibility of the currencyat a fixed rate must be rather secure. It is because New York and London are the only two open moneymarkets of any size that the dollar and sterling are the only two significant reserve currencies. And it isbecause exchange rates between sterling and other currencies have not been secure that the dollar, supportedby large gold reserves, supplanted sterling as the reserve currency of the system.

Other currencies have not become reserve currencies either because the central bank discouragesplacements of funds at interest by foreign central banks or because their convertibility at a fixed rate doesnot seem reasonably assured over the longer run. Continental European countries have not wanted to becomereserve centres; they are reluctant to have their markets and reserves disturbed by large-scale operations offoreign central banks and some, also, see no point in their country bearing the interest burden attached tohaving their currency held as reserves.

It has been said that dollars are kept in reserves primarily as a matter of convenience, since dollars can beused directly in the exchange market whereas gold must first be converted into dollars for the purpose ofmarket intervention. However, dollars held in money-market paper or on fixed-term deposit must equally beconverted into cash to be available for market intervention, and there is no great difficulty in convertinggold into cash.

While a variety of developments went to make the dollar the reserve currency of the system, the UnitedStates took no initiative in the matter; its action was only permissive.

4 With respect to its reserves, also, the United States is an exception. While other countries are free tohold their reserves in any combination of gold and dollars they wish, including 100 per cent in dollars, theUnited States must hold its reserves essentially in gold. This is because there is no other currency besidesthe dollar that can be used for general intervention in the exchange market; hence, any foreign currencies heldby the United States cannot be used for general support of the dollar in the way that other countries use thedollar as a general support for their currencies. The United States can generally use foreign-exchangeholdings only for bilateral settlements. To underline the importance of this point, France could hold all itsreserves in dollars if it were so minded, but the United States could only hold a quite small fraction of itsreserves in French francs.

THE GOLD-DOLLAR SYSTEM 211

The foreign-exchange assets that have appeared in the reserve statistics of the United States in recentyears were always acquired for specific purposes. For example, there may be a temporary holding of D-Marks which were acquired in the market in anticipation of repaying D-Mark Roosa bonds to theBundesbank. Or, there may be small holdings of Swiss francs to be fed into the market when the dollar isunder pressure so that the Swiss National Bank will not have to acquire the excess of dollars which it mightthen want to convert into gold.

The only currency that the United States has held in large amounts has been sterling. These holdingsarose mainly because American assistance to the Bank of England was given in the form of swaps of dollarsagainst sterling—rather than as simple advances. The sterling, of course, could not be used at the same timeby the United States to meet its own deficit or to avoid gold losses and, therefore, was not ‘reserves’ in theordinary sense of immediately marketable assets. To count such sterling assets in reserves is about asappropriate as it would be, say, for a business firm to include its accounts receivable in its cash.

5 Since the United States is the only country obliged to hold its reserves in gold, the function of gold as a‘discipline’ against excessive money creation is primarily applicable to the United States. Other countriesare subject to balance-of-payemnts discipline, but the discipline lies in the loss of any reserves—whetherdollars or gold. Even so, a loss of gold makes a much greater impression on public opinion in manycountries than a loss of foreign-exchange reserves or foreign borrowing by the central bank. If othercountries entirely stopped acquiring gold, the discipline of gold on the United States would become rathertheoretical. This is particularly so because increases in its liabilities to foreign official institutions seem tohave exerted little discipline on the United States.

6 Why do central banks, apart from the United States, hold non-interestbearing gold at all and whatdetermines the proportion between their holdings of gold and dollars? Several considerations are involved,to which the various countries attach different importance.

(a) Gold is unique in that the asset of the holding country is not a liability of another country. For dollarassets, on the other hand, there must be a liability in the United States—either money-market paperor bank deposits. Hence, the disposition of gold is entirely in the hands of the holding country,while the use of dollars may require the acquiescence of the United States. However remote it mayseem, countries take account of the possibility that exchange balances may be blocked in times ofpolitical trouble, such as war, and they hold some gold over which they are the sole masters. This isthe ‘war-chest’ motive and it is often said that holding gold is an aspect of sovereignty. Gold-buyingby China in 1965 and 1966 probably reflected this motive, and one sees its influence frequently intimes of political stress. The war-chest motive is sometimes disparaged by writers who look upongold as anachronistic, but it is evident that every major country gives it some weight in its reservepolicy. Apart from the fact that gold reserves have been drawn upon in past wars, their existencesupports a country’s credit standing in such troubled times. It is a fact also that foreign-exchangebalances have been blocked for political reasons.

(b) The exchange risk to a central bank on its gold reserves is limited to a possible fall in the price ofgold or to an appreciation of its own currency vis-à-vis gold. On dollar reserves there is theadditional risk of its own currency appreciating vis-à-vis the dollar as a result of a rise in the dollarprice of gold while its own price of gold remains unchanged. When sterling depreciated in 1931,some central banks had large balance-sheet losses on their sterling holdings by the change inexchange relationships. The small group of central banks that hold almost all of their reserves ingold are concerned primarily to avoid such risk to their balance-sheet position. They do not considerit a primary function of the central bank to earn interest on its reserves; they took their increases of


reserves in gold even when there seemed to be no possible threat to the convertibility of the dollar,because they knew, if the risk should arise, there might be practical limitations to conversion ofdollars into gold. Having their reserves in gold, they believe, gives them greater independence ofaction in the event of a future monetary crisis; that is, if the United States should reduce the goldcontent of the dollar, they would be free to fix their exchange rate with the dollar onprospective balance-of-payments considerations alone—without the complication of a possible lossin the domestic-currency value of their reserves.

In holding gold, of course, a central bank must have confidence in the intrinsic value of gold interms of its own currency. But, then, they all do—and that is putting it mildly.

(c) Another consideration in reserve policy is the possibility of a universal rise in the price of gold. Acentral bank which held only dollars in that event would not take any loss on its reserve holdings,but it would not have the benefit of the marked-up value of reserves that gold-holding countrieswould have. I know of only one central bank which calculated years ago that the book profit from apossible rise in the gold price was too uncertain to set against realizable interest earnings; it,therefore, made the decision to hold a minimum in gold, to take its interest earnings on dollarholdings, and to stick to this policy even if a rise in the gold price became more of a possibility. Butmany other central banks have not been that unwavering. Besides, some feel themselves open tointernal political criticism when their ratio of gold reserves gets much out of line with that ofneighbouring countries.

(d) A large number of central banks have a very low gold ratio. These are mainly capital-importingcountries. They have not given much weight to the exchange risk because they expect to maintain afixed rate with the reserve currency under almost any circumstances. Furthermore, they look upontheir reserves partly as overborrowing by their country from abroad, and they see their interestearnings as a partial offset to the interest payments which have to be made abroad. Some centralbanks, also, have little scope for earnings on domestic operations and it is only interest receipts ontheir reserves that allow them some independence from the government.

(e) A few central banks have always considered it an obligation to take at least part of any increase ofreserves in gold so that the ‘discipline of gold’ should be a reality for the United States. Likewise,some have continued to buy some gold from the United States to maintain the principle of theconvertibility of the dollar, even after it became impractical to exercise the right of convertibility tothe limit. They have felt it necessary to resist full acceptance of a dollar standard and have beenstrengthened in this view by what they consider to be an inadequate priority which the United Statesgives to correction of its payments deficit. As one official put it, if we accept a full dollar standard,it would be like having a country with two central banks sometimes working at cross purposes.

(f) In several countries the law requires the central bank to maintain reserves in gold as backing for thedomestic currency. This legal provision is a leftover from the days when gold coins were in activecirculation and has little relation to present-day conditions. It is the only motive for central banksholding reserves in gold that is entirely traditional.

Given this variety of motives, it is apparent that the comparative benefits of holding gold relative to dollarscannot be calculated. In other words, central banks cannot know what reserve policy will make their countrybetter off—and, perhaps, they cannot even define precisely what being ‘better off is. What many do, therefore,is work to some rule of thumb. A few years ago, for example, there were several central banks that aimed tohave about a 50–50 ratio between gold and dollars, whereas others held mostly gold and still others mostlydollars. Reserve ratios generally are not set once for all, however, but are subject to change according to


circumstances—particularly to changes in the degree of certainty regarding the gold convertibility of thedollar at its fixed price.

7 Besides reserves, IMF facilities are available in the system to assist countries that encounter balance-of-payments difficulties. The amount any country may draw is originally fixed by its quota, which broadlyreflects its size and economic strength. In establishing its quota, each country as a rule pays 25 per cent tothe IMF in gold and 75 per cent in its own currency and agrees that its currency may be drawn upon in caseof need to finance other countries’ drawings. The Fund may also finance drawings partly by selling gold inorder to acquire the needed currencies.

The right of a country to draw on its gold subscription is practically automatic; so also is its right to drawon any credit balance it may have built up by having had its own currency drawn upon. These two amountshave come to be called a ‘reserve position in the Fund’. If a country draws on its quota above its reserveposition in the Fund, it is taking credit and its right to this credit is conditional upon the IMF judging that itspolicies are likely to correct its external deficit.

From the standpoint of meeting a deficit, therefore, a Fund reserve position is equivalent to a country’sown reserves. But it differs from reserves in three respects:

(i) A drawing on the 25 per cent gold tranche of its quota carries repayment obligations. (ii) Public confidence in a currency depends more on the size of reserves than on the country’s reserve

position in the Fund.(iii) A Fund reserve position—except for credits under General Arrangements to Borrow (GAB)—does

not yield interest like dollar reserves, nor does it have the charactertistics that induce countries to holdnon-interest-bearing gold; thus, when surplus countries supply resources for drawings by deficitcountries, they do so as an act of co-operation rather than an act of investment for its own sake.

There has been one instance of a country making a deposit with the IMF; the consideration involved wasthat the deposit was covered by a full gold guarantee, in contrast to the normal gold-value guaranteeincorporated in the IMF Articles. Hence, it is not an arrangement with large possibilities of expansion,because the IMF could not assume the risk on the price of gold.

The credit tranches of a country’s quota are not comparable to reserves; they are conditional facilities andany credit obtained carries definite repayment obligations and interest charges. Some high officials have hadthe hope that drawings on the IMF would become fairly routine central-bank operations—like the use ofbank credit by a business firm to supplement its working capital. Thus far, however, this idea has not beenrealized; drawings on the IMF have been indicative of a strained or crisis situation in which the IMF iscalled upon as a rescue organization. In fact, countries have at times emphasized drawing on the IMF so asto gain public support for necessary corrective policy actions.

Since the IMF was established in 1944, there have been two general increases in countries’ quotas, by 50per cent in 1959 and by 25 per cent in 1966, as well as special increases for particular countries. Also, theGeneral Arrangements to Borrow was agreed to by the Group of Ten industrial countries in 1962, wherebythey could lend additional resources to the Fund to help meet large drawings by members of the Group. Theneed for this arrangement arose because the Fund’s stock of convertible currencies could be inadequate tomeet large drawings under conditions when a balance-of-payments deficit of the United States limited theFund’s use of its dollar holdings or when the United States itself wanted to make a large drawing on theFund. In either of these cases, therefore, the Fund could have a problem of liquidity; in fact, in the BrettonWoods’ arrangements it was probably not contemplated that the Fund’s dollar holdings might not be freelyusable because of a large deterioration in the reserve position of the United States.


It should be noted that transactions under the GAB are covered by a full gold guarantee. 8 In addition to the IMF, short-term central-bank credit facilities have been arranged among a number of

countries. These may be used on an ad hoc basis and are designed essentially to help meet reversiblemovements of private funds and to relieve the pressure on the reserves temporarily while the character ofthe demand for foreign exchange is being appraised. Such assistance is provided on the credit standing ofthe central bank, in which the size of the reserves is an essential consideration. They are not conditional inthe sense of IMF credit facilities, since the borrowing central bank cannot make commitments about theadjustment policies of its government.

9 Besides reserves and official credit facilities, extensive use is made of foreign commercial-bank creditand other private liquid funds to meet strains on the exchange market. Central banks may do this on theirown account, or they may arrange matters so that it is done by their own commercial banks. The scope forsuch operations has been much enlarged by the development of the Euro-dollar market and the market hasin recent years been drawn upon by several countries for quite large amounts. Private credit facilities arecertainly a flexible supplement to official resources and are likely to be of growing importance. It would begoing too far, however, to consider them a substitute for monetary reserves, especially since a country withinadequate reserves is not likely to have a high credit rating with private banks.

The adjustment process: countries in deficit

1 A country with a balance-of-payments deficit can for a time hold its exchange rate by drawing on reservesand available borrowing facilities. As these are limited, however, and as drawing on them too much maymake matters worse by leading to a flight from the currency, the authorities must sooner or later take actionto get out of deficit. When this adjustment is not brought about, and the exchange parity depreciates,economic and financial policy are considered by the general public to have failed. Whether maintaining therate is a reasonable objective in given circumstances, however, depends on whether the authorities can takesufficient policy action to eliminate the deficit.

2 The policy actions available to eliminate a deficit and some limitations on them are, briefly, as follows:

(a) Fiscal and monetary restraint on total domestic demand so as to limit imports and, possibly,encourage exports. After a bout of inflationary pressure, the curtailment of demand to restoreexternal balance often results in a short period of domestic recession. But the Bretton Woodsexperts did not expect countries to subject themselves to prolonged stagnation in order to maintainthe fixed exchange rate of the currency. While this does happen, of course, it is the country itselfwhich sets the priorities among its objectives.

(b) Monetary restraint to raise interest rates relative to rates abroad so as to improve the net externalbalance on short- and long-term capital account. This instrument, too, has limitations because it willcause domestic recession and stagnation if pushed too far—though the limits can be widened bycompensatory fiscal action.

(c) Long-term borrowing abroad by the government or by other authorities of the public sector. This iseasier for some countries than for others.

(d) Reduction of government expenditures abroad, in cases where such expenditures are relatively large.The limitations on this technique are political but, none the less, real.

(e) Direct controls on imports and invisibles. Import controls are subject to severe limitations byinternational agreement and their use by an industrial country implies a rather desperate situation.Indeed, as the major purpose of a fixed rate is to encourage liberal trading practices, there is not


much sense in maintaining the rate by restrictions on trade. The limitations on controls overinvisibles are less severe, but such controls are used much more for protection than for balance-of-payments purposes.

(f) Direct controls on capital exports. Such controls are not limited by international agreement and maybe used freely, not only by deficit countries, but even by countries in surplus, without any sanctionsbeing available to other countries. In fact, a country with a balance-of-payments deficit attributableto a deficit on capital account is not normally supposed to be eligible for IMF assistance. Manycountries maintain controls on capital movements, either for balance-of-payments or for domesticreasons; others have little need of them because the combination of monetary and fiscal policiesthey follow leads to market conditions which limit capital outflows anyhow.

Hence, in adhering to the principle of no direct capital controls until a few years ago, the United States wasalmost alone among the convertible-currency countries. Because of its high per capita income and the hugevolume of savings generated by its economy, the United States would have been the dominant capitalmarket in the world in any case. But this position was reinforced by the controls and policies maintained inother countries.

In the last few years the United States has imposed direct capital controls to limit its gold losses. Mostother industrial countries find this course perfectly natural and desirable. Indeed, some seem to believe thatthe deficit of the United States, apart from the effects of the Vietnam War, could be cured by stringentenough capital controls. This view, to my mind, does not take sufficient account of all the links there arebetween the capital and current accounts, or of the shifts that take place between the various categories ofcapital outflow and inflow when controls are applied.

3 The force of these instruments can be very substantial when they are used vigorously and there havebeen many instances in the postwar period of countries emerging successfully from a period of deficit bymeans of them—without undergoing deep recession or prolonged stagnation. However, cases can and doarise in which they are unable to restore external balance—usually because domestic inflation has broughtinternal prices and costs too far out of line. Hence, the aim of maintaining a fixed rate cannot be consideredabsolute.

A deficit position which requires a change in the exchange parity to bring about correction is called a‘fundamental disequilibrium’ and it is provided in the IMF Articles that a country in such a situation maychange its rate without sanctions. There is no legal definition of fundamental disequilibrium, but in practicecountries do not apply to the IMF for a change in rate before the situation is perfectly obvious; they havealways obtained approval. A country that resorts to extensive exchange restrictions in such a situationinstead of adjusting the rate is not supposed to be eligible for IMF assistance, though, if the truth be told,some countries have got away with murder. It is far from pleasant for the IMF to insist that a countrydevalue as a condition to drawing IMF credit.

While the evidence of fundamental disequilibrium in some cases is unmistakable, the distinction betweentransitory and basic imbalance is difficult to make in others. There is no computer program by which theprecise equilibrium rate of exchange can be determined, and, even if there were, no country would changeits rate to correct a small disadvantage in the structure of exchange rates. There are several reasons whichjustify this attitude. First, to depreciate the rate by, say, 3 per cent or 5 per cent would be likely to do moreharm than good, because of the distrust in the currency that it would engender. Secondly, the policyinstruments available for maintaining external balance are sufficient to prevent prolonged reserve losses insuch cases, without undue sacrifice of other objectives, for example, by rather small changes in the capitalaccount. And thirdly, there is an adjustment process constantly at work which tends to correct small


imbalances, particularly when it is helped along by appropriate demand policy and when it is not negated bycontinual wage inflation. This adjustment process takes place both within the given country and in the worldecon omy on the outside; its reality is evident from the fact that reasonably well-managed countries are ableto maintain fixed rates over long time spans.

Thus, the existence of fundamental disequilibrium is a matter of degree and to specify it in any given caseis a matter of judgement. Such a judgement is particularly hazardous when external imbalance isaccompanied by excess domestic demand and when there is likely to be some flight of capital contributingto the imbalance. For example, there were observers who considered that the lira had become overvalued in1963, but this was proven to be a misjudgement as soon as the domestic inflation was brought undercontrol. On the other hand, all competent analysts considered the French franc to be in fundamentaldisequilibrium in 1957—and they were right.

4 For the generality of countries in deficit, the availability of a change in rate, which improves thecompetitive position of exports relative to imports, means that a balance in external payments can always berestored. In fact, it always is restored. When a country delays action until it runs out of reserves and runsout of credit, it must in the end devalue. It may hide this fact from itself by tying its economy into knotswith extreme exchange restrictions and multiple exchange rates and by turning its eyes from the blackmarket which always springs into life in such circumstances. But, then, the currency has effectively beendevalued de facto, if not de jure.

Consequently, there is nothing wrong with the adjustment process when it is viewed as including achange in exchange rate as the ultimate policy instrument. We have seen it work perfectly adequately incase after case. Where the external deficit was due merely to excess domestic demand, as in the Netherlandsin 1956–7, Italy in 1963–4, or Germany in 1965–6, the deficit disappeared when effective monetary andfiscal measures were taken to restrict internal demand. And where such action would not do, because therewas a fundamental disequilibrium, as in France in 1956–7, or Spain in 1957–9, the deficit disappeared whenappropriate devaluation was undertaken in combination with restricting excess demand, which was thecause of the imbalance in the first place.

It would be far more satisfying, of course, if the monetary and economic behaviour of countries werealways such that they avoided falling into fundamental disequilibrium. But if they do not, it is no reflectionon the system. And if they choose to suffer the distortions and stagnation of an overvalued currency insteadof adjusting to an equilibrium exchange rate, it is on their own responsibility as sovereign nations.

5 Here again, however, the United States is a significant exception because as a practical matter it cannotact directly on exchange rates. This follows to some degree from the fact that the dollar is fixed to gold,rather than to any particular currency. But it is a consequence even more of the weight of the United Statesin the world economy and the significance of the dollar in the international monetary system. Suppose theUnited States decided that its balance-of-payments deficit could not be corrected by acceptable adjustmentpolicies, and that it had gone the limit in using its gold reserves and taking IMF and central-bank assistance.It could then either raise the dollar price at which the Treasury buys and sells gold or simply suspend goldsales by the Treasury without fixing a new price for the time being. Whether any changes would then bemade in exchange rates vis-à-vis the dollar would depend upon the reaction of other countries. In the firstcase they could maintain their fixed parties to the dollar, with the result that the price of gold would behigher in all currencies. In the second case, also, they could intervene in the exchange market to maintainthe peg to the dollar and let the price of gold in their own currencies be free to move with market forces.

I leave until later the question of what they might do under various conditions and here wish only tostress two points: the first is that the process available to the United States for removing a persistent deficitis different than for other countries; the second is that the equilibrium of the dollar involves the equilibrium


of the whole system in a way that is different than for other currencies and is necessarily related to the priceof gold. The difference between the dollar and other currencies in this regard may seem to be a difference ofdegree, but it is so large as to constitute a difference in kind.

6 This position of the dollar is what lies behind the official insistence on improvement of the adjustmentprocess. There is not great concern about the adjustment process in general, because other countries cannotavoid adjustment. And even if they have to adjust by means of a change in exchange rate, it is largely alocal affair which does not involve the system as a whole. The key target of the demand for betteradjustment is the persistent deficit of the United States because it is likely to involve the stability of thesystem as a whole. There has been a strong feeling that somehow its deficit has reflected misbehaviour onthe part of the United States—even when the United States was clearly not having excess demand, when themargin of unemployed resources was unnecessarily large, and when it could not be convincingly shown bywhat combination of policy measures the United States could meet the demand to eliminate its deficit.However, there have been very few in official circles bold enough to draw the apparently logical conclusionthat the dollar was in fundamental disequilibrium; very few have felt that their exchange rate vis-à-vis thedollar ought to have been revalued. For its part, the United States took refuge in the idea that the troublewas with lack of adjustment by the surplus countries—and the charges back and forth left matters more orless at a standstill.

For the past several years, also, criticism of the adjustment process has been directed at the UnitedKingdom. However, the United Kingdom was having substantial excess demand, domestic inflation, andoverfull employment. And at the same time it was asking for very large assistance from abroad to financeits external deficit. Hence, the grounds for complaint were quite different than in the case of the UnitedStates before the start of the Vietnam inflation.

7 A final point with regard to deficits. Given the nature of the policy instruments available for correctinga significant deficit position, it will be apparent that the process of adjustment is necessarily a relativelyshort-term affair. When it does not take place fairly quickly, it simply means that the authorities have nottaken the appropriate measures—either deflation and capital controls, if the imbalance is not fundamental, ordevaluation, if there is fundamental disequilibrium. And when the exchange rate is significantly overvalued,there is no way to adjust other than by changing the exchange rate.

Governments are often reluctant to accept this proposition because of the stigma usually connected with achange in the exchange rate; so they think up all sorts of pseudo-measures for the long-run correction of thedeficit. In recent times, however, there is not a single successful case of long-run adjustment of a sizeablebalance-of-payments deficit—apart from the special case of reconstruction of war damage to the productivepotential of the economy. And even those cases did not take very long. In former times, when stagnation ofthe economy led to declining wages and prices, such adjustments often occurred. In our day of downwardrigidity of wages and prices and of the high priority given to full employment, however, such an adjustmentcan take place only through wages in the deficit country rising less than in the outside world—and themargin of correction that has been possible by this process has proven relatively small.

The United States, in particular, has had a long-term programme to restore balance for seven years andyet the goal is as elusive as ever. Failure to face up to this reflects political attitudes—not economicanalysis.

The adjustment process: countries in surplus

1 It is often said, from the standpoint of the system as a whole, that both surplus and deficit countries mustshare the responsibility for achieving balance in international payments. However, the primary


responsibility, and the active role in the adjustment process, falls in fact on deficit countries because it is theirexchange rates that are in jeopardy. When a country is in surplus, the central bank can feel free toconcentrate on domestic objectives of full employment and growth. But when the country is in deficit, itcannot. Thus, there is a natural bias toward being in surplus, since the surest way to avoid any risks to theexchange rate is to stay on the right side of the line. When a country is in moderate surplus, therefore, it willnot take deliberate action to reduce the surplus and, even when the surplus is fairly large, deliberatecorrective action is rather limited. The cooperative actions taken by surplus countries have been confinedlargely to facilitating the financing problem—such as prepayment of long-term debt, provision of specialfacilities to the banks to acquire foreign-exchange assets, and accepting special exchange-guaranteed assetsinstead of gold. Several countries also have used special techniques to limit the inflow of funds from abroad.

2 Besides the general aim of protecting the exchange rate, several more specific factors militate againstan active adjustment policy by surplus countries:

(a) The basic objectives of economic policy are usually stated as full employment, stable prices, andexternal balance. Now, reasonable judgements can be made about full employment and pricestability, but external balance is too hazy a concept to serve as a guide to operating policy. It needs alot of interpretation. It is a normative, longer-run idea, whereas policy is made for a shorter run inwhich true ‘equilibrium’ can hardly ever be said to exist. The authorities are acutely aware that theexternal position may change rapidly and are inclined to expect that a surplus this year maydisappear or be smaller next year. The relative cyclical position of the country may favour asizeable surplus at the moment, but it is likely that both the cyclical position and the surplus will bedifferent a year from now. The surplus may reflect other temporary influences which can alwayschange and probably will change, such as an inflow of liquid funds or unusual imports of long-termcapital. Appraisal of the underlying situation is often difficult because of changes in the foreignposition of the banking system. Exports may be quite favourable, but there is the possibility thatwage increases may erode the country’s competitive position. And so on.

When the European countries were receiving Marshall aid, the external position was alwaysappraised with the aid apart—quite sensibly, too, because if the aid had not been looked upon astemporary how would the countries have ever arrived at a position in which aid was no longerneeded? Similarly, American military expenditures in Europe tended to be regarded as a temporaryfactor in the balance of payments—quite different from a country’s own exports as a means ofearning a foreign-exchange surplus. In sum, to the extent that surplus countries aim at adjustment,they tend to discount what they regard as temporary elements of the surplus.

(b) A somewhat similar influence arises from countries’ objectives with regard to the structure of thebalance of payments. This affects their willingness to pursue adjustment policies, the policyinstruments they use, and the view they take of the position of the currency in the structure ofexchange rates. The point may be illustrated by contrasting Canada and the Netherlands.

Both countries devote about the same percentage of GNP to gross investment and there is not asignificant difference in their per capita real output. Yet, no doubt largely for historical reasons,their balance-of-payments aims are quite different.

Canada looks upon itself as a developing country with a capital requirement that cannot be fullymet by internal savings. It expects the current account of its balance of payments to be in deficit andto be compensated by net capital imports. Thus, while Canada aims to avoid a deficit in its overallpayments position, its use of fiscal and monetary policy is conditioned by the view that net capitalimports are normal. When Canada fixed a lower rate of exchange in 1962, the obvious purpose was


to reduce the deficit on current account. But obviously, also, the rate chosen, of 0.93 Canadian tothe US dollar, was not intended to secure current-account balance but allowed for continuing capitalimports.

The Netherlands, on the other hand, considers itself among the relatively well-off countries of theworld and thinks that it should have a surplus on current account to allow for foreign aid and netcapital exports—particularly to finance the foreign investment of Dutch international enterprises.Hence, it uses fiscal and monetary policy to secure a volume of domestic savings, including savingsin the public sector, that exceeds domestic investment, and it generally manages to do so. If this aimis frustrated, say by an inflow of foreign direct investment or foreign purchases of Dutch securities,the authorities are unwilling to allow the current account to adjust to this situation. The result is anincrease in official reserves or improvement in the external position of the banking system. Whenthe Netherlands followed Germany in revaluation of the currency in 1961, it was certainly not withthe idea that the current-account surplus would be wiped out.

This view of balance-of-payments aims is characteristic of many continental countries; that is,they aim at being in surplus on current account and some even extend the objective to the tradeaccount.

They consider it inappropriate to allow adjustment of their current-account surplus to compensatefor capital exports by the United States. They could, and do to some extent, arrange their ‘policymix’ or relax controls so as to obtain some offset by exports of domestic capital funds. But thisoften poses difficulties in the domestic sphere—political and other. They think it is mainly up to theUnited States to take action against its ‘excessive’ capital exports.

Since the imposition of direct controls on capital exports by the United States, there has been asizeable volume of international issues on the Euro-bond market which has tended to increaseEurope’s net capital exports. In addition, some countries have arranged capital exports by public-sector institutions, including the central bank, to help absorb a current-account surplus. But it wouldtake a great change of attitude indeed for the European countries to allow these techniques toexpand enough to deteriorate their reserve positions, and this would not be a sustainable situation inany case.

(c) Finally, the tolerance for surpluses is influenced by the fact that international transactions have astrong upward trend. It is recognized, therefore, that reserves must increase also if they are toremain adequate to defend fixed rates against balance-of-payments fluctuations. The growth inreserves and transactions may not need to be at the same rate, since policy action may become moreeffective in narrowing balance-of-payments fluctuations. Also, a country’s reserves for the momentmay be more than adequate, so that their growth could be allowed to slide for several years withoutcausing trouble. But after a time their upward trend would have to be resumed. This is true forindividual countries and for all countries taken together. Furthermore, as an increase in reservescomes about by having an external surplus, one must say that the norm is for countries to be insurplus. Hence, equilibrium for an individual country, and for all countries together, is not simply asituation of balance between external receipts and payments; there must on average be surpluses. Acountry that fails to achieve a reasonable growth in reserves is bound to meet with difficulties onexternal account and declining confidence in its currency. Sterling is a striking example of acurrency falling into this kind of situation. Since the early 1950s the United Kingdom authoritieshave wished to secure an upward trend in their reserves to provide better support for the sterling-area system. But the aim was constantly being crowded out by other objectives—economic andpolitical—with the result that sterling has been subjected to repeated exchange difficulties.


The mercantile flavour of official attitudes toward the balance of payments comes from recognition that asurplus approximates equilibrium in a growing world better than does a constant level of reserves. And, ofcourse, economists have stressed the need for global reserves to rise. The Netherlands central bank formerlyset a quantitive target for the growth of its foreign reserves and tried to arrange domestic liquidity creationso that the target would be met. Other central banks are influenced by the same idea, if in a less precisefashion.

3 While a reasonable, or even moderately large, surplus may be in the vicinity of equilibrium, cases ariseof persistent extreme surplus. This is a situation in which, after total demand has been pushed to a full-employment level, the surplus does not fall to reasonable proportions. In the conventions of international co-operation, there is no obligation on surplus countries to pursue inflationary demand policies in order tobring down the surplus.

The only provision in the Fund Articles for such cases is the ‘scarce currency’ clause. It has never beencalled upon and, indeed, was expected only to be applicable to the United States. A country declared to havea ‘scarce currency’ would not have to do anything itself, but other countries would have a right todiscriminate against it in their trade and payments regulations.

To be parallel with devaluation in cases of fundamental disequilibrium, countries in exreme surplusought to revalue their currencies. But the high priority on fixed rates holds for currency appreciation as wellas devaluation, and revaluation is a rare occurrence. The only recent case was the revaluation in 1961 of theGerman mark and the Dutch florin by 5 per cent.

However, revaluation must be recognized as the ultimate policy weapon available to countries that wantto stop the inflationary consequences of an extreme surplus. The surplus is an inflationary force because, inpegging the rate, the central bank has to buy the excess of dollars offered in the market against domesticcurrency and there are practical limits to the extent to which the authorities can neutralize this increase ofdomestic currency by other policy actions. The revaluation of the Deutsche Mark was undertaken preciselyon these grounds.

Two other cases show that the authorities do have the power to act against an extreme surplus when theyfeel that it constitutes an intolerable danger to internal monetary stability. Canada adopted a floating rate in1950 to combat a huge inflow of investment funds from the United States; the Canadian dollar appreciatedbetween 5 and 8 per cent and a balance was achieved in the market with a less inflationary inflow of fundsfrom abroad. Switzerland, for some years after the war, also allowed its rate to float for most transactionsother than trade and a part of tourism, when faced with an unmanageable inflow of funds from abroad. TheSwiss franc on the free market appreciated as much as 30 per cent to the dollar, against the par of 4.37. Itshould be noted that both Canada and Switzerland could have converted their dollar inflows into gold inthose days without any reproach from Washington. But it was the large surplus they did not want even in gold.

Although rare, these instances of revaluation are significant from two standpoints. First, they show thatthe United States cannot foist any amount of dollars on the rest of the world; beyond a certain point othercountries would sever their fixed ties to the dollar. The fact that they have not done so indicates that theyhave not considered their currencies to be undervalued. Secondly, the system provides countries in extremesurplus with a remedy, if they care to use it. If they do not, it is on their own responsibility. The savinggrace from the standpoint of other countries is that an extreme surplus tends to be eroded by internalinflation.

4 By comparison with any previous time, there has been a high degree of international monetary co-operation in these postwar years. It must be recognized, however, that every country gives priority to itsown basic interests and that the demands on co-operation cannot violate those interests. As the world is


made up of sovereign states, ultimate responsibility both for the convertibility of each currency and for thecontrol of inflation within each country is a national responsibility.[…]

Deterioriation of the system

Since the eruption of the market price of gold in 1960, there has been a steady deterioration in the operationof the system and a change in its character. I summarize the main aspects of this deterioration from thereview of developments already recounted.

1 The standing of the dollar as the reserve currency of the system has become compromised as there isless readiness to hold dollars freely. To minimize conversion of dollars to gold under these circumstances,the United States has resorted to giving guarantees on various of its external liabilities.

2 The United States has used moral suasion to prevent dollars being converted to gold. It is no secret thatsuch conversions are considered to be at least uncooperative, and in some cases unfriendly. Some foreigncentral banks have refrained from demanding gold for dollars so as not to rock the boat. Hence, centralbanks no longer have full freedom over the composition of their reserves; nor is it quite right to say that thedollar is still freely convertible de facto.

3 After the rise of the market price of gold in 1960, the principal central banks formed the gold pool inorder to keep control over the price. At the start, the assumption of the pool was that there would normallybe an excess of market supply over demand—which may include buying by central banks that are notmembers of the pool. By 1967 the pool had to supply not only the deficit of gold for private demand but themarket demand of nonmember central banks as well. The residual supplier, of course, was the UnitedStates, since the other pool members could offset their gold losses to the pool by purchases from the UnitedStates.

4 The threat overhanging gold has restricted its use in official settlements; except in desperatecircumstances or for political ends, central banks try to meet temporary difficulties by other means.

5 While gold losses act as a discipline on the United States, they have become an uncertain guide forjudging its balance-of-payments performance. At one moment the authorities expressed a firm intention tobalance the external accounts. When they realized that this unilateral undertaking was impossible, they saidthat the surplus countries must carry a fair share of the burden of adjustment—without specifying what afair share for the United States would be. The latest posture seems to be a resigned attitude towards thebalance-of-payments deficit, with its persistence and size being attributed to the war in Vietnam. One canonly conclude that the authorities of the United States have not formulated a set of standards for judgingwhether its responsibilities for the reserve currency of the system are being fulfilled. In fact, of course, withthe shortage of new gold anything like as severe as it is at present, it cannot be done.

6 With the growing tightness of the system, it has become a matter of high priority to prevent anyexcitement on the exchange markets and to resort to extreme means to gain market confidence. One aspectof this is a fear of changes in exchange rates and a belief that almost any change in rates constitutes a threatto the stability of the system. This is in the face of the necessity for rates to be much more finely adjusted inconditions of a gold shortage than would be needed with an adequate flow of gold.

7 Owing to limitations on the growth of reserves through gold and dollars, the system no longer has abuilt-in mechanism for the increase in reserves. As a consequence, the growth of reserves has depended toa large extent on special credits negotiated to finance deficits. Such arrangements are often influenced bypolitical considerations, to the general detriment of strictly monetary and financial standards in the system.And, as the repayment of such credits would require a substantial contraction of global reserves, it is noteasy to visualize their orderly liquidation.


The changes in the system that have occurred since 1960 are often presented as an evolution andstrengthening of the system. While there have been innovations of permanent value, the essence of thematter has been a series of shoring-up operations to accommodate to a basic disequilibrium of the system.Far from the system being strengthened, it has been disintegrating. This can hardly be considered anevolution of the gold-dollar system, since it consists of replacing both gold and dollars with quite differentinstruments for the growth of reserves. The past six years have been transitional, and it is evident that thepattern of gold and credit financing followed in those years cannot be repeated in the next six years. If waysare found in the years ahead to suppress the official demand for gold, it will mean that a basically newsystem has come into being.

Note

1 My use of the terms ‘equilibrium’ and ‘disequilibrium’ to characterize concrete situations in the real world seemsto unbalance the editorial equanimity of Fritz Machlup, who has long and resolutely maintained that these termsshould be used only with reference to theoretical models with all variables fully specified. For want of moresuitable terms, I shall continue to use the proscribed ones in their real-world meanings.


16The Bretton Woods system: paradise lost?

Barry Eichengreen*

The Bretton Woods System frequently appears in the scholarly literature as a model for internationalmonetary reform. This paper briefly considers its operation. It argues that the system’s principalachievement, the maintenance of stable exchange rates, was the product not of the agreement finalized atthe Bretton Woods Conference alone but of two exceptional features of the postwar world. One was thelimited international mobility of capital. Capital controls provided policymakers room for maneuver; theysoftened the tradeoff between domestic objectives and defense of the exchange-rate peg. The other wassingular scope for growth resulting from postwar reconstruction and catch-up. In these circumstances,countries felt little need to engage in discretionary monetary and fiscal policies that might have underminedthe currency peg.

Introduction

A reader turning for the first time to the literature on Bretton Woods might be forgiven for thinking that hehad stumbled upon a forgotten sequel to Paradise Lost. Paradise, in the form of pegged but adjustableexchange rates, prevailed from the 1950s until 1971. Its pleasures included price stability, full employment,and effortless balance of payments adjustment. Paradise was lost in 1971-3, the system having beendestroyed by reckless policies, principally in the United States. The world was banished to a purgatory offluctuating exchange rates, rapid inflation, and high unemployment.

Or so the myth would have it. In fact, conditions were never so heavenly under Bretton Woods. And itsprincipal achievement, the maintenance of stable exchange rates, was a product not of the agreementfinalized at the Bretton Woods Conference alone but of two exceptional features of the postwar world.

One was the limited international mobility of capital. Governments applied capital controls during WorldWar II and retained them subsequently. Convertibility for current account transactions was only resumed inEurope on 31 December 1958. The restoration of convertibility for capital-account transactions had to waituntil years later.

The effectiveness of controls was buttressed by restrictions on international banking legislated inresponse to the Great Depression and by the fact that international bond markets had not yet recovered fromthe sovereign defaults of the 1930s. In this environment, controls could work. Together with quiescentmarkets, they limited international financial flows and provided policymakers room for maneuver. Theysoftened the tradeoff between domestic objectives and defense of the exchange-rate peg. Though never

* In Thierry Walrafen (ed.), Bretton Woods, Mélanges pour un Cinquantenaire, Association d’économie financière,1994, pp. 263–76.

impermeable and progressively less effective as time passed, they reduced the cost of defending a currencypeg and provided breathing space for governments to consult prior to devaluations.

The other special feature of the postwar world was singular scope for growth. The United States had lostnearly a decade of growth as a result of the Great Depression. Two decades of depression and war had leftEurope far behind. Now this lost time could be made up. With its economy growing at cent per annum inthe 1950s, Europe was content with undervalued exchange rates and modest real wage increases. While USgrowth was not as rapid, it was still impressive by historical standards and worked to soften policy tradeoffs.In these circumstances, countries felt little need to engage in discretionary monetary and fiscal policies.Aside from France’s problems, which were largely a consequence of overseas military commitments,pegged exchange rates posed few policy dilemmas.

By the end of the 1950s, this spurt of ‘catch-up’ growth was spent. Budget constraints tightened, andpolicymakers were torn between the maintenance of external balance and financing objectives like Europe’swelfare state and America’s Vietnam War. Meanwhile, the revival of international financial markets wasunderway. The restoration of current-account convertibility opened a back door through which capital couldflow. International bond markets reawakened, the Eurodollar market was born, and banks began syndicatinginternational loans. America’s gold reserves were no longer ample relative to its foreign liabilities. Balanceof payments constraints began to bind.

Rising capital mobility and new questions about the depth of governments’ commitments to theirexchange-rate pegs strained the Bretton Woods System. When currencies were thought vulnerable, theywere attacked. No longer was there time for consultation. Officials could not afford to be thought to becontemplating a devaluation, or a crisis would ensue. The system grew rigid and brittle. Instead of a regimeof pegged but adjustable exchange rates, Bretton Woods degenerated into a system of ‘fixed’ but ultimatelyunsustainable rates. Its collapse in the 1970s was rendered all but inevitable.

This paper develops these themes and draws out their implications for international monetary reform atthe turn of the century. The principal implication is that changes in the political and economic environmentwill rule out pegged but adjustable exchange rates like those established at Bretton Woods. Technologicalprogress in financial markets, by complicating efforts to apply capital controls, coupled with the moresclerotic state of labor and commodity markets, which creates a more troubled environment for growth, hasremoved the exceptional features of the postwar world that allowed the Bretton Woods System’s peggedbut adjustable rates to work. A move back toward pegged exchange rates like that proposed by the IMF’sMichel Camdessus is simply not on. This leaves two choices: continued floating and monetary unification.While Europe has opted for the second alternative, the first will continue to govern exchange rates betweenthe currencies of other advanced countries for the foreseeable future.

The role of capital controls

No one—certainly not John Maynard Keynes nor Harry Dexter White—advocated the early restoration ofcapital-account convertibility at Bretton Woods. Interwar experience with ‘hot money’, or ‘destabilizingspeculation’ as it was also known, led to a permanent disenchantment with uncontrolled capital movements.The IMF Articles of Agreement authorized the retention of controls on capital-account transactions. ArticleXIV also provided for a postwar transitional period, presumed to last for three to five years, when controlsmight extend to current-account transactions. Starting in 1950, under the aegis of the European PaymentsUnion, the continent made slow progress toward current-account convertibility. But its restoration tooklonger than anticipated: Europe completed the process only on 31 December 1958. And, even after 1958,controls on capital-account transactions remained the rule rather than the exception.

THE BRETTON WOODS SYSTEM 225

Obstfeld (1993) has analyzed the impact of controls using monthly data on interest rates on German andUK medium-term government bonds, from 1950 and 1947 respectively, computing interest differentialsrelative to comparable US interest rates. German and US bond rates generally moved in step over the longrun. But in the intermediate term, divergences could be considerable. In 1956, for example, German rateswere more than double the analogous US rates (at 7 versus 3 per cent). This differential reflected the BankDeutscher Lander’s efforts to restrain inflation. These differentials again opened up prior to the 1961 and1969 deutsche mark revaluations. Similarly, UK rates were some two percentage points above US ratesthroughout the Bretton Woods period.

The question is whether these differentials reflected expectations of devaluation and revaluation orobstacles to capital flows. To focus on the latter, Obstfeld computes covered interest differentials (interestrate differentials adjusted for the forward discount).1 He finds covered differentials as large as twopercentage points for the UK and larger than one percentage point for Germany in the 1960s. Thesedifferentials, which cannot be attributed to expected exchange-rate changes, confirm that capital controlsmattered.

Marston (1993) compares covered interest differentials between Eurosterling (offshore) rates and British(onshore) rates. The advantage of this comparison, relative to Obstfeld’s, is that it eliminates country risk(the danger that one country is more likely than the other to default on its interest-bearing obligations); thedrawback is that the time series concerned are relatively short. Between April 1961, when Eurosterlinginterest rates were first reported by the Bank of England, and April 1971, the beginning of the end for theBretton Woods System, the differential averaged 0.78 per cent. The Eurosterling-sterling differentialremained at least that large in 1973–9, when controls were maintained but exchange rates were floated, beforedeclining to essentially zero in 1979–87. Marston concludes that controls ‘clearly…had a very substantialeffect on interest differentials’. Similarly, covered interest differentials between Eurodollars and US CDsfell by two thirds between the control and post-control periods. The contrast is less pronounced whencovered interest differentials are calculated for Euromark and German interest rates, presumably becauseGermany, unlike Britain, was seeking to control capital inflows rather than outflows, but it points in the samedirection. The effectiveness of controls can be overstated, as Truman (1994) warns, but Marston’s evidenceconfirms that they mattered.

Controls made it possible for national authorities to defend their pegged exchange rates againstspeculative attacks not prompted by significant divergences in economic policy. Firms and brokers stillcould find ways of spiriting domestic currency out of the country, through over- and under-invoicing andthe operation of leads and lags, but the need to circumvent controls meant that there was expense involved.There had to be a reasonable expectation that a devaluation would follow in finite time for this to beworthwhile. Minor policy divergences that led to the modest overvaluation of a currency might not providesufficient motivation. This in turn gave national authorities some leeway to utilize their monetaryindependence.

Even when serious imbalances developed, controls were useful for providing the breathing space neededto organize orderly realignments.2 It was possible for policymakers to consult and agree on the magnitudeof the parity change without being submerged by a tidal wave of capital outflows. When devaluation wasimminent, speculative pressure could be intense, but officials still had several weeks or months to ponder acourse of action.

Not until the 1980s did most of the industrial countries remove their capital controls. Well before,however, the writing was on the wall. The financial system developed a variety of new channels tocircumvent controls, raising the costs of implementation and enforcement. Already in the 1960s, USinterest-rate ceilings and controls on capital outflows stimulated the development of the Eurodollar market.


Once it became apparent that US banking business could move offshore, the days of interest-rate ceilingswere numbered. And as Eurodollars were joined by Eurosterling and Euromarks, the process generalized.By the 1960s, the international bond market had recovered, and commercial banks, which increasinglymaintained offices in several countries, discovered the business of international lending.

As capital controls grew porous, pegged-but-adjustable exchange rates grew rigid. The mere hint of adevaluation could unleash an avalanche of capital flows. There being no possibility that sterling would berevalued in 1967 or that the dollar would be revalued in 1971, currency speculators were presented with anincreasingly low cost one-way bet. Policymakers could not display the slightest inclination of contemplatingthe possibility of a devaluation, or they would be overwhelmed by capital flows; this was true most of all ofthe United States, which was obligated to maintain the convertibility of the dollar into gold and whoseforeign liabilities, as a result of the dollar’s key currency status, greatly exceeded her official reserves. Theforeign-exchange value of the dollar could still be changed through simultaneous realignments by othercountries, but this required international consensus and coordination more ambitious than was feasible toarrange. Instead of being vented by parity adjustments, therefore, competitive imbalances were allowed tomount.

It is sometimes argued that the mistake of the officials responsible for the operation of the Bretton WoodsSystem was that they failed to realign in a more timely fashion.3 This criticism overlooks the fact that thegrowing permeability of capital controls undermined their ability to realign. If they displayed a willingnessto do so once, their reputations for defending their exchange rates would be tarnished thereafter.4

Speculators offered a low cost one-way bet would have every incentive to bet against the currency, and theinterest-rate premia that holders of domestic assets would consequently demand would increase the drain onofficial resources and weaken the government’s ability to defend the exchange rate peg.5 As a result, theBretton Woods System grew rigid. Eventually it was toppled by its own weight.

Post-war reconstruction and Bretton Woods

The other condition that helped Bretton Woods to work was rapid economic growth attributable to postwarreconstruction and catch-up. The (unweighted average) rate of growth of output per annum in WesternEuropean countries reached 4.7 per cent in the 1950s and 5.5 per cent in the 1960s.6 In Japan, growthaccelerated even more rapidly relative to prewar levels. Cross-section regressions suggest that Europe enjoyedsome two additional percentage points of growth per annum during the Bretton Woods years as a result ofthe backlog of available technology and its productivity gap vis-à-vis the United States.7

Rapid growth moderated adjustment difficulties and softened distributional conflicts. It was easier to getlabor and capital to agree to shift resources out of declining sectors when other sectors were growingrapidly. It was easier to enforce wage moderation when trends in living standards were strongly upward. Itwas easier to limit rates of profit taxation when the tax base was growing.

Wage moderation was buttressed by other exogenous conditions. Labor militancy was muted bymemories of high unemployment in the 1930s. Labor-market tightness was relaxed by the elastic suppliesof underemployed labor that Europe enjoyed.8 Ample supplies of labor flooding into Germany from itsEast, into Holland from Indonesia, and into France from Algeria encouraged docility on the part of unions.Underemployed labor in the agricultural sector had similar effects in other countries.

Rapid postwar growth and its concommitants had several favorable implications for the operation of theBretton Woods System. Where wage moderation prevailed, there was little danger that excessive inflationdue to rising labor costs would create a competitiveness problem sufficient to call into question theexchange-rate peg.9 The stability of money wages enhanced the effectiveness of exchange-rate changes;


because money wages did not rise in step with the exchange rate, parity changes could have powerful, long-lived effects.10 Wage flexibility and rapid productivity growth meant that full employment was relativelyeasy to maintain.

An implication was that governments came under less pressure than subsequently to actively employmonetary and fiscal instruments. In Germany, where ordo-liberalism was the order of the day, monetarypolicy was used to restrain inflation but rarely directed toward output and employment targets. Bordo(1993) and Eichenbaum and Evans (1993) report various measures of the magnitude of monetary policyshocks for a cross section of countries during the Bretton Woods years, concluding that these were smallerthan before and after. In Eichengreen (1993) I use a different methodology to derive estimates of aggregatedemand disturbances, to which monetary policy is one important contributor, finding that these weresmaller between 1959–70 than in surrounding periods. The fact that monetary policymakers were not tornbetween the imperatives of internal and external balance to the extent they would become subsequentlymeant that monetary instruments could be directed primarily toward balance-of-payments targets.

Such arguments must come to grips with the objection that these circumstances were a consequence, nota cause, of Bretton Woods’ pegged but adjustable rates. The rapid growth of the period could have beenassisted by the nominal anchor provided by Bretton Woods, which minimized price uncertainty andencouraged investment. But the fact that the acceleration in growth varied across countries and that much ofit was associated with the shortfall in output relative to postwar levels (the reconstruction effect) and gapsvis-à-vis the United States (the catch-up effect) suggests that it also had independent roots. It is moreplausible to argue that the lesser tendency for price-level increases to be passed through into wage inflationwas due not merely to memories of unemployment in the 1930s and elastic supplies of underemployedlabor but to the Bretton Woods System itself. But the extent of wage moderation also varied acrosscountries in ways that were associated with distinctive domestic labor-market conditions rather than simplyreflecting the operation of an international monetary regime common to all of them.11

This process played itself out in the 1960s. As growth decelerated, distributional conflicts intensified.Elastic supplies of labor from Eastern Europe were no longer available following the construction of theBerlin Wall, and a decade of growth had completed the process of drawing unemployed labor out of theagricultural sector. Meanwhile, unemployment rates continued to fall. The hot summer of 1968, punctuatedby a Europe-wide strike wave, signalled that a decline in labor market flexibility was underway. Bayoumiand Eichengreen (1996) estimate the elasticity of aggregate supply curves (which are steeper the moreflexible are wages and prices) for the Bretton Woods years and the post-Bretton Woods float; they find anoticeable decline in flexibility around the time of the breakdown of Bretton Woods. This meant that shocksto the economy increasingly displaced output and employment.12 Accordingly, political pressuresintensified for policymakers to direct the instruments at their command toward achieving internal ratherthan external balance. This new environment was less conducive to the maintenance of a system of peggedexchange rates.

Implications for international monetary reform 13

Both of the trends described above aggravate the difficulties of operating systems of pegged but adjustableexchange rates. Recent events in Europe underscore the problem. The inertial character European labormarkets compounds the difficulty of adjusting to shocks in the absence of the exchange-rate instrument. Thelimited independence of central banks heightens their susceptibility to political pressures and casts doubtover their commitment to robust monetary rules.14 The growth of international financial transactions,reputed to exceed $1 trillion a day, limits the effectiveness of capital controls.


One response to this problem is to live with floating exchange rates. In this conclusion I ask whetherthere exist alternatives at the global level.

A single world currency

A first option is to make exchange rates truly inflexible and unadjustable—irrevocably fixed, as is truewithin the United States, Canada, and other federations—by creating a single world currency.15 Byeliminating the exchange rate, monetary unification eliminates exchange-rate fluctuations. This is the paththat the European Union has opted to follow. But a clear lesson of the Maastricht process is that politicalsolidarity and economic convergence are prerequisites for monetary unification. Europe has been followingthis path for nearly half a century, as anyone familiar with the history of the EEC can attest. A web ofoverlapping bargains makes it feasible to extend side payments to countries that are reluctant to participatein the European Monetary Union. Thus, Germany, which hesitates to give up its beloved deutschemark forthe uncertainties of an ‘esperanto money’, might allow itself to be dragged into monetary union in returnfor an expanded foreign policy role in the context of a EU foreign policy. Any reservations France mightevince are offset by the political advantages of the Common Agricultural Policy, whose viability isthreatened by exchange-rate fluctuations.16 Small countries like Belgium and the Netherlands derivedisproportionate benefit from the access to the larger French and German markets which they enjoycourtesy of the Single Market Program. And Spanish and Portuguese qualms are assuaged by the benefitsthey reap from the EU’s Structural Funds. The existence of a European Parliament with expanding powersprovides reassurance that there exists an entity that might eventually possess the political authority to holdthe European Central Bank accountable.

It is unrealistic to hope that the major industrial countries can achieve comparable political and socialsolidarity in our lifetimes. ‘Federalism’ may be a dirty word in large parts of Europe, but the progress theEU has made in establishing a web of interlocking agreements underscores the very considerable stridesthat other parts of the world will have to take to emulate its example. It is difficult to imagine that the G-7,much less the entire world, will succeed in doing so in our lifetimes.

Currency boards

A currency board is designed to minimize uncertainty about the authorities’ commitment to defending theirexchange-rate peg. The currency board statute prohibits the authorities from issuing money except whenthey acquire foreign exchange reserves adequate to convert it into foreign exchange at a fixed rate. Forevery dollar’s worth of domestic currency they issue, for example, they must possess a dollar’s worth ofreserves. Credibility will be buttressed, giving speculators no incentive to test the resolve of the monetaryauthorities.

The question is whether credibility will be complete. The best way of answering it is to consider theoperation of a specific currency board arrangement. A good example is that of Estonia. The Estoniancurrency board statute separates the Bank of Estonia into Issue and Banking Departments and requires theformer to peg the exchange rate against the deutsche mark and to issue currency only upon acquiringdeutsche mark reserves.17 But although the Bank of Estonia is independent of the government, nothingprevents the parliament from changing the relevant law. Though the central bank currently has no discretionover the level of the exchange-rate peg, there remains the possibility that the currency board law will oneday be changed. It could be revoked or modified by Parliament in response to changing economic orpolitical conditions. Lainela and Sutela (1993) argue that Estonian officials in fact understand their currency


board to be a transitional arrangement to be abandoned in the not-too-distant future. For speculators, solvingbackward, this raises questions about the credibility of the peg.

The implication is that a currency board statute provides less than complete insulation against speculativeattacks. Requiring an act of parliament before the exchange-rate peg can be abandoned would presumablycompel the Bank of Estonia, in the event of an attack, to maintain the interest-rate defense for longer,irrespective of the domestic consequences of high interest rates, than if it was authorized to unilaterally alterthe peg. But the political fallout from high interest rates would be deflected onto the parliament. Thoughsignificant political costs might be incurred when revising the statute prohibiting changes in the exchangerate, nothing else would insulate the parliament from pressures to do so. The knowledge that there maycome a point where it has the incentive to change the currency board statute could provide speculators theincentive to mount an attack.

Target zones

In Williamson’s original proposal for a target zone system for the industrial countries, participantspreannounce bands for their real effective exchange rates, specifying a central rate surrounded by a 10 percent margin on either side.18 Given relative national inflation rates, this implies at any point in time acentral parity for the nominal exchange rate and a corresponding band. Governments would manage theirnominal exchange rates using foreign exchange market intervention and monetary policy so as to keep themin the band. Periodic realignments, to be undertaken before the edges of the band are reached, would avertthe danger of speculative attacks. In this respect, the arrangement would resemble a system of crawling pegs(surrounded by bands), in which the rate of crawl is governed by relative national inflation rates. The systemwould feature ‘soft buffers’ allowing the rate to move outside the band under exceptional circumstances.

It is worth considering the contrasts between this proposal and the European Monetary System, since thelatter arrangement proved so problematic in the early 1990s. Like the Williamson proposal, the EMSspecifies central rates and bands for each participating currency vis-à-vis baskets of other Europeancurrencies. It allows intervention by governments and central banks to keep currencies within their bandsand mandates intervention when the edge of the band is reached. It allows for periodic realignments of thecentral rate. But the Williamson proposal differs from the EMS in the width of its bands (wider than the per cent bands of the pre-1993 EMS, narrower than the 15 per cent bands of the subsequent system). Itdiffers in requiring the bands to be shifted before their edges are reached if the weakness of an exchangerate reflects an underlying competitiveness problem. It differs in allowing commitments to intervene to besuspended when that weakness reflects speculation not prompted by underlying competitive difficulties.

These features are attractive in many ways. The provision requiring the bands to be shifted before theiredges are reached would prevent a build-up of competitiveness problems when the bottom of the band wasapproached by offering a one-way bet to speculators and prompting them to attack. The soft-bufferprovision, allowing the band to be disregarded in the event of an attack not grounded in fundamentals,would allow the authorities to let the rate depreciate rather than raising domestic interest rates, ensuring thesurvival of the system. Once it became clear to speculators that the authorities were not inclined to alter thepolicies governing the evolution of fundamentals in response to the attack, the exchange rate should recoverand move back into the band.

The question is whether such a system would differ significantly from floating. The advantage of targetzones is the ‘bias in the band’, the fact that a credible commitment to defense of a target zone reduces theamount of exchange-rate variability associated with given fundamentals, creating a ‘target zone


honeymoon’. Less monetary policy intervention will therefore be needed to stabilize the rate. Hence, thetradeoff between exchange-rate stability and domestic monetary policy autonomy is relaxed.

Will the Williamson proposal create a target zone honeymoon? If bands are shifted as soon as adifferential develops between domestic and foreign interest rates, there is no reason for the markets toanticipate that the band will be defended, and there will be no bias in the band. Indeed, this is precisely thecircumstance in which the target zone honeymoon may give way to Bertola and Caballero’s target zonedivorce: an acceleration in inflation which increases expectations of realignment can increase the exchange-rate volatility associated with given fundamentals within the band.19 A more complicated set of monetary-policy intervention rules might give rise to more complex dynamics, but the resulting exchange-ratebehavior would not be obviously superior to that which would result from the kind of managed floating thatwould exist in the absence of target zones.

If, on the other hand, policymakers resist pressures to shift the band, allowing its boundaries to bereached and then intervening to prevent the rate from moving further, then they expose themselves to thekind of crises that upset the narrow-band EMS in 1993. In order to produce the bias in the band, they willhave to raise interest rates to defend the band’s edges. In an environment of virtually unlimited marketliquidity and no capital controls, the requisite interest rate increases, as in Europe in the summer of1993, may prove infeasible. Defending the band may only produce crises and no target zone honeymoon.

A Tobin tax

A final option is to tax foreign currency transactions as a way of mimicking some of the effects of capitalcontrols.20 This would enhance policy-makers’ ability to defend themselves against speculative attacks andhelp to reconcile the desire for exchange-rate stability with national policy autonomy. Speculators would beless likely to launch attacks against pegged currencies if they had to pay a tax to get in and out; while thiswould not permit seriously overvalued exchange rates to be defended indefinitely, it would decrease thelikelihood of attacks not motivated by serious policy imbalances. Countries wishing a modicum of policyautonomy could exercise it without immediately exposing themselves to the danger of violent exchange-ratefluctuations.

To be effective this policy would have to be implemented globally. The tax would have to apply to alljurisdictions, and the rate would have to be equalized across markets. Were it imposed unilaterally by onecountry, that country’s foreign exchange market would simply move offshore. If the tax was only applied byFrance, for example, French banks could ship francs to their foreign branches, where they would be sold forforeign currency free of tax.

Thus, the policy would have to be universal. Its implementation and coordination would have to be theresponsibility of a multilateral agency like the Bank for International Settlements or the InternationalMonetary Fund, which would have to possess enforcement capabilities. The IMF or BIS might beauthorized to set the size of the tax within limits. That organization will have to possess sanctions that canbe levied on countries that fail to comply with the measure. This is not something that will occur overnight.An international consensus supporting an amendment to the IMF Articles of Agreement would be needed,for example, to empower that institution to oversee the global adoption of the tax. There is little sign thatthis is a realistic possibility.


Conclusion

None of the options enumerated above—monetary unification, currency boards, target zones, or a Tobin tax—is likely to provide a viable solution to the problem of international monetary reform in the short run. Butjust as the difficulty of advancing multilateral trade liberalization through the GATT has encouragedregional trade liberalization, the difficulty of reforming the international monetary system globally is likelyto encourage regional initiatives. These tendencies have long been evident in Europe, where nations tradeextensively with one another. Ever since the collapse of the Bretton Woods System, the members of what isnow the European Union have pursued a series of initiatives to stabilize intra-European exchange rates.Intra-European exchange rates can be stabilized once and for all by abolishing them; there is noinsurmountable obstacle to completing this process on the schedule set out in the Maastricht Treaty,although there are reasons to question whether the three-step blueprint set out there is optimally designed.21

Other countries will have no choice but to resign themselves to floating exchange rates. With the passageof time, an international consensus might be built to allow a foreign exchange transactions tax to beimplemented globally. The political and social solidarity developed in Europe over the course of nearly 50years might be replicated in other parts of the world, allowing more monetary unions to be established. Butcultivating it will take time. For the world as a whole, there is no practical short-run alternative to livingwith floating rates.

Notes

1 Aliber (1978) and Dooley and Isard (1980) undertake similar exercises, reaching generally sympatheticconclusions.

2 This is the effect emphasized in the theoretical writings of Wyplosz (1986). Brown (1987, Chapter 4) describesthe principal capital movements and speculative crises of the Bretton Woods period. One impression conveyedby his account is that the length of time between the markets growing unsettled and the forceable abandonmentof exchange-rate pegs was longer than in the ERM crises of 1992–3.

3 This is, of course, also the official explanation for the 1992 crisis in the European Monetary System.4 Obstfeld (1993) cites a number of contemporaries who articulated this view.5 This last point is an implication of models of self-fulfilling balance-of-payments crises. See Flood and Garber

(1984) and Obstfeld (1986, 1994).6 Figures in this paragraph, as well as some of the analysis, are from Eichengreen and Kenen (1994).7 See Dumke (1990) and Crafts (1992).8 On the elastic-labor-supply argument, see Kaldor (1966) and Kindleberger (1967). 9 Alogoskoufis and Smith (1991) show that policies affecting prices produced smaller increases in wages and larger

increases in output and employment during the Bretton Woods period than subsequently.10 Obstfeld (1993) emphasizes the persistence of the effects of exchange-rate changes during the Bretton Woods

period.11 Evidence to this effect is provided by Eichengreen and Vazquez (1994).12 Bayoumi and Eichengreen (1994) document that a principal difference between the Bretton Woods period and

that of the post-Bretton Woods float was the slower speed of adjustment to shocks as time passed and hence thepersistence of fluctuations in output and employment.

13 This section draws on Eichengreen (1995).14 Measures to buttress central bank independence, undertaken in conjunction with Stage II of the monetary

unification process set out in the Maastricht Treaty, have helped to ameliorate this problem in Europe, but theindependence of major European central banks, in the UK and France for example, remains incomplete.

15 For arguments to this effect, see Cooper (1990) and Bergsten (1993).


16 Giavazzi and Giovannini (1989) have emphasized the disruptive impact of currency fluctuations on the CAP andthe impetus this lent to efforts to limit intra-European exchange-rate movements.

17 Under the law passed by the Estonian Parliament in May 1992, the currency (the kroon) must be fully backed bygold and foreign exchange. The Bank of Estonia can alter the quantity of notes and coin in circulation only byacquiring additional gold and foreign reserves. The Bank stands ready to convert kroons into deutsche marks formost current-account transactions. The exchange rate is pegged to the deutsche mark at the rate of 1 DM=8 EEK,with allowable fluctuations of plus or minus 3 per cent. See Hanke, Jonung, and Schuler (1993).

18 Williamson (1985). The proposal is generalized in Williamson and Miller (1987). Later variants propose settingthe central rate bilaterally against a reference currency such as the US dollar.

19 Bertola and Caballero (1991). Here ‘an acceleration in inflation’ is used as shorthand for any development thatwould undermine the international competitiveness of a country. Soft buffers, which allow the edges of the bandto be breached even in the absence of an acceleration in inflation, may provide motivation for speculators tomount an attack even when there is no inflationary event to prompt them.

20 The original proposal is Tobin (1978). It is updated and discussed in Dornbusch (1990) and Eichengreen, Tobinand Wyplosz (1994).

21 A critique of the Maastricht transition strategy, with various proposals for revision, can be found in the report andworking papers of Association for the Monetary Union of Europe’s report for the European Parliament(Collignon et al. 1993).

References

Aliber, Robert Z. (1978), ‘The Integration of National Financial Markets: A Review of Theory and Findings’,Weltwirtschaftsliches Archiv, 114, 448–79.

Alogoskoufis, George and Ron Smith (1991), ‘The Phillips Curve, the Persistence of Inflation, and the Lucas Critique:Evidence from Exchange Rate Regimes’, American Economic Review, 81, 1254–75.

Bayoumi, Tamim and Barry Eichengreen (1994b), ‘Macroeconomic Adjustment Under Bretton Woods and the Post-Bretton Woods Float: An Impulse-Response Analysis’, Economic Journal, 104, 813–27.

——(1996), ‘The Stability of the Gold Standard and the Evolution of the International Monetary System’, in TamimBayoumi, Mark Taylor, and Barry Eichengreen (eds), Modern Perspectives on the Classical Gold Standardpp. 165–88.

Bergsten, C.Fred (1993), ‘The Rationale for a Rosy View: What a Global Economy Will Look Like’, The Economist(11 September), 57–9.

Bertola, G. and R.J.Caballero (1991), ‘Target Zones and Realignments’, CEPR Discussion Paper No. 398.Bordo, Michael (1993), ‘The Bretton Woods International Monetary System: An Historical Overview’, in Michael

D.Bordo and Barry Eichengreen (eds), A Retrospective on the Bretton Woods System, Chicago, University ofChicago Press, 3–108.

Brown, Brendan (1987), The Flight of International Capital: A Contemporary History, London, Croom Helm.Collingnon, Stefan, with Peter Bofinger, Christopher Johnson, and Bertrand de Maigret (1993), The EMS in Transition,

Paris, Association for the Monetary Union of Europe.Cooper, Richard (1990), ‘What Future for the International Monetary System?’, in Yoshi Suzuki, Junichi Miyake, and

Mitsuaki Okabe (eds), The Evolution of the International Monetary System: How Can Efficiency and Stability beAttained? Tokyo, University of Tokyo Press, 277–300.

Crafts, N.F.R. (1992), ‘Institutions and Economic Growth: Recent British Experience in an International Context’,Western European Politics, 15, 16–38.

Dornbusch, Rudiger (1990), ‘It’s Time for a Financial Transactions Tax’, International Economy (August/September),95–6.

Dooley, Michael and Peter Isard (1980), ‘Capital Controls, Political Risk, and Deviations from Interest-Rate Parity’,Journal of Political Economy, 88, 370–84.


Dumke, Rolf (1990), ‘Reassessing the Wirtschaftswunder: Reconstruction and Postwar Growth in West Germany in anInternational Context’, Oxford Bulletin of Economics and Statistics, 52, 451–91.

Eichenbaum, Martin and Charles Evans (1993), ‘Some Empirical Evidence on the Effects of Monetary Policy Shockson Exchange Rates’, NBER Working Paper No. 4271.

Eichengreen, Barry (1993), ‘Epilogue: Three Perspectives on the Bretton Woods System’, in Michael D.Bordo andBarry Eichengreen (eds), A Retrospective on the Bretton Woods System, Chicago, University of Chicago Press,pp. 621–57.

——(1994), International Monetary Arrangements for the 21st Century, Washington, DC, Brookings Institution.——(1995), ‘History and Reform of the International Monetary System’, Economia e Sociedede 4, 53–78.Eichengreen, Barry and Peter B.Kenen (1994), ‘Managing the World Economy Fifty Years After Bretton Woods: An

Overview’, in C.Fred Bergsten and Peter B. Kenen (eds), Managing the World Economy, Washington, DC,Institute for International Economics pp. 3–57.

Eichengreen Barry, James Tobin, and Charles Wyplosz (1994), ‘Two Cases for Sand in the Wheels of InternationalFinance’, Economic Journal 105, 162–72.

Eichengreen, Barry and Pablo Vazquez (1994), ‘Institutions and Postwar European Economic Growth: An EconometricInvestigation’, unpublished manuscript, University of California at Berkeley.

Flood, Robert and Peter Garber (1984), ‘Gold Monetization and Gold Discipline’, Journal of Political Economy, 92,90–107.

Giavazzi, Francesco and Albert Giovannini (1989), Limiting Exchange Rate Flexibility: The European Monetary System,Cambridge, Mass., MIT Press.

Hanke, S. H.L.Jonung and K.Schuler (1993), Monetary Reform for a Free Estonia, Stockholm, SNS Norlag.Kaldor, Nicholas (1966), Causes of the Slow Rate of Economic Growth of the United Kingdom, Cambridge, Cambridge

University Press.Kindleberger, Charles P. (1967), Europe’s Postwar Growth, New York: Oxford University Press.Lainela, Seija and Pekka Sutela (1993), ‘Escaping from the Ruble Zone: Estonia and Latvia Compared’, unpublished

manuscript, Bank of Finland.Marston, Richard (1993), ‘Interest Differentials under Bretton Woods and the Post-Bretton Woods Float: The Effects of

Capital Controls and Exchange Risk’, in Michael Bordo and Barry Eichengreen (eds), A Retrospective on theBretton Woods System, Chicago, University of Chicago Press, pp. 515–46.

Obstfeld, Maurice (1986), ‘Rational and Self-Fulfilling Balance-of-Payments Crises’, American Economic Review, 76,72–81.

——(1993), ‘The Adjustment Mechanism’, in Michael D. Bordo and Barry Eichengreen (eds), A Retrospective on theBretton Woods System, Chicago, University of Chicago Press, 201–68.

——(1994), The Logic of Currency Crises’, NBER Working Paper No. 4640.Tobin, James (1978), ‘A Proposal for International Monetary Reform’, Eastern Economic Journal, 4, 153–9.Truman, Edward (1994), ‘Review of “A Retrospective of the Bretton Woods System” ’, Journal of Economic

Literature, 32, 721–3.Williamson, John (1985), The Exchange Rate System, Policy Analyses in International Economics No. 5, Washington,

DC, Institute for International Economics, revised edn (first edn 1983).Williamson, John and Marcus Miller (1987), Targets and Indicators: A Blueprint for the International Coordination of

Economic Policy, Policy Analyses in International Economics No. 22, Washington, DC, Institute for InternationalEconomics.

Wyplosz, Charles (1986), ‘Capital Controls and Balance of Payments Crises’, Journal of International Money andFinance, 5, 167–80.


Further reading

Readers seeking to pursue the issues addressed in this volume must cast their net widely. No single workprovides comprehensive coverage of the history, theory, and doctrine of the gold standard, much less allthree. This guide provides some suggestions about where to begin the search for further information.

The best surveys of doctrine on the gold standard remain Viner (1937) and Fetter (1965). Readersdesiring a more modern perspective should consult Bordo (1984).

Models of the price specie-flow mechanism appear in the standard textbooks of international economics,but the literature contains few unified theoretical treatments of the roles of interest rates, capital flows, andchanges in output and spending in the gold-standard adjustment mechanism. Although all of these effectsappear in the early theoretical work of Meade (1951), it is difficult to pick them out. Mundell (1971) andDornbusch and Frenkel (1984) provide more modern but more stylized treatments. A model of thedetermination of the price level under the gold standard is developed by Barro (1979) and extended to thecase of gold as an exhaustible resource by Bordo and Ellson (1985).

Readers interested in the structure of nineteenth-century financial markets should start with Kindleberger(1993) and Neal (1990). For overviews of the structure of the international monetary system, Bloomfield’s(1959, 1963) classic surveys remain unsurpassed. Many of the issues raised by Bloomfield are analyzedstatistically by Morgenstern (1959) and contributors to Bordo and Schwartz (1984). McKinnon (1996) andEichengreen (1996) contrast international adjustment under the gold standard with adjustment undersubsequent international monetary regimes. Overviews of the operation of the gold standard at the peripheryare provided by Keynes (1913), Ford (1962), and de Cecco (1974).

Hawtrey (1935) is a lively introduction to the origins of the gold standard and its nineteenth-centuryspread. Redish (1990) focuses on the transition from bimetallism to gold. On central banking under the goldstandard, useful central bank histories include Clapham (1944) and Sayers (1976) on the Bank of Englandand Ramon (1929) on the Bank of France. Whale (1944), Horsefield (1944), and Sayers (1957) take up therole of the Bank Charter Act in establishing the Bank of England’s public role. Insight into the operatingprocedures of central banks can be gleaned from interviews with central bankers conducted by the USNational Monetary Commission (1910). Central bank operations have been extensively analyzed: ondiscount policy see Goodhart (1972), Dutton (1984), and Pippinger (1984); on the gold devices Sayers(1936); and on the practice of holding foreign exchange reserves Lindert (1969).

Histories of the international economy between the wars are Lewis (1948) and Kindleberger (1973). Thedefinitive account of the international monetary relations of the period is Brown (1940). Clarke (1967,1973) is concerned with the reconstruction of the gold standard after World War I. Moggridge (1969, 1972)and Moure (1989) study the British and French decisions to return to gold and their consequences. Keynes’sviews of the problems with the reconstructed system, as summarized in the Macmillan Report, areelaborated in Keynes (1930). Studies of the collapse of the interwar gold standard are many: see forexample Gregory (1935), Cassel (1936), and Fraser (1933). Eichengreen (1990), Wheelock (1991), andSchubert (1991) are recent contributions to this literature. Temin (1989), Bernanke and James (1991), andEichengreen (1992) emphasize the connections between the gold standard and the Great Depression.

The definitive history of the Bretton Woods System is James (1996). Gardner (1969), Horsefield (1969),and Ikenberry (1993) consider the origins of the system in more detail. Dam (1982) and Obstfeld (1993)

draw out the parallels between the Bretton Woods System and the gold standard. Triffin’s (1960) critique ofBretton Woods is trenchant for its analysis of the potential for instability and prescient for its prediction ofthe system’s demise. Kenen (1969), Solomon (1977), and Williamson (1977) describe proposals forreforming the relationship between gold and the dollar toward the end of the Bretton Woods period, whileGarber (1993) models the system’s collapse.

On proposals to restore a form of gold standard, the reader should survey the range of conflicting opinionin the report of the US Gold Commission (1982) and its appendices. The Minority Report, which makes thecase for gold, is conveniently available as Paul and Lehrman (1982). It is criticized by Cooper (1982) andSchwartz (1982).

References

Barro, Robert (1979), ‘Money and the Price Level Under the Gold Standard,’ Economic Journal 89, 13–33.Bernanke, Ben and Harold James (1991), ‘The Gold Standard, Deflation, and Financial Crisis in the Great Depression:

An International Comparison’, in R. Glenn Hubbard (ed.), Financial Markets and Financial Crises, Chicago,University of Chicago Press, pp. 11–32.

Bloomfield, Arthur (1959), Monetary Policy under the International Gold Standard, 1880–1914, New York, FederalReserve Bank of New York.

——(1963), ‘Short-Term Capital Movements under the Pre-1914 Gold Standard’, Princeton Studies in InternationalFinance, No. 11, International Finance Section, Department of Economics, Princeton University.

Bordo, Michael D. (1984), ‘The Gold Standard: The Traditional Approach’, in Michael D.Bordo and Anna J.Schwartz(eds), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press,pp. 23–120.

Bordo, Michael D. and Richard Ellson (1985), ‘A Model of the Classical Gold Standard with Depletion’, Journal ofMonetary Economics, 16, 109–20.

Bordo, Michael D. and Anna J.Schwartz (1984), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago,University of Chicago Press.

Brown, William Adams (1940), The International Gold Standard Reinterpreted, 1914–1934, New York, NationalBureau of Economic Research.

Cassel, Gustav (1936), The Downfall of the Gold Standard, Oxford, Clarendon Press.Clapham, John (1944), The Bank of England: A History, Cambridge, Cambridge University Press.Clarke, S.V.O. (1967), Central Bank Cooperation, 1924–1931, New York, Federal Reserve Bank of New York.——(1973), ‘The Reconstruction of the International Monetary System: The Attempts of 1922 and 1933’, Princeton

Studies in International Finance, No. 41, International Finance Section, Department of Economics, PrincetonUniversity.

Cooper, Richard (1982), The Gold Standard: Historical Facts and Future Prospects’, Brookings Papers on EconomicActivity, 1, 1–56.

Dam, Kenneth W. (1982), The Rules of the Game: Reform and Evolution in the International Monetary System,Chicago, University of Chicago Press.

De Cecco, Marcello (1974), Money and Empire: The International Gold Standard, Oxford, Blackwell.Dornbusch, Rudiger and Jacob Frenkel (1984), ‘The Gold Standard and the Bank of England in the Crisis of 1847’, in

Michael D.Bordo and Anna J.Schwartz (eds), A Retrospective on the Classical Gold Standard, 1821–1931,Chicago, University of Chicago Press, pp. 233–75.

Dutton, John (1984), ‘The Bank of England and the Rules of the Game under the International Gold Standard: NewEvidence’, in Michael D.Bordo and Anna J. Schwartz (eds), A Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press, pp. 173–202.

Eichengreen, Barry (1990), Elusive Stability: Essays in the History of International Finance, 1919–1939, Cambridge,Cambridge University Press.

236 FURTHER READING

——(1992), Golden Fetters: The Gold Standard and the Great Depression, 1919– 1939, New York, Oxford UniversityPress.

——(1996), Globalizing Capital: A History of the International Monetary System, Princeton, Princeton UniversityPress.

Fetter, Frank (1965), The Development of British Monetary Orthodoxy, 1717–1875, Cambridge, Mass., HarvardUniversity Press.

Ford, A.G. (1962), The Gold Standard, 1880–1913: Britain and Argentina, London, Oxford University Press.Fraser, Herbert (1933), Great Britain and the Gold Standard, London, Macmillan.Garber, Peter (1993), ‘The Collapse of the Bretton Woods Fixed Exchange Rate System’, in Michael D.Bordo and

Barry Eichengreen (eds), A Retrospective on the Bretton Woods System, Chicago, University of Chicago Press,pp. 461–94.

Gardner, Richard (1969), Sterling-Dollar Diplomacy, New York, McGraw-Hill, 2nd edition.Goodhart, Charles A.E. (1972), The Business of Banking, London, Weidenfeld and Nicolson.Gregory, T.E. (1935), The Gold Standard and its Future, New York, E.P.Dutton, 3rd edition.Hawtrey, Ralph (1935), The Gold Standard in Theory and Practice, London, Longmans Green, 5th edition.Horsefield, J.Keith (1944), ‘The Origins of the Bank Charter Act, 1844’, Economica 11, 180–9.——(1969), The International Monetary Fund, 1945–1965: Twenty Years of International Monetary Cooperation,

Washington, DC, International Monetary Fund.Ikenberry, G.John (1993), ‘The Political Origins of Bretton Woods’, in Michael D. Bordo and Barry Eichengreen (eds),

A Retrospective on the Bretton Woods System, Chicago, University of Chicago Press, pp. 155–200.James, Harold (1996), International Monetary Cooperation Since Bretton Woods, New York, Oxford University Press.Kenen, Peter (1969), ‘The International Position of the Dolar in a Changing World’, International Organization, 3,

705–18.Keynes, John Maynard (1913), Indian Currency and Finance, London, Macmillan, volume I of The Collected Writings

of John Maynard Keynes, Cambridge, Cambridge University Press for the Royal Economic Society.——(1930), A Treatise on Money, London, Macmillan, volume VI of The Collected Writings of John Maynard Keynes,

Cambridge, Cambridge University Press for the Royal Economic Society.Kindleberger, Charles P. (1973), The World in Depression, 1929–1939, Berkeley, University of California Press.——(1993), A Financial History of Western Europe, New York, Oxford University Press, 2nd edition.Lewis, W.Arthur (1948), Economic Survey 1919–1939, London, Allen & Unwin.Lindert, Peter (1969), ‘Key Currencies and Gold, 1900–1913’, Princeton Studies in International Finance, No. 24,

International Finance Section, Department of Economics, Princeton University.McKinnon, Ronald (1996), The Rules of the Game: International Money and Exchange Rates, Cambridge, Mass., MIT

Press.Meade, James (1951), The Theory of International Economic Policy, London, Oxford University Press.Moggridge, Donald E. (1969), The Return to Gold, 1925, Cambridge, Cambridge University Press.——(1972), British Monetary Policy, 1924–1931: The Norman Conquest of $4.86, New York, Cambridge University

Press.Morgenstern, Oskar (1959), International Financial Transactions and Business Cycles, Princeton, Princeton University

Press.Moure, Kenneth (1989), As Good As Gold: French Monetary Management, 1925– 1936, Cambridge, Cambridge

University Press.Mundell, Robert (1971), Monetary Theory, Pacific Palisades, California, Goodyear Publishers.Neal, Larry (1990), The Rise of Financial Capitalism, Cambridge, Cambridge University Press.Obstfeld, Maurice (1993), ‘The Adjustment Mechanism’, in Michael D.Bordo and Barry Eichengreen (eds), A

Retrospective on the Bretton Woods System, Chicago, University of Chicago Press, pp. 201–68.Paul, R. and L.Lehrman (1982), The Case for Gold, Washington, DC., The Cato Institute.Pippinger, John (1984), ‘Bank of England Operations, 1893–1913’, in Michael D. Bordo and Anna J.Schwartz (eds), A

Retrospective on the Classical Gold Standard, 1821–1931, Chicago, University of Chicago Press, pp. 203–32.

FURTHER READING 237

Ramon, G. (1929), Histoire de la Banque de France, Paris, B.Grasset.Redish, Angela (1990), ‘The Evolution of the Gold Standard in England’, Journal of Economic History, 50, 789–805.Sayers, Richard S. (1936), Bank of England Operations, 1890–1914, London, P.S. King.——(1957), Central Banking After Bagehot, Oxford, Clarendon Press.——(1976), The Bank of England, 1891–1944, Cambridge, Cambridge University Press.Schubert, Aurel (1991), The Credit-Anstalt Crisis of 1931, Cambridge, Cambridge University Press.Schwartz, Anna J. (1982), ‘Reflections on the Gold Commission Report’, Journal of Money, Credit and Banking, 14,

538–51.Solomon, Robert (1977), The International Monetary System, 1945–1976: An Insider’s View, New York, Harper and

Row.Temin, Peter (1989), Lessons from the Great Depression, Cambridge, Mass., MIT Press.Triffin, Robert (1960), Gold and the Dollar Crisis: The Future of Convertibility, New Haven, Yale University Press.United States Gold Commission (1982), Report to the Congress of the Commission on the Role of Gold in the Domestic

and International Monetary Systems, 2 vols., Washington, DC, GPO.United States National Monetary Commission (1910), ‘Interviews on the Banking and Currency Systems of England,

Scotland, France, Germany, Switzerland and Italy’, Senate Document no. 405, Washington, DC, GPO.Viner, Jacob (1937), Studies in the Theory of International Trade, New York, Harper Brothers.Whale, P.B. (1944), ‘A Retrospective View of the Bank Charter Act, 1844’, Economica xi, 109–11.Williamson, John (1977), The Failure of World Monetary Reform, 1971–74, London, Athlone Press.Wheelock, David (1991), The Strategy and Consistency of Federal Reserve Monetary Policy, 1924–1933, Cambridge,

Cambridge University Press.

238 FURTHER READING

Index

adjustment mechanism 10–11, 13, 140–58;in Argentina 175–6;capital movements 13;monetary approach 57–75

Albania, foreign reserve ratio in 266Andrew, A.Piatt 46, 51Angell, J. 63arbitrage:

international 15, 60, 64–6, 92, 109, 293Argentina:

economy of 10, 112, 170;as example of periphery 169–81, 221–3;exchange rates in 171;and exchange standard 265–6;monetary reform in 170–1

Austria:central banking legislation 265–6;decline of foreign exchange 269;gold reserves 283;reconstruction loans to 267

Austria-Hungary:adopts exchange reserve system 264

Australia 202;foreign exchange reserves 266;overseas banks 170

balance of payments 188–90, 194, 198, 200;adjustment mechanism 140–58, 200–1, 300, 305;crises 10;discipline 295;equilibrium 11, 140, 144, 248, 273;and monetary theory 66–75 ;policy, Canada and the Nertherlands 307;settlements 34;temporary vs. permanent disturbances 248

balance of trade 80–2Bank Charter Act (1844) 5, 232, 242;

suspended 234bank deposits 232, 235, 250, 272Bank for International Settlements 22Bank of Belgium 46Bank of England 5, 39, 46, 58, 132–3, 135–6, 200–3, 267,

293;mentioned in Cunliffe Committee Report 230–45,259, 275, 295;operated gold standard 143, 164

Bank of France 46, 47, 132, 136, 143, 203;on gold-exchange standard 271–4, 278

Bank Rate 58;as controlled by the Bank of England 163, 164–6, 193,195, 198, 200, 235, 239, 250, 252

Banking School 99banks:

influence of commercial 270Baring Crisis (1890) 112, 203Belgium 5;

foreign reserve ratio 266;gold reserves 275;reconstruction loan 267

bimetallic standard 4–7, 101, 115, 214–16, 330Bismarck 217–18Bland-Allison Act 221Bloomfield, A.I. 143, 146Bolivia, foreign exchange in 266Bowley, A.L. 65brassage 101Bretton Woods System 1, 2, 9, 23–4, 101, 109, 120, 121,

290–312, 330Britain:

early gold standard 4, 119;gold movements 46, 51;and gold standard 57–75, 161–83, 246–61;as hegemon 21–2;international lender 20;

239

Napoleonic Wars 4–5, 116, 217;return to gold 114

Brussels, International Conference in 246Bryan, W.J. 19, 221–2Bulgaria 10;

foreign exchange 266, 269;transfers reserve balance 275

Bullionist Controversy 213Bullion Committee (1810) 213

California Gold Rush 5Canada 137, 202;

balance of payments policy 307;floating rate in 309;gold standard 76–93

capital:transfer of 52–3;mobility 91

capital flows:international 146, 161, 165, 182, 267–8, 275;mobility 91;short-term 265

Cassel, G. 149–50, 244central banks 118, 330, 149, 250, 292;

autonomy under gold standard 15–18, 210;cooperation among 22, 136, 203, 281;European 146, 264;and foreign balances 53–4;and foreign exchange reserves 267, 270;policy, internal repercussions of 252, 277;prewar policy of 46, 55;reserves of 293, 296, 311;and ‘rules of the game’ 60;statutes modified 265, 279;stocks of gold 237

Chile 10;foreign exchange reserves 266

China 7, 36Cleveland, G. 221Coinage Act (1792) 5Coinage Act (1834) 5Columbia:

foreign exchange 266commodity:

exports 3;money 107;prices 12, 275

consols 93, 112convertibility 7, 78, 109, 155, 247;

in Argentina 171, 181;capital flows 17;check on inflationary finance 13–14;defence by central banks 237, 297;legal and constitutional guarantees 104, 118;restrictions on 310–1;rule 107;suspensions 19–20, 100, 213

Cooper, P. 220cover system 7credibility 15–19, 23, 101–2, 105–6, 113, 188, 202–3, 213credibility bands 109–10credit:

expansion 233, 237–8, 255, 279;facilities, central bank 300;policies 148–9

creditor states 275crises:

see financial crisesCunliffe Committee 12, 230–45currency:

boards 24;convertible 299;devalued 303;discussed in Cunliffe Committee report 246–61;foreign 174, 177;gold and silver 152;inconvertible 241;paper 152, 155, 241, 246, 250;principle 99;reserve 293–4, 311;revalued 307;sterling 293, 308

Currency Act (1928) 133Currency and Bank Notes Act (1914) 235, 243Currency School 99Czechoslovakia:

foreign exchange 266, 269;transfers reserve balance 275

Danzig:foreign reserves 265

Dawes Plan 267deadweight losses 115debt, government 238debtor states 275–6deficits 300–5;

American 311deflation 8, 19–20, 47, 263, 267, 277, 305

240 INDEX

Denmark 268;foreign reserves 266

devaluation 305;in case of surplus 309;of sterling 257–8

discount rates (see also Bank Rate) 46–7;impact on capital movements 143

disequilibrium 291, 303, 305, 312dollar:

price in gold 101Douglas, P. 65Ecuador:

foreign exchange 266Edinburgh 40, 264Egypt:

foreign exchange 266El Salvador 10employment 56England:

and obstructions to trade 41Estonia:

foreign exchange 266Euro-bond market 308Euro-dollar market 300European Monetary System 2European:

Community 207;Union 24

exchange rates:fixed 131, 207–12, 292;flexibility 24;

floating 9, 141, 210, 255–6, 309–10;pegged 2;stability of 3, 10, 209, 146–7, 157–8, 172, 209;and stability of system 311;during suspension periods 113–15

exchange standard, gold 262–87;collapse of 268

expectations 9export cycles 200

Farmers’ Alliance 220Federal Reserve Bank 132, 135, 250, 267fiduciary system 7financial crises 10, 17–18, 89, 104, 203, 219, 221Finland:

foreign exchange 266, 269fiscal policy 101–3, 111Ford, A.G. 161–83

foreign exchange:reserves 8, 20, 45, 77, 201, 330;(table) 283;rate differentials 134, 136

France 18;bimetallic standard 5, 7, 115;disequilibrium in 303;fall of prices in 47;and foreign exchange reserves 266, 271–4, 278;as gold center 281;on the gold exchange standard 271;gold reserves in 275;and postwar exchange rates 271;postwar lending power of 253;repatriation of capital to 268, 272, 274

franc, French 152;depreciation of 267;in disequilibrium 303;recovery of 272;undervaluation of 272

Franco-Prussian War 6, 215Freycinet Plan (1878) 18

Genoa Conference (1922) 246, 263, 265, 277–8Germany 6–7, 18;

external deficit 303;foreign exchange reserve 278;and foreign reserve ratio 266, 269;reconstruction loan 267

Gilbert, M. 290–312gold:

arbitrage 15;brassage 101;bullion 101, 234, 236–7, circulation as money 232;coin 5, 7, 20, 101, 232, 234, 236–7, 298;discovery of 5, 8, 150, 155, 157–8, 183;domestic reserves of 157–8, 268–9, 271–2, 273;(table) 283;flows 14–15, 48, 51, 54, 70–1, 161, 177–9, 233;import and export of 167, 180–1;insufficiency of stocks 263;mining 277;newly mined 45, 150;outflow 233;price of 296, 310;reserves 200

Gold and Silver Export Control Act (1925) 115gold bloc 277Gold Commission (US) 330

INDEX 241

gold devices 15–16gold-dollar standard 290–312gold exchange standard 262–87, 291gold points 15–16, 106, 109, 131–9gold standard:

adjustment mechanism 3, 13, 77, 107;agitation against 18–19, 23, 217–23;actual operation of 54–5;asset equilibrium 83–5;balance of payments 79, 189–90;Bretton Woods System 290–310;Canada 76–93;capital flows 76–93;domestic commitment mechanism 105–10;emergence 212–17;exchange-rate stability 3;financial criscs 17–18;international 254–5;interwar 230;‘limping’ 46;monetary autonomy under 15–18, 79;and monetary theory 57–75;myth of 140;political economy of support for 207–24;portfolio-balance model 76–93;price stability 2–3, 9–10;postwar (I) 44, 230;prewar 234;restoration 330;suspension 112;variants 7–8;wartime 113–17

Grant, U.S. 219Great Depression 210, 262Greece 10;

foreign exchange 266, 269;inconvertibility episodes 116

Greenback:dollar 113, 218;Party 220

Group of Ten 153, 299

Harrison, B. 221Harrod, R. 48Hayes, R.B. 220Heckscher-Ohlin model 64hegemonic stability 21–2, 101, 120, 121Hoffman, Walther 64Honduras 10

Hume, D.:price-specie-flow mechanism of 11, 32–43, 57

Hungary:foreign exchange 266, 269;reconstruction loan 267

index, price 62, 150–1, 176India 133;

exchange standard 265–6inflation 9, 106–7, 272–3;

and balance-of-payments deficit 148;in Britain 235;CPI 107;as a consequence of extreme surplus 309–10;and gold exchange standard 277;persistence 106–9

interest rates:appropriate to the war 255;change in 49, 163;and deflation 267, differentials 46, 83, 134;and discount rate 233;in monetary theory 69–71;in 1920s 278;and prices 51;real 9;short-term 106;during suspension periods 113–15

International Monetary Fund (IMF) 121, 298–9, 302, 309international law of one price 12Ireland 231, 241Italy 2, 10;

external deficit 303;and foreign exchange 266, 270, 278;gold reserves in 275;stabilization credit 267

Japan 7;and exchange standard 265–6

Jevons, W.Stanley 62Junkers 217

Kenen, P.B. 144Keynes, J.M. 1, 14, 23, 58, 212, 230Kondratieff cycles 150, 151

Latin America 147Latin Monetary Union 7, 215–16

Latvia:foreign exchange 266, 269

242 INDEX

League of Nations 152, 230;Financial Committee 265;reconstruction loans 267

lender of last resort 22Lithuania 269London, competing with New York 281;

as a financial centre 58, 144, 161, 233, 238, 275, 294

McKinley, W. 221–2McKinley Tariff (1890) 221McCloskey, D.N. 32, 57–75Macmillan Committee 230;

report of 246–61Maier, K.F. 48Marshall, A. 8, 57Mexico 10;

peso crisis 2Mill, J.S. 8Minority Report 330monetary:

theory 57–75money:

commodity 149, 152;multiplier 68

money, high-powered 68money, paper 37–8;

demand for 66–71

Napoleon 115Napoleonic Wars 4–5, 116, 152, 217National Monetary Commission (US) 330Netherlands 6;

balance-of-payments, policy of 307;central bank policy 309;external deficit 303;gold reserves in 275

New York, as a financial centre 275, 281, 294New Zealand 170;

foreign exchange 266Newton, I. 4, 213non-traded goods 80–1, 86Norway 268;

foreign exchange reserves 266, 269note issue 8, 178–9, 240–1Nurske, R. 53, 143, 230

Ohlin, B. 58, 63

Panic of 1873 219

Panic of 1893 221Paris, as minor gold exchange centre 275Paris Club 153patent law 103–4periphery 111–13, 119, 146, 147, 161, 182, 187, 222–3;

capital flows 111- 13, 118, 187Persia 7Peru:

foreign exchange 266Poland:

foreign exchange in 278;foreign reserve ratio 266, 269;reconstruction loan 267;transfers reserve balance 275

Populism 220–22Portugal:

foreign exchange reserve 266, 269prices:

relative 88, 91;stability 3, 9;balance of trade 80–82

price-specie flow mechanism 11, 14, 77–8, 86–8, 195, 329primary products 3, 187, 222proportional system 7purchasing power parity 91

quantity theory 45, 50

railroad expansion 76Reichsbank 269, 270reserve ratio 89–90reserves:

central gold 243–4;foreign exchange 8, 20, 77, 330; 262–87, 293;gold 200;in gold-dollar system 311–2;sterling 293;surplus countries 308

Resumption Act (1875) 112Ricardo, D. 217risk premium 82, 83, 111–12, 118Romania, foreign reserve ratio 266‘rules of the game’ 14, 47, 54, 58, 60, 92, 110, 141, 213,

251, 280Russia 203;

exchange reserves 264Russian Revolution 272

Sayers, R.S. 46

INDEX 243

Scandinavian Monetary Union 216Schumpeter, J.A. 152Scotland 36, 39–40, 47, 231, 241seigniorage 115, 116Sherman Silver Purchase Act 110, 221silver 5, 10, 155Silver Crisis (1893) 112Sino-Japanese War 216small open economy 80Soviet Union:

foreign exchange 266Spain 10, 34;

foreign exchange 266;fundamental disequilibrium 303

speculative attack models 110stability, price 249;

exchange rate 140sterilization 54, 255sterling:

assets, French 274;declining confidence in 308;decreasing value of 152;as international currency 293;standard 161, 181, 183, 262

Sweden 247;foreign reserves 269;gold reserves in 275

Swift, Jonathan 34, 41Switzerland 6, 153;

floating exchange rate 310;gold reserves in 275

target zones 15, 119tariffs 20, 57, 162Taussig, F.W. 12, 45–6, 57, 63terms of trade 53, 144, 183, 198–9, 201time consistency 99, 103, 188Triffin, R. 140–58Triffin Effect 195

United Kingdom:and adjustment process 305;

as capital exporter 214–15;ERM 2;reduction in gold stocks 275;suspension of convertibility 19

United States:and bimetallic standard 5;conflict over gold standard 218–223;

Populism 220–2;and gold dollar system 291–312;and gold exchange standard 265;gold movements 46, 51;and gold standard 57–75;postwar lending power of 253;reduction in gold stock 275;restored gold convertibility 7;silver coinage 18–19

Uruguay:foreign exchange 266

USSR see Soviet Union

Venezuela:foreign exchange 266

Vietnam War 121, 302, 311Viner, J. 45

wages, downward adjustments 142Weimar Republic 19Whale, P.B. 32, 44–56, 161–2White, H.D. 1World War I. 19. 76

Young Plan 269Yugoslavia:

foreign exchange 266, 269

Zecher. J.Richard 32. 57–75

244 INDEX

Date post:	10-Oct-2014
Category:	Documents
Upload:	luis-gandarias
View:	453 times
Download:	39 times

The Gold Standard; Theory and History

Documents