Information about Western Hemisphere Hemisferio Occidental

3. Changing Perspectives on the International Monetary System

Jacob Frenkel, and Morris Goldstein
Published Date:
September 1991
  • ShareShare
Information about Western Hemisphere Hemisferio Occidental
Show Summary Details
Robert Solomon

In his long and distinguished career at the International Monetary Fund, Jacques Polak was intimately involved with most of its activities. Among those were the functioning and reform of the international monetary system. He was “present at the creation” of Bretton Woods in 1944 and has witnessed, and contributed to, the many alterations in the system since then. As one reviews the evolution of the international monetary system, one finds it almost impossible to detect an aspect of that subject about which Jacques Polak has not written.

What I propose to do in this paper is to examine the way preoccupations about the international monetary system have changed over the years since Bretton Woods. Whose preoccupations? Mainly those of officials, who are the ones with the power to reform the system, but I shall not ignore the views of academics. I shall follow a “then” and “now” procedure: depicting the concerns that were prevalent in the 1950s and the 1960s and comparing them with current concerns.

The functioning of the international monetary system is usually discussed in terms of adjustment, liquidity, and confidence. I shall concentrate on the first two of these. Confidence refers to the potential switching by monetary authorities from one reserve asset to another, especially conversions of reserve currencies into gold. While other stability problems may afflict the system, that particular danger no longer exists. What is possible is the switching from one reserve currency to another and that has been happening but apparently has not created significant instability in recent years.

I. Balance of Payments Adjustment—Then

In the 1950s and 1960s, exchange rates were pegged and were expected to be altered, if at all, only in the event of “fundamental disequilibrium.” Such alterations occurred only in extremis; a crisis was often required to bring about or make acceptable a change in par values under the Bretton Woods system. In the early 1960s, the prevailing attitude among officials of the major industrial countries was that their exchange rates should remain fixed. In 1964 the Group of Ten characterized the international monetary system as a structure based on “fixed exchange rates and the established price of gold.”1

In the early postwar years, it was widely believed that the “dollar shortage” would persist for a long time on the basis of the technological prowess of the United States and its supposed rapid rate of productivity growth. The dollar shortage implied a potentially larger U.S. current account surplus than actually existed, since most countries were restricting imports from the “dollar area.” It was thought that other countries would be unable to finance the corresponding current account deficits. Thus the United States accepted other countries’ restrictions on dollar imports. U.S. policy also sought to encourage capital flows to the rest of the world so as to make it possible for other countries to increase both their dollar imports and their gold and dollar reserves. The focus of adjustment was on Europe and Japan; the aim was to help those countries achieve sustainable balance of payments positions and to do so without resort to import restrictions. It was assumed that the United States could ignore its own balance of payments. The decline in the U.S. trade surplus from its 1947 peak of $10.1 billion (4.3 percent of gross national product (GNP)) was welcomed, but it was believed that a repressed demand for U.S. exports existed throughout the world.

As it turned out, Western Europe staged a remarkable recovery during and after the Marshall Plan (1948–52). Output and productivity rose rapidly (more rapidly than in the United States), exports grew markedly, international reserves were reconstituted, and, in 1958, convertibility on current account was established.

In these conditions, the flow of private U.S. capital to Europe increased. The United States developed an overall balance of payments deficit as reflected in a decline in its reserve assets (mainly gold) and an increase in its foreign liabilities. Actually this deficit appeared as early as 1950, during the Marshall Plan, but it was not labeled as a “deficit” at that time. It was called, in IMF Annual Reports and elsewhere, “a net transfer of gold and dollars” to the rest of the world and was welcomed as a sign that war-torn countries were recovering, since they were able to accumulate reserves instead of spending all their receipts on imports. Only in the late 1950s and early 1960s did U.S. deficits come to be frowned upon in Europe and to be worried about by U.S. officials. I believe that the first time an IMF Annual Report used the word deficit to characterize the U.S. balance of payments was in 1960, covering the year 1959.

Attention was focused much more on reserve changes than on current account positions. In fact, the U.S. current account was almost steadily in surplus; between 1950 and 1970 it was in deficit only in three years: 1950, 1953, and 1959. In the 1950s, the current account surplus averaged $0.6 billion a year; in the 1960s, it was $3.3 billion a year, dwindling late in the decade under the impact of the Viet Nam war. The overall deficit was the product of an excess of capital outflow over the current account surplus. The U.S. balance of payments statistics showed what was then called a “liquidity deficit” (the decrease in reserves plus the increase in liabilities to both official and private holders abroad). The liquidity deficit amounted to $1.7 billion a year, on average in 1950–59, but ranged above $3 billion a year in 1958–60. In the 1960s, the liquidity deficit averaged $2.6 billion a year. The official settlements deficit (adopted in the mid-1960s and initially called the balance settled by official transactions, which counted only liabilities to foreign monetary authorities) was smaller: $1.2 billion a year in the 1950s and $1.1 billion a year in the 1960s.

Finance ministers and central bank governors became quite exercised about the U.S. deficits in the late 1950s and early 1960s. In fact, President Kennedy is said to have coupled the U.S. balance of payments deficit with nuclear war as the two dangers that he feared most. The concern expressed in Europe was partly that the United States was exporting inflation via its balance of payments deficits. And both European and U.S. officials worried about the deteriorating net reserve position of the United States.

By today’s standards those deficits appear trivial. Were they trivial when we scale them back to the magnitudes of those earlier decades? The official settlements deficit amounted to 0.3 percent of U.S. GNP in the 1950s and less in the 1960s. Relative to the GNP of the six member countries of the European Community plus the United Kingdom, the entire U.S. official settlements deficit came to about 0.6 percent a year in the 1950s and half that much in the 1960s. It seems evident that in most of the 1950s and 1960s, the U.S. deficits were rather small, but they appeared large in 1958–60. In those three years, the official settlements deficit averaged $3 billion a year, and the gold and dollar reserves of the industrial countries of Western Europe rose by about 50 percent.

While the United States was in “overall deficit” and Europe as a whole was in overall surplus in most of the 1960s, some European countries had occasional deficits that created problems for them: Italy in 1963–64, the United Kingdom in 1964–67 (culminating in the sterling devaluation of November 1967), and France in 1968–69 (leading to the franc devaluation of August 1969). Each of these countries was forced to adopt stringent domestic policy measures to cope with its deficit.

But in the United States and the surplus nations of Europe (Japan was not yet regarded as a country in surplus) balance of payments adjustment measures were few and far between. The U.S. Treasury and Federal Reserve devised a number of techniques to protect the U.S. reserve position but these can hardly be called adjustment measures. The income tax cut proposed by the Kennedy Administration in 1963 was regarded as an opportunity to alter the mix of fiscal and monetary policy so that higher interest rates might discourage capital outflows. And beginning in 1963, a number of direct restrictions on capital outflows were adopted.

Among European countries in surplus, Germany and the Netherlands appreciated their currencies by 5 percent in 1961. Germany and Switzerland later attempted to restrict capital inflows with penalty reserve requirements on foreign-owned deposits. In late 1968, Germany raised border taxes on exports and reduced them on imports and, in September 1969, this was replaced by a revaluation of the deutsche mark.

It is evident that an asymmetry existed in the reasons for exchange rate adjustment as between countries in surplus and countries in deficit. Those in deficit devalued when their reserves were running low and doubt arose as to their ability to finance continued external deficits. Those in surplus revalued because they wished to avoid the monetary effects of the inflow of foreign exchange that their central banks were forced to purchase in order to maintain their exchange parities.

In the late 1960s, the current account surplus of the United States declined sharply, while that of Europe and Japan increased. And in 1967–69, the exchange rates of three major countries had been changed. This led to greater attention even in official circles to the adjustment process in general and to the possible need for greater exchange rate flexibility in particular. The earlier notion that exchange rates should remain fixed among industrial countries was giving way. The IMF undertook a study of The Role of Exchange Rates in the Adjustment of International Payments, published in the autumn of 1970, which guardedly suggested more prompt adjustment of par values.

Members of the academic community organized the Bellagio group in 1963–64 (under the leadership of Fritz Machlup, William Fellner, and Robert Triffin) and issued a report that, among other proposals, recommended more frequent changes in exchange rates “than currently contemplated by major governments.2 Later the so-called Bürgenstock group held conferences and published a volume of papers on exchange rate flexibility.3

While some officials in Germany, Italy, and the Netherlands were sympathetic to increased flexibility, those in the United States who had such sympathy faced a dilemma: under the existing international monetary arrangements, with the dollar convertible into gold for monetary authorities at the fixed price of $35 an ounce, the dollar could not be devalued without setting off a run on the U.S. gold stock. All the United States could do was to urge countries in surplus to revalue their currencies upward. And Europeans looked upon this asymmetrical approach with little enthusiasm.

Thus nothing came of the impulse to reform the system so as to provide for greater flexibility of exchange rates.

And apart from Germany’s small revaluation in 1969, no exchange rate adjustments were made to reverse the sharp decline of the U.S. current account surplus. As for other adjustment measures, the United States imposed stringent controls on the outflow of capital in early 1968 and, rather too late, tightened its fiscal policy while the Federal Reserve pursued a restrictive monetary policy. This led to a large flow of short-term capital to the United States, via U.S. bank borrowings from their branches in the Eurodollar market, which created an official settlements surplus that masked the deterioration in the current account. But the United States moved into mild recession in late 1969. As Federal Reserve policy eased and interest rates declined, short-term capital flowed out and the dollar holdings of European countries rose rapidly, setting the stage for the crisis of 1971.

In August 1971, the United States abandoned convertibility into gold and the way was prepared for the floating regime that started in March 1973.

II. Balance of Payments Adjustment—Now

The past decade has featured substantial current account imbalances—principally the U.S. deficit and the surpluses of Germany and Japan. In the first half of the decade these imbalances were associated with the fiscal policies being pursued; the United States developed a large budget deficit, while budgetary policy in Germany and Japan shifted toward surplus. Exchange rates moved to accommodate the fiscal-monetary mixes, and the 80 percent appreciation of the U.S. dollar in 1980–85—the largest movement of a real exchange rate of an industrial country in modern history—helped to produce the U.S. current account deficit, which at its peak in 1987 amounted to 3.6 percent of GNP. Japan’s surplus also peaked in 1987 at 3.6 percent of GNP (in 1986, it was only slightly smaller in dollar terms but amounted to 4.4 percent of GNP). Germany’s surplus was at its maximum in 1989 at 4.6 percent of GNP. In the 1960s, the United States had been in steady current account surplus; its largest surplus, in 1964, came to 1.0 percent of GNP. Japan was mostly in deficit in the 1960s, but in 1969 it had a current account surplus equal to 1.2 percent of GNP. Germany’s largest surplus of the 1960s, in 1968, was equal to 2.2 percent of GNP. Thus, current account imbalances were larger, relatively as well as absolutely, in the 1980s than earlier.

What can be said about balance of payments adjustment in recent years? First, it has to be noted that the notion of an overall balance or even an official settlements balance is hardly mentioned these days. With a few exceptions—notably 1987, when official intervention in foreign exchange markets was heavy—current account imbalances have been financed mainly by flows of private capital.

If we ask what policies have been directed to reducing the current account imbalances, we can point to the Plaza Agreement of September 1985, which was aimed at encouraging a depreciation of the dollar that had already begun in February of that year. But the Louvre Accord of February 1987 was aimed at stabilizing exchange rates among the Group of Six even though most forecasts pointed to a continuing U.S. current account deficit at those exchange rates. We can also take note of the “6 trillion yen” domestic expansion program adopted by Japan in 1987 based on the Mayekawa report and the recent budgetary measures in the United States.

It is evident that payments imbalances do not evoke the same degree of concern as in the past. The rest of the world seems quite prepared to live with what might be called a persistent U.S. current account deficit. Some Americans express concern because of the buildup of gross foreign debt that will burden the future. But many observers, official as well as academic, have come to the view that current account imbalances reflect decisions regarding saving and investment and do not require policy reactions. Moreover, such imbalances can be financed easily in today’s world of high capital mobility.

One of the most dramatic changes between “then” and “now” is the increase in the mobility of capital. As the result of the abandonment of capital controls and the development of telecommunications and computer technology, new financial instruments have appeared, information now spreads instantaneously throughout the world, and transactions can be carried out without delay. The consequence has been an enormous increase in gross international capital flows. For example, between 1973 and 1989, the foreign liabilities of banks in industrial countries increased elevenfold while their aggregate liabilities rose fivefold; annual foreign purchases and sales of U.S. Treasury bonds and notes rose from $4 billion to $4 trillion.

What can be said about attitudes toward exchange rates? In the 1970s, there was discussion about “clean” versus “dirty” floating but that has disappeared. Official intervention in foreign exchange markets is now regarded as normal but its limitations as a basic influence on exchange rates are recognized.

Exchange rate overshooting is a live concern and this has led to well-known proposals for target zones but no action in this direction has been taken. The attempt at exchange rate management begun at the Plaza by the Group of Seven and continued at the Louvre does not qualify, in my view, as a major reform.

The European Monetary System (EMS)—the most important reform of the exchange rate regime since 1973—seems to have evolved toward a fully fixed rate system. By this I do not mean to say that no further EMS realignments will occur but that most EMS members will try very hard to avoid such changes, given the credibility they have achieved and their aspirations to form a European monetary union. Whether the world will swing back toward something like the exchange rate regime of the 1950s and the 1960s is not an irrelevant question.

One other element in the present system is the procedure of macroeconomic policy coordination among the nations of the Group of Seven. This process is still in its infancy, but it strikes me as a valuable recognition of the increased interdependence that has occurred in the past quarter century. I regard it as a preoccupation that aims at more fundamental goals than the preoccupations of the 1960s.

III. International Liquidity—Then

By international liquidity was meant world reserves. For an individual country, international liquidity “means, most commonly, the command of its monetary authority over foreign exchange for use in intervening in the foreign-exchange market to support the exchange value of its currency. By intervening, the monetary authority can delay or avoid (1) the adoption of the domestic economic policies that would be required to adjust the economy so as to restore immediate payments balance at the current exchange rate, or (2) the adjustments that would be brought about by a change in the exchange rate.”4 Generally, liquidity consisted of gold and foreign exchange holdings. Among industrial countries, foreign exchange was mostly in the form of dollars.

In the system of almost fixed exchange rates of the 1950s and the 1960s, monetary authorities were sensitive about the level of their reserves and regarded it as desirable that reserves should grow as their economies grew and their international transactions increased over time. The Bretton Woods agreement made no provision, systematic or otherwise, for the growth of world reserves (except for the provision in the IMF Articles for a uniform change in par values—that is, a general increase in the price of gold—which was impractical if the dollar was to remain a reserve currency). As it turned out, the main sources of reserve growth were that portion of gold production that found its way to official reserves ($300 million a year in 1952–69) and increases in the official liabilities of the United States ($750 million a year in 1950–69). A U.S. official settlements deficit, like a deficit of any other country, increased the reserves of the rest of the world. But a U.S. deficit financed by increased dollar holdings of monetary authorities abroad increased other countries’ reserves without reducing those of the United States.

Thus the world relied on U.S. official settlements deficits for growth of reserves. This led to two types of dissatisfaction with the existing system. In Europe, Jacques Rueff, who became an economic adviser to General de Gaulle when the latter returned to head the French Government in 1958, expressed unhappiness with an asymmetrical system in which U.S. deficits did not, as in the case of other countries, lead to a loss of reserves and consequent pressure to adopt policies to reduce the deficits; President de Gaulle in turn spoke of the “exorbitant privilege” the United States had under existing international monetary arrangements. And French officials came forward with a proposed reform: a new “collective reserve unit” (CRU) linked to gold and designed to replace the dollar as a reserve asset as well as to meet the need for reserve growth.

In the United States, at Yale, Robert Triffin formulated his famous dilemma: if the U.S. deficits were eliminated, the world supply of reserves would not grow enough to satisfy the demand; but if this source of reserve growth continued, instability might ensue as U.S. liabilities increased relative to U.S. reserve assets.

In these circumstances, it was decided in the autumn of 1963 that the Group of Ten would undertake an “examination of the outlook for the functioning of the international monetary system and its probable future need for liquidity.” While U.S.-French differences were papered over in this study, it in turn led to the formation of a study group to examine proposals for the creation of reserve assets.

Over the next five years, treasury and central bank officials and IMF staff members busied themselves with this subject—in the Group of Ten, in combined meetings of the Group of Ten with the Executive Directors of the IMF, and, finally, in crisis meetings of the finance ministers and central bank governors of the Group of Ten in 1967 and again in March 1968. All this activity led, in 1969, to agreement on the creation of SDRs in the IMF.

This agreement was facilitated by the gold crisis that erupted after the devaluation of sterling in November 1967. The central banks of the major industrial countries had formed a gold pool in 1960 to try to stabilize the market price of gold. While this arrangement permitted the central banks to buy gold for a while, the devaluation of sterling led to market expectations of a possible dollar devaluation, which would mean an increase in the dollar price of gold. Thus, private purchases of gold increased sharply. In early 1968 the speculation shifted from an expected dollar devaluation to the prospect that the gold pool would be abandoned and the market price would rise. At a weekend meeting at the Federal Reserve Board in March 1968 the remaining gold pool members (France had dropped out) agreed to adopt a two-tier system: to stop selling gold in the market in an effort to stabilize the market price but to continue to transact gold with each other at the official price of $35 an ounce. This meant that there probably would be no additional flows of gold into official reserves. In a sense, gold was demonetized at the margin.

Also, from 1964 through 1968, world reserves in the form of gold and dollars had decreased by more than $4 billion as the result of official sales of gold into the market up to March 1968 and U.S. official settlements surpluses in 1966 and 1968 (and continuing in 1969).

Another rationale for the creation of SDRs was that additions to world reserves would serve to reconcile countries’ balance of payments aims. It was argued that if one added up the goals of countries concerning growth of their reserves, including the desire of the United Kingdom and France to reconstitute reserves that had been depleted during their balance of payments crises, the positive number that resulted was greater than any potential supply of new reserves. The implication was that unless the demand for reserves was satisfied, nations might pursue policies that were too restrictive. Thus a decision was taken in 1969 to create SDRs.

By the time the first SDR activation occurred at the beginning of 1970, short-term capital was flowing out of the United States back to the Eurodollar market and other countries’ dollar reserves were rising again. The system was headed for the dollar crisis of 1971.

IV. International Liquidity—Now

Very little attention is paid now, at least among industrial countries, to international liquidity. Developing countries, which have been oppressed by heavy debts and much-reduced capital inflows, have had more reason to be concerned with their reserve levels.

Gold reserves have barely altered over the past twenty years, except for the transfer by members of the EMS of some of their gold holdings to a central institution. While gold remains in reserves and is so reported, it has lost most of its monetary function. It has become a commodity and is part of the national patrimony like the Louvre or the Prado.

Foreign exchange reserves grew, on balance, in the 1980s after declining from 1980 through 1983. The increase from 1980 to 1989 was $339 billion, or about 90 percent. Of this amount, only about one third consisted of reported official claims on the United States. Another 31 percent represented claims on other countries. The share of dollars in total foreign exchange holdings fell from 78 percent in 1978 to 60 percent in 1989. But these changes in the amount and composition of foreign exchange reserves have aroused little interest in recent years. Discussion of a substitution account in the IMF, which would permit monetary authorities to turn in foreign exchange for SDRs, has virtually ceased. And one hears little about additional SDR allocations—the latest of which was in 1981.

Within the EMS, of course, reserves are of greater importance since intervention in foreign exchange markets is obligatory.

If one were to identify the world’s principal economic and financial problems, international liquidity would be far down on the list. That is why this section of my paper is so short. There is little to say on the subject in regard to the present system. One can imagine reforms that would enhance the function of and need for reserves, including the SDR, but that is beyond the scope of this paper.

V. Concluding Comment

The key factor explaining the contrasts between “then” and “now” with respect to both adjustment and liquidity is the enormous increase in the international mobility of capital. It helps to account for the larger current account imbalances now and the relative indifference to them, since they are mostly financed privately. It helps to explain the volatility of exchange rates. It helps to explain the lack of concern about the level and composition of reserves. And, finally, it accounts for the increased interest in international policy coordination.


Ministerial Statement of the Group of Ten and Annex Prepared by Deputies. August 1964.


International Monetary Arrangements: The Problem of Choice—Report on the Deliberations of an International Study Group of 32Economists (Princeton, New Jersey: Princeton University Press, 1964), p. 102.


Approaches to Greater Flexibility of Exchange Rates: The Bürgenstock Papers, edited by George N. Halm (Princeton, New Jersey: Princeton University Press, 1970).


International Monetary Arrangements: The Problem of Choice, op. cit., p. 29.

    Other Resources Citing This Publication