3 Decline of Multiple Rates, 1947–67

Margaret De Vries
Published Date:
June 1986
  • ShareShare
Show Summary Details

One of the most immediate questions with which the newly created International Monetary Fund was confronted in 1946 was that of multiple exchange rates, used here as another term for multiple currency practices. Of the 40 countries that were members at the end of 1946, 13—all in Latin America except one—were using such rates.1 Although many members maintained exchange controls, to which the Fund was also opposed in principle, and had inconvertible currencies at the same time the Fund was emphasizing the achievement of widespread convertibility, the problem of multiple exchange rates was thought to require particular attention. Multiple rates were feared partly because of the competitive ways in which they had been used during the 1930s. More important, because multiple rates are not merely restrictive practices but also rates of exchange, they were viewed as a major source of exchange rate instability. Exchange rate stabilization and the attainment of fixed rates of exchange were points often stressed in the Fund’s deliberations about multiple rates in its first 20 or so years. Achievement of fixed rates was a primary objective, to be attained as quickly as possible. Indeed, it was in order to move toward this objective that the Fund decided as early as 1946 to ask members to set initial par values. Fund officials argued that while the mere existence of multiple exchange rates might not endanger exchange stability, the frequent changes of rates inherent in the use of multiple rates could well undermine the whole par value system.


For these reasons, among the key decisions of the Fund in its first two years was that multiple exchange rates must be subjected, without delay, to special review. In December 1947 the Fund circulated to its members a general pronouncement which came to be known as “the December 1947 Letter on Multiple Currency Practices.” This letter stated explicitly that members should, as a minimum, consult the Fund before introducing a multiple currency practice, before making a change in any of their existing multiple rates of exchange, before reclassifying transactions subject to different rates, and before making any other significant change in their exchange systems. This statement of general principles on multiple currency practices, moreover, emphasized that early steps should be taken toward the removal of these practices, especially those that were clearly not necessary for the balance of payments position of a member, although problems of timing and practical alternatives had to be recognized.2

In the next several years, the Fund was flooded with specific instances of multiple rates for consideration and action. The Executive Directors and the management and staff began to work closely with the members concerned to determine what progress might be made not only toward reducing but also toward the early removal of multiple rates.

However, as circumstances in individual member countries were revealed in detail, an early termination of multiple exchange rates seemed remote. It was quickly apparent that much of the use of multiple rates was caused by the existence of domestic inflation: multiple rates usually resulted from balance of payments deficits which, in turn, were most often caused by inflation. Until inflation was brought under control, the removal of multiple rates would necessitate the introduction or tightening of quantitative restrictions on imports and payments as a substitute. Inasmuch as quantitative restrictions are also a form of control and give rise to many difficulties on their own account, the replacement of multiple exchange rates with such restrictions would have achieved little.

It also became evident that members were making widespread use of multiple rates for purposes other than balance of payments, such as to tax or subsidize particular categories of exports and imports or generally to raise revenue. In fact, many economists began to regard multiple rates as performing useful economic functions that could not be accomplished by quantitative restrictions. These additional uses of multiple rates meant that, in practice, the removal of multiple rates would, in many instances, require the introduction of some acceptable substitute measures to perform these tax and revenue-raising functions. Were the Fund to insist on the immediate discontinuation of multiple rates, members might be compelled to put into effect hastily devised tax substitutes, different in form but not better in practice than the existing systems of multiple rates. Much better substitutes could be devised, but these would require overall reforms of the fiscal systems and customs tariff schedules of the members concerned. Such reforms first had to be carefully designed and then had to go through prolonged legislative procedures, both of which were time-consuming.

In these circumstances, the general principles outlined in the 1947 Letter could be put into effect only as permitted by the situations prevailing in the members concerned and would have to be adapted to the specific problems at hand. Accordingly, the Fund gradually evolved “a case-by-case approach” that is, the situation of each member was decided on its own merits. In implementing this approach, the Fund, rather than insisting on a hurried elimination of multiple rates, offered guidance to its members by putting emphasis on the achievement of the basic underlying economic conditions which would facilitate elimination of multiple rates as soon as practicable. The most important of these conditions was domestic financial stability. Members were also encouraged to avoid the use of many particular practices—such as auction systems, compensation arrangements, or “mixed rates”—which unduly complicated their exchange rates or were most likely to harm other members. Furthermore, to keep exchange rates as stable as possible, the Fund urged that free market rates be limited to items other than trade, such as capital or invisibles. Avoidance of these particular practices and limitations on free markets would help eliminate the most adverse forms of multiple currency practices and also help members exercise reasonable control over those retained. Within this framework, the Fund was able to agree to many multiple currency practices on a temporary basis, remaining in consultation with the members concerned.


In many respects, the 1947–55 period was a frustrating time for the Fund and for its members. On the one hand, many members with multiple currency practices, especially in Latin America, argued that the Fund, in a world of extensive exchange controls and continued inconvertibility of key currencies, was devoting excessive attention to the multiple rates of a few members that had only a small share of total world trade. On the other hand, some economists and public officials, eager for evidence of success by the new organization in attaining at least some of its objectives, frequently lamented the lack of progress, even in what seemed to be the relatively small area of multiple rates.

Definitely there was reason to lament: multiple currency practices actually increased among the Fund’s membership. Several countries that already had multiple exchange rates joined the Fund: Afghanistan, Argentina, Indonesia, Israel, Korea, the Syrian Arab Republic, Thailand, and Viet Nam. Many existing members found it necessary to introduce such practices: China, Egypt, Greece, Finland, Iceland, the Philippines, Turkey, and Yugoslavia. Thus by the early 1950s multiple rates were no longer confined to Latin America; they had spread to Asia and the Middle East and, indeed, to Western Europe. Even France, a major industrial member, undertook an abortive experiment with a free market for certain currencies in 1948, and during the early 1950s a number of other Western European members, while retaining unitary fixed rates for the bulk of their transactions, felt compelled to open free markets for the sale of certain currencies earned in bilateral arrangements or for the currencies of all countries in the European Payments Union, or for particular capital transactions. By the end of 1955, the Fund had grown to 58 members, 36 of which had some kind of exchange device that the Fund considered a multiple currency practice. Although some of these practices were minor, very complex systems with many different exchange rates had also gradually evolved in a number of members.

The increased use of multiple currency practices during these years reflected two distinct kinds of problems, each of which had given rise to different types of practices. One problem was what was commonly called at the time “the chronic world dollar shortage.” As a result of a widespread shortage of U.S. and Canadian dollars, the currencies of the world had become divided into “hard” currencies, such as the U.S. and Canadian dollars, which were freely convertible into other currencies and, through the U.S. dollar, into gold at a fixed price of $35 an ounce, and “soft” currencies, such as the pound sterling, the French and Belgian francs, the deutsche mark, and the Netherlands guilder, which were not freely convertible. While a country might have achieved overall equilibrium in its balance of payments, any surpluses for which it was paid in inconvertible currencies could not be used to settle deficits it might have with other countries where payment in convertible currencies was required. This situation had led to the emergence, in Western Europe as elsewhere, of some multiple currency practices—such as limited free markets and broken cross rates—which aimed at increasing a country’s supply of “hard” currencies and decreasing its holdings of “soft” ones.3

The second problem was confined largely to developing members. Whereas a few Latin American members had initially used multiple exchange rates to alleviate the depression-based balance of payments problems of the 1930s, these practices now seemed useful techniques for coping with the balance of payments problems generated by measures designed to stimulate economic development. Thus, persistent balance of payments deficits associated with long-term development expenditures, the need for protection of new domestic industries, the desire to differentiate various categories of exports and imports and to tax or subsidize various groups of commodities, and the scarcity of adequate revenue sources had, in the absence of appropriate alternative measures, given rise to widespread use of multiple rates by developing members in all geographic regions. Selective exchange rate devaluation for particular exports and imports often seemed preferable to general devaluation, which would be inflationary, or would raise the cost of essential imports, or would bring excessive profits to exporters of some primary products, and might still not help significantly to reduce the balance of payments deficit.


After 1955, however, a sudden and almost dramatic breakthrough occurred in the Fund’s efforts to reduce multiple exchange rates. In its Seventh Annual Report on Exchange Restrictions, 1956, the Fund was able to report great progress within a single year. A marked general improvement in the international payments situation in the mid-1950s was the main cause of the sudden trend toward reducing the use of multiple rates. The intensive investment programs, especially in Western Europe, after World War II had borne fruit, and large increases in production were coming forth. Increased supplies of goods, together with the improved domestic financial and monetary positions of Western European members and the realigned exchange rates of 1949, brought the world dollar shortage to an end. As a result, within the next few years, the major industrial members greatly reduced, or even abolished, many restrictions on their trade and payments and established convertibility for their currencies. The multiple currency practices which distinguished between “hard” and “soft” currencies could be, and were, quickly removed. By 1958 the partial free markets and other special devices of a number of Western European members had all been discontinued.

What was in some ways even more unexpected, although the Fund had been working closely with these members over several years, was that at the end of 1955 and during 1956 several developing members that had long maintained complex multiple rate systems (Bolivia, Chile, Paraguay, and Thailand) suddenly abolished them in favor of a single fluctuating exchange rate.

The time was ripe for an intensification of the Fund’s efforts to eliminate multiple rates. The members still employing multiple rates—members exporting mainly primary products—were more inclined to defend their use by emphasizing the practical difficulties of eliminating them than by claiming any special economic rationale for them as against unitary exchange rates. The Fund intensified its efforts to eliminate multiple rates in the mid-1950s also because the potentially adverse effects of multiple exchange rates on other Fund members had become more worrisome. Competition in world markets had become keener and members that had unitary exchange rates were becoming increasingly concerned that competitors using multiple export rates might derive unfair advantage in export markets.

The experiences of the Fund in its first decade had shown that it was much easier for a member to simplify its rate structure than to unify entirely. The stringency in internal monetary policies needed for simplification was significantly less than that needed for unification. Moreover, the foreign exchange reserves required were lower if a simplified system included a fluctuating rate. And in a simple system with fixed rates, an exchange spread might absorb part of any internal monetary expansion, while such absorption was not possible with a unitary rate. In addition, Fund experience had shown that it was the complexity of multiple rates that had led to their most damaging effects.

For these reasons, in June 1957 the Fund worked out a new policy on multiple currency practices. It urged members to simplify their existing rate structures and offered to help them work out specific exchange systems. Simplification meant more than merely a reduction in the number of existing rates: emphasis was placed on a maximum of two or three rates sufficiently realistic to maintain a satisfactory balance of payments position with only a minimum of quantitative restrictions. A simple system could include, where necessary, the temporary use of a widespread free market or of a fluctuating rate as a step toward an eventual stable rate; over the years it had become evident that a floating-rate mechanism had a greater tendency to be self-liquidating than a series of fixed rates. Technical assistance could also include Fund help with the formulation, where appropriate, of substitute fiscal arrangements. Most important of all, where the exchange reform was sufficiently broad and accompanied by a domestic stabilization program, the Fund was willing to consider use of its resources to support stabilization efforts. Meanwhile, it would be reluctant to approve changes in multiple rate systems which would make them more complex.4

In the next several years following this new announcement, other members, encouraged by the success of those that had already unified their exchange systems, eliminated their multiple rates. Significantly enough, in 1959 the members involved included Argentina, which in the 1930s had been the founder of multiple rates in Latin America. Many other members—Costa Rica, the Islamic Republic of Iran, Jordan, Nicaragua, and Saudi Arabia—began to eliminate their multiple rates in a series of steps. In 1962 this group was joined by Yugoslavia, which in the mid-1950s had the most complex system of all, a system involving nearly a hundred effective rates of exchange. Some members—Finland, Israel, Spain, and Turkey—took the whole plunge at once, devaluing, eliminating multiple rates, and setting a new fixed rate.

These major reforms of exchange systems usually involved both exchange devaluation and internal policies designed to remove price distortions and to help bring inflationary pressure under better control. Frequently changes in customs duties and tariffs, usually upward adjustments, also accompanied exchange reform, and in some instances, broad fiscal reforms were undertaken. Some members temporarily applied taxes or surcharges to a few categories of imports, such as luxury consumption items, while using a fluctuating unitary rate for the bulk of their transactions. Thus, from 1955 to 1962, 8 Western European members eliminated their various minor practices, and 18 other members unified their systems of multiple rates. Hence, by the end of 1962, only 15 Fund members out of a total of 82 had multiple rates, in contrast to 36 out of 58 eight years earlier.


Three types of developments in multiple exchange rates in the five years from 1963 to 1967 occurred. First, additional members that had long made use of multiple rates continued to join the list of those who had considerably simplified or unified their exchange rates. For example, in 1964–65 both Venezuela and Brazil substantially reduced their rate multiplicity. The Philippines, which introduced multiple rates in 1962, eliminated them in November 1965. In 1965 Korea replaced its series of rates by a unitary fluctuating rate. Second, some members, such as Afghanistan, Colombia, and Indonesia, continued to maintain complex exchange systems and, as of 1967, had not yet simplified them. Third, backsliding occurred. For example, following the virtual unification of its exchange rate system in 1956, Chile, in 1962, introduced a dual market which still continued in 1967. Some multiple rates even sprang up in a few members—for example, Sri Lanka (Ceylon) and Zaïre (the Democratic Republic of Congo)—where they had not been used before. If not multiple rates, then some other devices, such as import surcharges, advance deposit requirements, or special export bonuses, were also introduced by several members to deal with recurrent balance of payments or fiscal deficits. Some members that introduced these devices, such as Argentina, had eliminated multiple rates, while others, such as India and Pakistan, had not previously had them. Moreover, among the measures introduced by a number of industrial members to reduce the outflow of capital were special taxes or exchange markets with premiums for investment currencies. The latter were introduced, for example, in the United Kingdom.

In addition, from time to time economists renewed arguments for multiple rates as ways of solving a particular trade problem, especially for developing members. Some economists suggested—even for industrial members—that subsidies with tariffs (similar to multiple rates) could be useful for many purposes along with either a fixed or a fluctuating exchange rate.

Consequently, by the mid-1960s Fund officials realized that the issue of multiple exchange rates was not entirely at an end and that it might well never be. The nature of, and reasons for, multiple currency practices differed, and the locale shifted. Nonetheless, despite some continued problems and other emerging ones, and the need for constant work by the Fund with all its members, the wide use and the complex nature of multiple rates that had prevailed at the end of 1947 had been vastly diminished 20 years later. One of the important objectives of the Fund, to eliminate multiple currency practices, had, both because of the Fund’s flexible policies and because of the cooperation of its members, come at least as near to attainment as could have been hoped when the Fund was created.

Note: This article, written in December 1967 on the occasion of the twentieth anniversary of the Fund’s first general pronouncement on multiple exchange rates of December 1947, was published earlier in a slightly different form in Finance & Development (Washington), Vol. 4 (December 1967), pp. 297–303. A longer version can be found in Margaret G. de Vries, “Multiple Exchange Rates: Expectations and Experiences,” Staff Papers, International Monetary Fund (Washington), Vol. 12 (July 1965), pp. 282–313.

The Latin American members were Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Ecuador, Honduras, Nicaragua, Paraguay, Uruguay, and Venezuela; the other member using multiple rates was the Islamic Republic of Iran.

This letter can be found in History, 1945–65, Vol. III, pp. 261–65.

As described in the article on pp. 40–48, broken cross rates involved the sale of some currencies at discounts from their parity relations with the U.S. dollar.

The decision setting forth this policy is in History, 1945–65, Vol. III, pp. 265–66.

    Other Resources Citing This Publication