The European Monetary System 50 Years after Bretton Woods: A Comparison Between Two Systems

R. A. Mundell Columbia University

Paper presented at Project Europe 1985-95, the tenth edition of the "Incontri di Rocca Salimbeni" meetings, in Siena, 25 November 1994.

This paper evaluates key features of the international monetary system that emerged in the post-war period and contrasts it with the European Monetary System that originated in the late 1990s and which came to be regarded as the prelude to European Monetary Union.
1. Fifty Years Ago

A half century ago, when the conference at Bretton Woods was held, the international situation was very different from today. The world war was still raging but its initiative had passed to the Allies. Allied troops had landed in Normandy and were advancing across France; German forces were being pushed up the Italian spine; and the Japanese Empire was in full retreat. Thoughts had already turned to the task of reconstructing the post-war international economic system. Haunting the deliberations were the four horsemen of the 1930s: economic isolationism, depression, nationalism and instability. The specific challenge was the need to set up institutional machinery to temper the business cycle, avoid currency chaos, protectionism, trade restrictions, exchange control, dried-up lending and discrimination. It was generally agreed that management of interdependence would require some surrender of national sovereignty in exchange for a voice in supranational power. The challenges came to be divided into two categories. One was the financial problem concerning the balance of payments, exchange rates, and international lending; the other was the commercial problem involving protectionism, discrimination and the growth of trade and employment. The main initiatives came from the United States and Great Britain. In the United States, the financial problems came to be centered in the Treasury, under Henry Moregenthau Jr., Harry Dexter White and others; and the commercial problem at State, under Will Clayton and others. A companion division of labor developed across the Atlantic with Lord Keynes, involved in his Clearing Union Plan, James Meade with his plan for a Commercial Union, and Lionel Robbins as head of the Economic Section of the War Cabinet Secretariat. The financial negotiations gave birth to the Bretton Woods twins. The IMF was given responsibility over exchange rates, liquidity, and short-term balance-of-payments finance; and the IBRD (World Bank) over long-term lending and development policy. Both institutions were handicapped in the early years by resources inadequate in relation to their task. More important, however, was that their functions in the early years were preempted by Marshall Plan aid; the recipients of that bilateral aid could

not double-dip by drawing on the Fund. The first decade served as a formative period that prepared the two institutions for more important work in the 1960s. The other half of the grand design, however, was less lucky in its institutional promise. It fell afoul of complicated negotiations, bad compromises, and finally, the United States Congress. The ITO Charter, signed in Havana in March 1948, came under attack in both countries: In Britain, which wanted commercial rules to be contingent on full employment, and in the United States, because of its escape clauses, lack of provision for investor protection and failure to rule out discrimination and exchange control; it was charged that only the United States and Switzerland would adhere to its provisions. (1) Fortunately, the less controversial commercial provisions of the Havana Charter (which was rejected by the U.S. Congress in 1950) had earlier been incorporated into the GATT.(2) The failure of the Havana Charter meant that the post-war airplane had to limp along on three engines, the IMF, the IBRD and the GATT. There was no explicit arrangements for ensuring price stability or full employment at the international level. Global macroeconomic stability had to rest on the stability of the international monetary system.
2. Order vs System

Twenty five years ago, at an earlier Bretton Woods retrospect, I made a distinction between a monetary system and a monetary order: A system is an aggregation of diverse entities united by regular interaction according to some form of control. When we speak of the international monetary system we are concerned with the mechanisms governing the interaction between trading nations, and in particular between the money and credit instruments of national communities in foreign exchange, capital, and commodity markets. The control is exerted through policies at the national level interacting with one another in that loose form of supervision that we call cooperation. An order, as distinct from a system, represent the framework and setting in which the system operates. It is a framework of laws, conventions, regulations, and mores that establish the setting of the system and the understanding of the environment by the participants in it. A monetary order is to a monetary system somewhat like a constitution is to a political or electoral system. We can think of the monetary system as the modus operandi of the monetary order. We are accustomed to thinking in terms of a given monetary system. In what follows I shall have to treat as variables what are usually, in economic analysis, regarded as constants. But the system may be undergoing change without our noticing it. The

"monetary order" may be rigid and unable to cope with the problems of the new system. If we fail to distinguish between system problems and order problems we may wrongly discard ideas about the system that no longer appear to work, or blame the order because it was created to house a system that had grown beyond it. In the latter case we have to ask whether it would be better to strengthen the order and suppress changes in the system, or modify the order to accommodate change in the system.
3. The Post-War System

The IMF Articles of Agreement signed at Bretton Woods, New Hampshire did not create a new international monetary system. On the contrary, it almost made it impossible for the existing international monetary system to function.(3) According to the agreement, countries were required to maintain exchange rates within one percent of the par value. This clause would have forced a revolutionary change in operating procedures in the United States which did not, as a rule, intervene in the foreign exchange market. As the system had operated since the devaluation of the franc in 1936 and the Tripartite Agreement in the same year, the United States bought and sold gold within narrow margins of its fixed parity. Most of the other countries fixed their currencies to the dollar, directly or indirectly through the pound sterling or one of the other reserve currencies. The exchange rate rule would have required the United States to support all the foreign currencies in the New York market or else close it. To negate this obligation--almost as an afterthought--the United States had a crucial sub-clause, Article IV-4(b) inserted into the articles. This subclause enabled a member to fix the price of gold in lieu of adhering to its exchange rate obligations. When the United States notif ied the Fund that it was buying and selling gold freely (to foreign monetary authorities for official monetary purposes) it was relieved of its exchange rate obligations. (4) This sub-clause established the legal basis for the asymmetrical post-war international monetary system, a system that been in existence since the late 1930s. (5) It was a dollar standard, anchored to gold. Was the anchored dollar standard a "system"? A system is an aggregation of diverse entities united by regular interaction according to a form of control. The anchored dollar standard can be identified as a system if we can perceive the order in the interaction of its components and outline the form of control. In a presentation before the Subcommittee on International Exchange and Payments of the U.S. Congress, I presented, in a paper entitled "Rules of the Gold Exchange Standard," the first complete analysis of the gold exchange standard as coherent system:

"The size of the deficit of a key currency "inner" country is a measure of the extent to which it is supplementing
increases in world gold holdings by currency reserves. The composition of the deficit provides a measure of the extent to which non-reserve "outer" countries judge, in their interests, its size to be excessive or deficient. The outer countries peg (directly or, through other currencies, indirectly) their currencies to the inner country's currency (the U.S. dollar) and thus act as residual purchasers or sellers of dollars, while the inner country (the United States) pegs the currency to the ultimate asset (gold), and thus acts as the residual buyer or seller of gold. This means that the size of the U.S. deficit determines the increase in reserves of the rest of the world, while its composition determines the change in reserves of the United States, given the rate of increase of monetary gold holdings in the world. When U.S. monetary policy is very expansive the outer countries have to buy up large amounts of dollars and this has direct and indirect inflationary consequences for the outer countries; similar, when U.S. monetary policy is restrictive there is a scarcity of dollars and this has deflationary consequences for the rest of the world.

The outer countries' protection against an excessive or deficit excess of dollars is to alter the composition of the U.S. deficit and thus affect the reserve position of the United States. When U.S. monetary policy is excessively expansive the outer countries can convert dollars into gold; this leaves the aggregate level of their own reserves unchanged, but it destroys world reserves because it reduces U.S. reserves. And similarly, when U.S. policy is unduly restrictive, the outer countries can convert gold into dollars, leaving their own reserves unchanged, but improving the reserve position of the United States. The composition of the U.S. deficit, which is under the control of the outer countries, is the mechanism by which the outer countries in their role as governors of the gold exchange standard cast their votes with respect to the appropriateness or inappropriateness of the aggregate size of the U.S. deficit. The vigor with wh ich the votes are cast, however, is circumscribed by the attachment of the inner and outer countries alike to the existing system. The outer countries can warn the United States by gold conversions, but they cannot lower the U.S. gold stock below the point at which it no longer pays the United States to continue running it; overly aggressive conversions would reinforce the "go-it-alone" forces in the United States...On the other hand, the U.S. freedom of action is also circumscribed in the sense that U.S. monetary policy must not be so inimical to the interests of the rest of the world that the outer countries decide, in their own protection, to opt out of the system by abandoning the dollar for gold. This is the system as it is supposed to operate. But within these constraints monetary and fiscal policies have to be carefully coordinated as both the U.S. and the outer countries try to preserve internal balance and external equilibrium..."

The rules of the game of the system constitute a combination of laws, commitments, conventions and gentlemen's agreements by which the inner country (the United States) pegs its currency (the dollar) to gold and the outer countries (Let us call them Europe) peg their currencies to the dollar, either directly or indirectly through another currency (such as the pound sterling or the franc). This means that the United States acts as the residual buyer or seller of gold, whereas Europe acts as the residual buyer or seller of dollars. The U.S. has to buy up any excess supply of gold on world markets and satisfy any excess demand out of its own reserves; failure to do so would result in the dollar price of gold moving away from the dollar parity. Europe, on its part, has to take up any excess of dollars offered to it or supply any excess of dollars demanded; failure to do so would result in the exchange rate moving away from its dollar parity. (6)

The boundary conditions are given by the U.S. stock of gold and Europe's stock of dollars; the United States cannot supply gold, nor Europe dollars, they lack. But there is an asymmetry in these conditions because, as long as gold and dollars can be supplied at the U.S. Treasury, Europe has access to additional dollars in exchange for gold. The total reserves (dollars and gold) of Europe therefore constitutes Europe's boundary condition, whereas the gold reserve of the United States represents the U.S. constraint. Control of the system rests on U.S. monetary policy, on the one hand, and Europe's gold-dollar portfolio on the other. When the United States expands the dollar supply it puts upward pressure on world incomes and prices--directly, because of interest rate effects and spending changes in the United States, and indirectly because of increases in European reserves. Similarly, when the United States contracts the dollar supply, it puts downward pressure on world prices. Europe's gold-dollar portfolio is the other control variable. When Europe converts dollars into gold it weakens the U.S. reserve position and stimulates or compels a monetary contraction(7) and when it converts gold into dollars it strengthens the reserve position and permits or compels a monetary expansion. Europe's goldpurchase policy thus influences U.S. monetary policy, while the latter "determines" world prices and incomes. When U.S. monetary policy is forcing inflation on the rest of the world, Europe can stimulate or compel a contraction by gold purchases; and when U.S. monetary policy is deflationary, Europe can entice an expansion by gold sales. We may thus express the control mechanism of the system as follows: The United States expands or contracts its monetary policy according to whether its gold position is excessive or deficient, and Europe buys or sells gold from the United States according to whether U.S. policy is causing inflation or deflation. The gold exchange standard therefore constitutes a "system" and it is with its implications that we must now be concerned.
4. The Mechanism of Adjustment

The system can be formulated in precise mathematical terms. (8) The purpose of doing so, of course, is not to lend an impression of spurious precision to central bank behavior.(9) It is rather to be explicit about the possibilities inherent in the mechanism, and to see what help the formulation is in uncovering possibilities easily missed by unaided intuition.(10) Figure 1 plots Europe gold holdings (G) on the abscissa and U.S. dollar liabilities (D) on the ordinate. Given the stock of gold in the worlds as a whole, the line RS plots the

relation between Europe's gold holdings (and therefore U.S. gold holdings as the residual) and U.S. dollar liabilities, which preserves the U.S. gold-reserve ratio at its desired level. Thus if Europe holds OR of gold, none is left over for the United States and dollar liabilities must be zero, whereas if Europe holds no gold (so that the United States holds it all), U.S. dollar liabilities are ORS. (Given one dollar's worth of gold as the unit in which gold is measured, the slope of RS with a negative sign, is the reciprocal of the U.S. reserve ratio, which is assumed to be less than 1.(11)) The line PQ plots the relation between Europe's gold holdings and U.S. dollar liabilities that preserves a given money supply (composed of dollars and gold) in the world as a whole, and therefore (supposing no lags) a given level of prices or incomes. Thus a dollar supply OQ would be exactly sufficient to maintain a given level of prices so Europe's gold holdings would have to be zero to preserve that price level; and, similarly, if the dollar supply were nil, gold holdings in Europe would have to be OP. The slope of PQ (with negative sign) is the reciprocal of the dollar price of gold, which, as already noted, it taken to be unity. The two lines represent targets of policy. When the U.S. reserve ratio is below (above) the desired ratio, which is the case whenever D is above (below) RS, U.S. authorities will reduce (increase) D; and when the world money supply is

Figure 1. The Adjustment Mechanism

above (below) that which preserves the price level implied by PQ, which is the case whenever G is right (left) of PQ, Europe will purchase (sell) gold to encourage restraint of expansion in the U.S. We thus get "direction forces" at each point in the diagram indicated, in each quadrant, by arrows. It may be seen at once that the system may be stable or unstable, depending on the intensity of the forces acting on the control variables. Let be the rate of dollar expansion in the United States expressed as a proportion of excess gold reserves; let be the rate of gold purchases in Europe expressed as a proportion of the excess of the world money supply over its equilibrium level; and let be the gold-dollar equilibrium reserve ratio in the United States. Then, assuming , and are constant, the system is stable or unstable according to whether / is greater or less than 1/. (In the borderline case, the system, from an initial point in the graph such as Z, would

oscillate hopelessly (and elliptically) around the equilibrium Q.) Thus if = 25 per cent, U.S. monetary contraction must be more than four times European gold purchases; if this were not the case the world money supply would increase (decrease) when the price level was too high and the dollar supply overexpanded (underexpanded). The potential for instability arises from the fact that the policy mix violates the principal of effective market classification.(12) The United States adjusts the dollar supply to protect its gold reserve position, while Europe adjusts its gold holdings to induce the United States to changes its monetary policy. Target variables are not matched to the targets they affect most strongly.. The system might break down even in the event that it is otherwise stable. Boundary constraints are given by the axes and by a vertical line from R1. Hitting the latter boundary would imply a suspension of gold payments in the United States--structural collapse--or a cessation of the rules of the game. Thus it is entirely possible that, given a sufficiently large disequilibrium such as that indicated by the point Z', the system could break down because it hits the barrier from R to the right, flattening both RRS and PQ.
5. The Mounting Crisis

At the end of 1950 the United States had 652.0 million ounces of gold and seven European countries had 95 million. By 1971, the United States stock had dropped to 291.6 million ounces, while Europe's seven countries held 481.7 million ounces. The seven countries were the original six countries that signed the Treaty of Rome and Switzerland. The bulk of this enormous shift of gold from the United States to Europe occurred in the late 1950s and early 1960s. It is clear from the analysis of the system that the anchored dollar standard had crisisladen potential if it is run in such a way that the United States policy is governed by its reserve ratio while Europe tries to control the world rate of inflation by pressure exerted on the composition of the US balance of payments. In the 1960s the United States and Europe were on a collision course with respect to the international monetary system--what Prime Minister Harold Wilson of the U.K. called a "monetary war." The risk lies that the variables would hit the boundary conditions determined by the stock and price of gold, bringing on a convertibility crisis. A higher price of gold--to make up for World War II and post-war inflation--would have provided more room for adjustment within the parameters of the system. But, in the 1960s, an increase in the price of gold was ruled out--mainly for political reasons.

The system problem was exacerbated in the late 1960s by the war in Viet-Nam which, in conjunction with the war on poverty, made U.S. financial policy overly expansionist, not just from Europe's point of view but from that of the United States as well. The fixed exchange rate system imposed correspondingly high inflation rates on Europe or the alternative of accumulating reserves and massive speculation. Especially in Germany, the system was criticized for producing "imported inflation"(13) into the surplus countries. Given the basic disequilibrium of the system--both because of the undervaluation of gold and the excessively expansionary monetary policies in the United States--it is perhaps surprising that it could last as long as it did. Beneath the threat-counterthreat struggle between the two continents, lay a powerful undercurrent of bluff. Neither-and least of all Europe--wanted the system to collapse. Rather than bring on a collapse of the system, countries changed their mode of operation. First, palliatives kept the system in operation for some years. Indeed, it was widely hoped that if the system could be kept in place for some years, efforts to modify it through the creation of SDRs--a kind of paper gold insofar as it was intended to be a gold substitute--would create a sounder basis for equilibrium. The palliatives were, first, mechanisms for reducing the excess demand for gold by economizing on it. The United States removed the 25% gold reserve requirement behind member bank deposits at the Federal Reserve in 1965, and behind Federal Reserve currency in 1968. These measures removed the internal gold reserve requirement and freed what was left of the US gold stock for the requirements of external convertibility.(14)
More important were the changes in the mode of operation as manifested in the dynamics of both parties. First, the elimination of the gold reserve requirements freed US monetary policy for more expansionary policies, and, as it turned out, reduced the pressure on the United States to preserve price stability; The US inflation rate, which had averaged less than 2% before 1965, rose to 4.4% in 1968, 5.4% in 1969 and 5.9% in 1970.(15) Second, Europe was now worried that, with US gold stocks now comparatively low, the United States would react to gold conversions by changing the system, scrapping convertibility; flexible exchange rates were anathema in Europe. Under the old mechanism, faced with the increased inflation rate in the United States, Europe would have pounced on the US gold stock. But to even threaten to do so in the late 1960s would have brought down the system. It had become an oligopoly in which countries could not act in their short-run self interest without taking into account the reactions of their partners. Another factor was the belief, as already mentioned, that the production of gold-guaranteed SDRs would save the system from destruction if it could be maintained for a few more years.

The system did break down, however, in the summer of 1971 after an earlier mark crisis in the spring. In August 1971 the United States rejected demands for conversions of dollars into gold by the U.K. and other countries. The gold window was declared closed and the dollar became officially inconvertible. This posed a problem for the other countries--whether to continue to peg to an inconvertible dollar, to let their currencies float jointly against the dollar, or to let their currencies float independently against the dollar and each other. 6. The Smithsonian System Already in the events leading up to the August crisis, a joint float had been considered. In the crisis of May 1971, when the mark (and guilder and Belgian franc) floated, Germany proposed a joint float with her other partners. At that time France and the European Commission proposed instead fixed parities and controls on capital movements. No agreement resulted. After the system crisis of August 15, 1971, Germany again proposed a joint float, again rejected by France. Exchange rates now floated, to the discomfort of most countries in Europe and small countries all over the world. Four months after the crisis, the major countries met at the Smithsonian Institution to reconstruct the international monetary system. The new system had some similarities to the old system. There was a modest realignment of exchange rates. Currencies were in effect pegged to the dollar. The main difference was in the U.S. commitment to covertibility. Although the official price of gold was maintained--it had been raised to $38 an ounce--the United States was not buying or selling any. The international monetary system was therefore no longer an anchored dollar standard. For all practical purposes it was a dollar standard.(16) How does the mode of operation of an unanchored dollar standard differ from an anchored dollar standard? It will be recalled that in an anchored dollar standard there are two targets (convertibility of the dollar and a stable price level) and two instruments (the supply of dollars and the U.S. gold reserve). The United States adjusts its policies to maintain convertibility and Europe controls the composition fo the US deficit to achieve inflation targets. Under the unanchored dollar standard the situation is different. Europe adheres to a monetary policy to achieve balance of payments equilibrium while the United States follows a monetary policy that achieves its inflation target. The adjustment process is reversed. Under these reversed assignments, the system does not exhibit the instability inherent in the anchored dollar standard. This is because it conforms to the principle of effective market classification. US monetary policy is matched to the price level which it affects most directly; and European monetary policy is matched to its balance of payments, which it affects most directly. International reserves are created as a result of the U.S. balance of payments deficit, which must equal, in equilibrium, both the excess supply of dollars in the United States and the excess demand for dollars in the rest of the world. In a growing world economy the equilibrium balance of payments deficit of the United States will be equal to the increase in dollar reserves desired in

the rest of the world. Given fixed exchange rates, U.S. monetary policy essentially determines the rate of monetary expansion in the world as a whole. These relationships can be expressed in terms of a model. Let R = k*Y* be the demand for reserves on the part of the rest of the world (ROW), where Y* is the GDP of ROW and k* is the fraction of Y* that ROW wants to hold in the form of dollar balances. Under a pure dollar standard, the US payments deficit P is equal to the growth of foreign exchange reserves so that dR/dt = P = n*k*Y*, where n is the rate of growth of nominal GDP in ROW. If in, say 1970, n* = .06, Y* = $4 trillion, and k* = .01, then P would equal $2.4 billion, not far from the average annual magnitude of the persistent balance of payments deficit of the period. Alternatively, if we express the US payments deficit as a fraction of US GDP, i.e., b = P/Y, we have b = n*k*, where is the ratio of the shares in world product of the rest of the world and the United States. The stability of the unanchored dollar standard, however, does not mean that it is an effective international monetary system. The United States has the ability to use monetary policy to achieve price stability, but it may not choose to do so. In the absence of the convertibility requirement, there is no mechanism (apart from moral suasion) by which Europe can induce the United States to change its monetary policy. The United States has complete freedom to impose its own inflation preferences on the rest of the world.(17) In short, it is a dominated system, the "Roman solution." In retrospect, as we look again at the post-war international monetary system, it is possible to see that features of the unanchored dollar standard had already crept into the post-war system in the late 1960s, especially after the elimination of the gold reserve requirements behind deposits and notes, and when the UN gold stock had dwindled to crisis levels.(18) To conclude, therefore, the unanchored dollar standard has the merit that it is free from the destabilizing possibilities of the anchored dollar standard; but the demerit that it is a dominated system which makes no provision for an accountability mechanism such as conve rtibility. By contrast, the anchored dollar standard has the possible demerit that its stability depends on careful timing of central banking policies, but the merit that it makes the reserve currency country accountable for its policies through the convertibility discipline.

The merits of the unanchored dollar system, formed at the Smithsonian Institution in December 1971, soon outlived their usefulness. The rules of the game required Europe to maintain balance of payments equilibrium and the United States to maintain price stability. But fulfillment of the two targets were not entirely independent. The United States, in the presidential election year of 1972, was following a monetary policy that was much too expansionary for Europe. The consequent aggravation of the European surpluses led to additional imported inflation in Europe, an inflation which many of the European countries tried to avoid by sterilization policies. Sterilization policies, however, cannot work in the long run. The surpluses, which automatically expand the money supply, can be temporarily offset by central bank sales of securities or changes in reserve requirements. However, these actions tend to raise interest rates and aggravate

capital inflows. With reserves rising rapidly, bullish speculation is aggravated and the central bank eventually has to abandon its exchange rate peg. This is indeed what occurred in the mark crisis of May 1971 when the Bundesbank had to buy up dollars at the rate exceeding $150 million an hour. A similar phenomenon occurred early in 1973. In February of 1973, Dr. George Schultz, a disciple of Milton Friedman and the new Secretary of the Treasury, sought a solution in devaluing the dollar.(19) The official price (at which the United States was neither buying nor selling) was raised $42.22, and some exchange rates were changed. But, in the absence of a more restrained monetary policy, devaluation only served to whet the appetites of speculators. Speculative capital outflows forced massive quantities of dollar balances on the rest of the world. In Japan, foreign exchange reserves (mainly dollar balances) soared from $2.6 billion at the end of 1969 to $15.2 billion at the end of 1972; and in Germany, foreign exchange reserves (again mainly dollar balances) had soared from $2.7 billion to $15.8 billion over the same period. By the spring of 1973, the unanchored dollar system was in the throes of a major system crisis. European countries were skeptical that the United States would take measures to tighten its monetary policy and reduce its deficit and the consequent flood of reserves to the rest of the world. Professor H.W.J.Bosman, in discussing the problem from the German point of view, wrote:

"Among these measures I do not include a substantial reduction of the U.S. balance of payments deficit. Of course, we would wish that such a reduction would take place, but we may not base ourselves on such a development as it is improbable that it will occur. It is for the U.S.A. from a technical point of view very difficult to use the instruments of the economic policy to improve the balance of payments position; it is furthermore politically still more difficult to do so, as imports and exports only play a minor role in producing and spending national income, and last but perhaps most important the political will does not exist to solve the problem. All statements concerning an improvement of the balance of payments position have not be taken too seriously. The recent devaluation of the dollar may improve somewhat the balance of payments position but it will not change it fundamentally, unless the U.S. are going to change fundamentally their commitments toward the world outside... Sticking to the dollar system...means that the dollar reserves of the other countries and especially of Europe will continue to grow and that there will be always a tendency to assume that the existing parities will not be the right ones, so that it can only pay to continue speculation against the dollar and in favor of the D.M. If nothing would be done against that speculative flow of dollars strengthening the already existing "normal" flow, the European and especially the German authorities would be faced with a continued creation of money.

Without maintaining that this can be regarded as the main cause of inflation, it makes the struggle against inflation still more difficult than it is already. Of course something can be done against the influence of large-scale dollar inflow on the banking system, e.g., by requiring a reserve of 100 per cent against foreign deposits. Experience tells us that this does not prevent outside funds continuing to flow into the country waiting for the currency to be revalued or to rise when left free. Moreover, the foreign funds can be invested in securities and then a measure like the 100 per cent reserve cannot be applied..." (20)

In June, the system was disbanded and the period of flexible exchange rates began.

7. Formation of the ERM

It will be recalled that Germany had proposed a joint float of the European currencies against the dollar both before and after the crisis that followed August 15, 1971, proposals that were rejected by the other countries. In the Spring of 1973, before the breakup of the unanchored dollar standard set up at the Smithsonian Institution, Germany again proposed a joint float of the European currencies, again rejected. One reason is not far to seek: A joint float would by no means have been a symmetrical outcome. Whenever the dollar weakened, a joint float would have had to rally around the mark. Neither France nor Britain were ready yet to acknowledge the mark as the natural center of the system. Nevertheless, the European countries had earlier indicated an interest in, if not consensus on, European monetary integration. Proposals for monetary integration go back to the late 1950s. The Treaty of Rome had called for individual policies for achieving equilibrium in overall balance of payments, maintaining confidence in currencies, and coordinating policies through collaborations of governments and central banks. Four years later, in October 1962, the European Commission submitted to the Council of Ministers a set of proposals for coordination of monetary and economic policies within the Community, leading to the eventual establishment of a monetary and economic union. In 1964, the Committee of governors of the Central Banks of the Member States was set up, along with budgetary and economic policy committees. In February 1968, the commission proposed that members commit themselves to adjust their exchange rate parities only by common agreement and to consider the elimination of margins on each others' currencies around the established parities. The next year, on February 12, 1969, the "Barre Report" called for concerted economic policies to ensure the attainment of agreed medium-term objectives. The Council agreed with many features on the Barre Report and committed members to prior consultation before a member altered its economic policies in such a way as to have an important impact on other members. The Community Summit conference at the Hague, on December 1 and 2, 1969, requested the Council to draw up a plan, based on the Barre Report, to establish by stages an economic and monetary union in the Community. On March 6, 1970, the Council authorized the creation of a committee, headed by Pierre Werner of Luxembourg, to draw up a plan for economic and monetary union. Along the lines of the Barre Report, central banks established a fund for balance of payments support by which members could draw up to $1 billion for a period of three, extendable to six

months. The Werner Report of October 8, 1970 recommended a program for the establishment by stages of an economic and monetary union by 1980. In its final form, the union was to have the following features: (1) a single Community currency (or else rigid and irrevocable fixing of exchange rates with zero margins and total interconvertibility); (2) complete liberalization of capital movements; (3) a common central banking system, organized along the lines of the Federal Reserve System; and (4) centralized responsibility in a Community "center of decision for economic policy" politically responsible to a European Parliament. These provisions were later watered down at French insistence, leaving undecided the exact division of powers between the Community and member states. The substance of the amended Werner Report was adopted by the Council of Ministers of February 9, 1971. Subsequent progress, however, was overtaken by the turmoil in the exchange markets in the spring and summer of 1971.(21) The impulse for European monetary integration fluctuates with the dollar cycle: it is strongest when the dollar is weak, as in the early and late 1960s and the early and late 1970s. After the implementation of the Snake under the Werner Plan, the next great thrust forward came in the wake of the weak dollar depreciation in the late 1970s. Following the Bremen summit in 1978, President Giscard d'Estaing and Chancellor Helmut Schmidt made the agreement to create the European Monetary System, which came into existence in March 1979.

8. Crisis of the ERM

The EMS was viewed as a prelude to monetary unification:

"The purpose of the European Monetary System is to establish a greater measure of monetary stability in the Community. It should be seen as a fundamental component of a more comprehensive strategy aimed at lasting growth with stability, a progressive return to full employment, the harmonization of living standards and the lessening of regional disparities within the Community. The Monetary System will facilitate the convergence of economic development and give fresh impetus to the process of European Union." (22)

The EMS went beyond the Snake in that it created an institution, the European Monetary Cooperation Fund, and introduced a kind of pre-money, the ecu, defined as a basket of the currencies of the EC countries, weighted by a formula that took account of both trade and GDP. The ecu was to serve as numeraire for the EMS

exchange rate mechanism; as a basis for indicating divergence; as the numeraire for central bank operations; and as a means of settlement between monetary authorities of the European Community. The Fund was to provide a source of ecus for settlement of central bank transactions against a deposit of 20% of gold and 20% of foreign exchange reserves. The exchange rate mechanism (ERM) within the EMS in theory was symmetrical with respect to its member countries, but in practice the DM became the "inflation anchor" of the system. The Bundesbank has been able to pursue monetary policies dictated by the requirements of internal balance (as its constitution requires). When a conflict exists--such as an appreciating mark (or a payments surplus) combined with inflationary pressure-- Germany has opted for tight money to prevent inflation rather than easy money to relieve the external pressure on itself or its partners in the EMS. By contrast, the other countries in the ERM have had to give priority to external balance, tightening or easing monetary policy according to whether their currencies ar at their upper or lower limit. In short, the ERM has all the hallmarks of a currency areas anchored to the mark and German monetary policy. The analogy with the dollar standard is apparent. But it should not be carried too far. The dollar standard was global, the mark standard, regional. As already noted, the size of a country's transactions domain plays a large role in determining its ability to cushion shocks in the system as a whole; the burden of international adjustment is distributed inversely in proportion to the size of country. With a country one-third the size of the US economy, Germany, qua anchor, is only one-third as stable. The turnaround in German fiscal policy due to unification brought about a reduction in the German current account from a surplus of $46 billion in 1990 to a deficit of $20.7 billion, a turnaround of $66.7 billion. This corresponded to an adjustment of 4.4% of German GDP, but the same absolute disturbance would have involved a turnaround of only 1.1% of US GDP. (23) The effect of the enormous inward transfer to (or reverse outward transfer from) Germany put pressure on the German price level, forcing the Bundesbank to react with higher interest rates--and the highest interest differentials (relative to the United States) favoring the mark--in years. Left alone with a neutral monetary policy (say a fixed rate of monetary expansion) the mark would temporarily have appreciated strongly in the spot market against all currencies, but going to a substantial discount in the forward market to reflect both the interest differential and a future weakness of the mark. Equilibrium would have been served by a substantial realignment involving a temporary appreciation of the mark or a downward realignment of the currencies of Germany's partners in the ERM. It should be emphasized, however, that had Germany been a larger country, the needed scale of real exchange rate adjustment would have been proportionately smaller.

Most economic events are spread across countries. The German unification disturbance was unique, a shock unparalleled since the oil price increases of the 1970s, but concentrated in a single country. One approach to the shock would have been for Germany to appreciate its currency against the dollar and against its partners in the Community. But this would have undermined its usefulness as an anchor and would have overvalued German labor (especially in the Eastern provinces) in the long run. In any case, France at this time would have resisted such a general appreciation of the mark and insisted on a proportional appreciation of the franc. The best policy--given the ERM--might have been for the Bundesbank to follow a monetary policy that would be neutral for Europe as a whole. Abstracting from ordinary economic growth, there are two candidates for a "neutral" monetary policy. One is that the growth of the money supply in Europe is kept unchanged. Under these circumstances, an increase in German government spending, financed by an increase in debt, would lead to somewhat higher interest rates in Europe and a somewhat higher ecu. The price of domestic goods would rise somewhat in Germany, and fall somewhat in the rest of Europe with international goods prices remaining constant. A Europe-wide monetary policy would have cushioned the impact of the German unification shock over the EMS part of the continent. It would have led to more inflation than the Bundesbank wanted, and more deflation than her partners wanted, but a more balanced equilibrium for the fixed exchange rate mechanism. It would have been the equilibrium imposed by an independent Board of Directors of the European Central Bank with power distributed among the Board in proportion to the economic sizes of its member countries. (24) The ERM crisis of September 1992 illustrates a basic defect of the EMS system. The mark anchor works as long as disturbances are not too large and arise from outside Germany. But disturbances in Germany would be neutralized only if Germany adopted a policy appropriate for Europe as a whole, not Germany alone.(25) The role of leader implies responsibility to the group not the individual. Self-centered behavior on the part of the leader undermines the whole system. European Monetary Union would eliminate much of that defect of the EMS. (26)
9.The Mechanism of Adjustment

A number of writers have identified the mechanism of adjustment in the ERM with the post-war or "Bretton Woods system." or even the gold standard. Thus De Cecco and Giovannini write:
"Indeed, the functioning of the EMS in its first ten years strikingly resembles the functioning of other fixed exchange rates regimes: the gold standard and the Bretton Woods regime. Like the earlier experiences, the conduct of monetary policy was under the control of a 'centre' country--West Germany. The other countries either largely accommodated Germany's monetary policy, as did Ireland, at an allegedly high price in terms of domestic

employment and welfare, or achieved temporary monetary independence with the use of capital controls, as did France and Italy. This pattern also characterizes also earlier experiences: monetary policy wad dominated by the United Kingdom during the gold standard and--at least to some extent the U.S. during the Bretton Woods years. Capital controls were also used by countries other than Britain during the gold standard, and by the European countries, including West Germany, during the Bretton Woods years."(27)

It should be clear by now that this view is incorrect. The ERM system was not like the gold standard or even the "Bretton Woods regime." First of all, let us consider the analogy to the gold standard. It is true that London was the center of the world's capital market, that sterling was widely used as an invoice currency, and that the sterling bill served as a secondary substitute means of payment. But the major determinant of the world price level was the stock of monetary gold, not British monetary policy. Decades before World War I the United States had become the largest economy in the world and the largest holder of gold. The British stock of gold was incredibly small relative to the other major holders and British interest rates followed, not determined, the world cycle. News about economic events originated in London, the world's main financial center, but they were not principally determined in London. The gold standard was anchored to gold, not the pound. It is also true that the gold exchange standard of the 1920s was anchored to gold, and it was the abortive restoration of gold that aggravated demand for it and brought about the deflation of the 1930s. After the Tripartite Agreement of 1936, the United States was the only country on some semblance of a gold standard (although private citizens could not hold gold and the dollar was not redeemable). The dollar had a key-currency role in the post-war period. But the gold-reserve and convertibility requirements imposed a discipline on US monetary policy. In the absence of these requirements US monetary policy would have been more inflationary than it was. This is suggested by the more expansionary policy after the gold reserve requirements were abolished in the late 1960s; and by the even more inflationary policy in the United States when the convertibility requirement was scrapped. The correct analogy for the ERM is the unanchored dollar system set up at the Smithsonian Institution. As the Bundesbank report for 1971 correctly claimed, the system had become a dollar standard. The United States pursued its own inflation preferences in its monetary policy and that determined the inflation rate for the currency area as a whole. The monetary policies of the outer countries were disciplined by their balance of payments, but the United States policy rested on its self-discipline, without the accountability of convertibility. When these inflation preferences clashed with those of Europe, a crisis emerged that ended in the partial breakup of the dollar standard system. (28)

The ERM had gravitated to a mark standard in the 1980s. The mark standard of 198092 had the same type of adjustment mechanism as the dollar standard of the early 1970s. Under the ERM system, the Bundesbank pursued its own inflation preferences. The other countries pursued policies to stabilize their exchange rates and preserve balance of payments equilibrium. The ERM crisis occurred because Germany's monetary policy was too tight for its neighbors. (29)
10. Conclusion

My task in this paper was to contrast two systems: (1) the International Monetary System that came into being after the signing of the Bretton Woods Agreements fifty year ago, and which broke down in steps in the late 1960s and early 1970s; and (2) the European Monetary System which came into being after an agreement signed in Bremen between France and Germany in 1978 and which threatened to break down in various steps during 1992 and 1993. Despite superficial similarities, there was a fundamental difference between the two system. Consider first the earlier system. This was an anchored reserve-currency system. Under the Bretton Woods arrangements, most of the other currencies were pegged to the dollar, whereas the US dollar was pegged to gold. US monetary policy was disciplined by its internal gold reserve ratio and by its commitment to external convertibility of the dollar (for foreign monetary authorities). European countries were constrained by the discipline of balance of payments equilibrium, but had an additional weapon--conversion of dollar balances--with which they could put pressure on the United States to contract or encourage it to expand. Although the system was asymmetrical in the sense that the dollar had a special role, there was an exchange of commitments that distributed control between the United States and the outer countries. The ERM system was basically different. Under the ERM system as it came to operate after the Plaza Accord, the outer countries were disciplined by the balance of payments under fixed exchange rates while the center country, Germany, could pursue an independent monetary policy geared to its version of price stability. Germany could pursue its own inflation preferences without any accountability mechanism; the other countries had no instruments to alter German monetary policy. It was a dominated system, a mark standard, the Roman solution. The correct analogy to the ERM is not the gold standard or the post-war system; it is the regime set up at the Smithsonian Institution in December 1971 and that lasted (albeit with a second devaluation of the dollar) until June 1973. This system was an unanchored dollar standard in which the United States could pursue its own inflation

preferences, without any accountability mechanism; the other countries had no instruments to alter US monetary policy. It was a dominated system, a dollar standard, the Roman solution. Both the dollar and mark standards threatened to break down when the center country pursued monetary policies that were at variance with the needs of the outer countries. The dollar standard broke down in 1973 because US monetary policy, taken in conjunction with the explosion of the Eurodollar market, flooded surplus countries with reserves. Rather than accept the inflationary consequences of expansion, or revalue the currencies in fundamental disequilibrium, the system was allowed to break up.(30) Similarly, the mark standard threatened to break down when German monetary policy, over-reacting to the unification shock, followed a money policy that was too tight for the rest of the ERM. Rather than import deflation (or less disinflation than was politically acceptable), several countries devalued or left the ERM. The mark standard broke up--or, more correctly, was transformed, because Germany monetary policy in the wake of the unification shock was too contractionary for the rest of the ERM. When, in the summer of 1992, the mark was soaring against the dollar-reaching a dollar low of DM1.385--the Bundesbank should have reacted to the error signal and moderated its policies. The defect of both the dollar and mark standards was that the monetary policies of the anchor countries were out of line with the interests of their partners. In the case of the United States, monetary policy was too inflationary. In the case of Germany, monetary policy was too deflationary. There is an inherent defect in any unanchored currency standard that lacks a mechanism by which the partner countries can have some influence over the monetary policy of the leader and therefore the currency area as a whole.

1. I have discussed some aspects of these issues in "Multilateral Commercial Diplomacy," The National Banking Review 3, No. 2 December 1965): 193-199. 2. Aware that negotiations for the charter would last a long time, interim agreements had been made soon after the war ended. The U.S. tariff, despite reductions since the Smoot-Hawley Tariff under the Reciprocal Trade Agreements Act, was still a main barrier to trade. Crucially, the Administration got authority from Congress to reduce tariffs to 50 percent of the 1945 level,

giving effect to a crucially important U.S. proposal at the 1946 London Conference to embody new commercial treaty rules, including those expected to be incorporated later in the ITO, in a General Agreement on Tariffs and Trade. It was out of this ad hoc memorandum that GATT--the only living remnant of the Havana Charter--was born. 3. I have discussed this issue in greater depth in "Tales from the Bretton Woods," in Retrospective on the Bretton Woods System (M. Bordo and B. Eichengreen, eds.) Cambridge, MA: National Bureau of Economic Research, 1991. 4. There were two other problems with the Articles of Agreement for the other countries: (1) Literally, Article IV required each country--unless it adopted, like the United States, Article IV4(b)--to keep the currencies of all member countries within one percent of parity, clearly a ridiculously wasteful arrangement; subsequently it was established that a country that was fixing its currency to one convertible currency would be deemed to be satisfying its obligations. (2) A country in, say, the sterling area that kept its currency fixed to the pound within 1 per cent of parity, whereas the pound was fixed within margins on the dollar, could experience exchange margins of 2 percent of parity or more against the dollar; arrangements of this type were subsequently condoned as a "multiple-currency practice." 5. Why did countries shift from intervention in the gold market to intervention in the dollar market? My conjecture is that the shift occurred after the "gold scare" of 1937 when a change in the U.S. commitment to gold was widely discussed. Transactions costs were much lower under the dollar standard than under gold. 6. Under the Bretton Woods arrangements, parities for currencies were specified in terms of the number of grams of gold or number (or fraction thereof) of 1944 U.S. gold dollars equivalent to one currency unit. Market exchange rates, however, were determined directly by the foreign exchange market including intervention by the exchange authorities. 7. I shall speak throughout of "expansion" and "contraction" and rising and falling prices rather than increases or decreases in the rate of expansion or contraction of money and prices, but the reader should have no difficulty in making his own translation into a growing economy. 8. See my paper, "The Crisis Problem," in Monetary Problems of the International Economy (R. A. Mundell and A. K. Swoboda, eds.) Chicago: University of Chicago Press 1969: 343-349; reprinted in R.A.Mundell, International Economics New York: Macmillan 1968: 282-288.. 9. I know of no central banker who wold like to think that his actions can be expressed in a system of differential equations. 10. For a mathematical formulation of the system and formal proofs about its behavior see my paper "The Crisis Problem" in Monetary Problems in the International Economy (eds. R.A. Mundell and A. K. Swoboda) Chicago: University of Chicago Press, 1968:343-349, reprinted in R.A. Mundell, International Economics, New York: Macmillan, 1968..

11. D may be taken t o represent the aggregate supply of non-gold money in the world as a whole, rather than simply the foreign central bank holdings of dollars, since we are not concerned here with exchange margins, imperfect substitutability of one currency for the other, or more than one currency areas, while the gold stock, equal to OR1, can readily be adjusted to take into account private gold hoarding. 12. The principle of effective market classification was first discussed in my paper, "The Monetary Dynamics of International Adjustment Under Fixed and Flexible Exchange Rates," Quarterly Journal of Economics 74 (May 1960): 227-257 (reprinted in International Economics, loc. cit., 152-176); and developed further in "The Appropriate Use of Monetary and Fiscal Policy for External and Interna l Stability," IMF Staff Papers (March 1962): 70-79 (reprinted in International Economics, loc. cit.,233-239). In the latter article the principle was formulated as follows: "...Policies should be paired with the objectives on which they have they most influence. If this principle is not followed, there will develop a tendency either for a cyclical approach to equilibrium or for instability...On a still more general level, we have the principle that Tinbergen made famous--that to attain a given number of independent targets there must be at least an equal number of instruments. Tinbergen's principle is concerned with the existence and location of a solution to the system. It does not assert that any given set of policy responses will, in fact, lead to that solution. To assert this, it is necessary to investigate the stability properties of a dynamic system. In this respect, the principle of effective market classification is a necessary companion to Tinbergen's Principle." 13. The term "imported inflation" was invented by Professor Wilhelm Rpke of the Graduate Institute of International Studies in Geneva. 14. By 1968 the private market demand for gold had caught up with market supply and the gold pool--composed of the few countries that were major gold holders--was disbanded. In March of that year, the central banks made the decision to withdrew from the private market, allowing the private market price to rise above the official price. This measure prevented any further drain of gold from official to private stocks. 15. Although the rate of increase of US money held by domestic residents (except in 1968) did not appear to surge--it was 4.6% in 1967, 7.2% in 1968, 5.8% in 1969 and 3.6% in 1970-exported money in the Eurodollar market was soaring, a phenomenon the Federal Reserve studiously ignored. See my article, "World Inflation and the Eurodollar," Economic Notes 1, No.2, 1971. 16. In its report for 1971 the German Bundesbank declared that the gold exchange standard had for the time being been replaced by a dollar standard. 17. A reserve currency country has, moreover, an incentive to inflate above what would otherwise be national inflation preferences. This arises because of the additional seigniorage provided by the inflation tax. I analyzed this issue in "The Optimum Balance of Payments Deficit," in Stabilization Policies in Interdependent Economies (Pascal Salin and Emil Claassen, eds.) Amsterdam: North-Holland Press 1972: 69-86.

18. In one year, 1969, the United States actually acquired 27 million ounces of gold, a consequence of the events of May 1969 in France. Nevertheless, gold conversions continued so that the United States lost about 60 million ounces between 1969 and the end of 1971. 19. It should be remembered that many economists--Keynesians and Monetarists alike-misdiagnosed the disequilibrium of the 1950s and 1960s as a national problem of correcting the U.S. balance of payments rather than a problem inherent in the structure of the international monetary system. The major problem of the 1950s and 1960s was the undervaluation and scarcity of gold, which could not be corrected by a small change in exchange rates. 20. H.W.JBosma n, "Exchange Rate Policies," in European Economic Integration and Monetary Unification Brussels: European Communities Commission Directorate-General for Economic and Financial Affairs (October 1973): 25-26. 21. Arthur Bloomfield has provided a detailed discussion of the early developments toward monetary integration in the community. See A. Bloomfield, "The Historical Setting," in L.B.Krause and W.S.Salant, European Monetary Integration and its Meaning for the United States, Washington, D.C.: The Brookings Institution, 1973: 1-19. 22. Statement in the conclusion of the Presidency of the European Council of 4 December 1978, quoted in M. De Cecco and A. Giovannini, "Does Europe Need a Central Bank," in M. De Cecco and A. Giovannini (eds.), A European Central Bank? Perspectives on Monetary Unification after Ten Years of the EMS. Cambridge: Cambridge University Press 1989: 2. 23. On this argument see my paper, "The Proper Distribution of the Burden of International Adjustment," National Banking Review 3(September 1965): 81-87; reprinted in my International Economics (Ch. 13); see also my "Uncommon Arguments for Common Currencies," in The Economics of Common Currencies (H.G.Johnson and A.K.Swoboda, eds) London: Allen & Unwin, 1973: 114-132. 24. The Maastricht Treaty, however, assumes an independent central bank with one director from each country; this solution may, however, prove to be a mistake. 25. The best defense for the Bundesbank's policy is that there is a critical rate of inflation at which expectations of inflation lead to a crack in the bond market and compensatory wage demands, that threaten to set in motion a wage-price spiral (the correction of which, later on, would entail a higher cost in terms of excess unemployment). To the extent that this argument is valid, it seems to be an argument for an appreciation of the mark within the ERM. 26. Critics may argue, however, that flexible exchange rates would have resulted in a better solution in the case of a shock of the size experienced by Germany. In the same view some advocates of flexible exchange rates have argued that flexible exchange rates between the NorthEast and South-West regions of the United States would have cushioned those regions against the oil shocks of the 1970s and 1980s; and, similarly, that a separate currency for New York City during its debt crisis of the mid-1970s would have allowed it to pay off its debt in depreciated currency. These arguments, however, seem to be short-run in nature; a flexible exchange rate

regime has the defect that it sets in motion expectations of future changes that nullify the effectiveness of the adjustment features of flexible exchange rates. The best solution, signalled by the sharp appreciation of the DEM against the dollar, would have been a more expansionary monetary policy in Germany. 27. De Cecco and Giovannini, loc. cit.,.3. 28. The breakup was only partial because, while the surplus European countries floated, most of the Fund members in the rest of the world that had fixed their currencies to the dollar stayed there. Whereas the inflation preferences of the United States were higher than that in the surplus European countries (or at least Germany and Holland), they were substantially lower than most other countries. 29. Differences in "inflation preferences" are not always easy to measure. For example, in a highly integrated currency area, there is a single inflation rate for the area as a whole and it is not, in general, possible to distinguish between inflation rates in the center and outer countries. During the late 1960s and early 1970s, for example, US inflation rates did not differ much from most of the European countries; and German inflation rates were not much if any lower than her neighbors. Even differential rates of monetary expansion (relative to output) may not correctly reflect inflation preferences where part of the money supply is held abroad. For example, US monetary expansion was not exceptionally rapid in the 1960s and 1970s (although it did increase substantially in the latter decade) when measured by domestic monetary aggregates, making no allowance for the inflationary effect of soaring dollar balances abroad both in official hands and for use as reserves in the Eurodollar market. 30. It would be a mistake to claim that the Committee of Twenty was in any sense forced to disband the system. Without denying the US blame for an excessively expansionary monetary policy, the Committee's decision to scrap the dollar standard for flexible exchange rates was an unwise one. A better alternative would have been for Germany, the main country in fundamental disequilibrium, to revalue the mark. German foreign exchange reserves had almost doubled, rising from $11.5 billion at the end of 1971 to $20.8 billion at the end of 1973. The other countries, including Japan, were not in fundamental disequilibrium; Japan's reserves, for example, fell from $12.6 billion to $8.5 billion over the same period. Rather than destroying the system because the exchange rate of a country comprising (then) 6% of world GDP was inappropriate, it would have been better for Germany alone to leave the system temporarily and then refix at a higher parity. (Germany could argue in defense, however, that because of world inflationary pressures it would be better for the rest of the world to adapt to Germany rather than the other way around even if Germany were only 6% of world GDP.) However, whatever the defects of the dollar standard, including the correct charge that it was too inflationary, its maintenance would have resulted in less inflation in most countries than actually occurred. Only a handful of countries have had less inflationary pressure, and depreciated against the dollar, since 1973.