ONE FUNCTION of exchange rates is to assist in bringing a country’s international payments into a tenable balance, one in which receipts on current and ordinary capital account are adequate to meet payments on current and ordinary capital account. With satisfactory exchange rates, a country ought not to find it necessary, when the world economic environment is reasonably favorable, to impose restrictions, to deplete its monetary reserves, or to seek extraordinary financing to meet a balance of payments deficit.


ONE FUNCTION of exchange rates is to assist in bringing a country’s international payments into a tenable balance, one in which receipts on current and ordinary capital account are adequate to meet payments on current and ordinary capital account. With satisfactory exchange rates, a country ought not to find it necessary, when the world economic environment is reasonably favorable, to impose restrictions, to deplete its monetary reserves, or to seek extraordinary financing to meet a balance of payments deficit.

ONE FUNCTION of exchange rates is to assist in bringing a country’s international payments into a tenable balance, one in which receipts on current and ordinary capital account are adequate to meet payments on current and ordinary capital account. With satisfactory exchange rates, a country ought not to find it necessary, when the world economic environment is reasonably favorable, to impose restrictions, to deplete its monetary reserves, or to seek extraordinary financing to meet a balance of payments deficit.

This function is performed in two ways. On the one hand, by bringing foreign prices and domestic costs into a relationship which is satisfactory for exports, the exchange rate should encourage a supply of exports sufficient to provide the foreign exchange needed for imports and other payments. On the other hand, by bringing foreign prices for its imports and domestic prices for home output into a satisfactory relationship, the exchange rate should limit the demand for imports to the amount that can be paid for with current exchange receipts and normal capital inflow. This does not imply, of course, that the exchange rate is the only factor that determines the flow of foreign exchange receipts and payments.

The effects which exchange rates have on exports and on imports might be examined independently of each other. Before doing this, however, two other general points should be emphasized:

First, in any system free from exchange controls, the export rate and the import rate (i.e., the buying rate and the selling rate) will be substantially the same.

Second, any country for whose exports the world demand is infinitely elastic, and for whose imports the world supply is infinitely elastic, will secure the maximum advantage from its international trade by having substantially the same rate for exports and imports.

Where recipients of exchange are free to dispose of the exchange, they will sell it at a rate uniform for all sellers and substantially equal to that paid by importers. Buyers of exchange have no reason to distinguish between the different sources from which exchange funds flow. They will pay at the same rate for dollars derived from capital inflow, from copper exports, or from rice exports. Nor will there be any difference between the rates paid by different buyers—whether for imports, private remittances, or business remittances. The maximum difference between buying and selling rates will be just sufficient to induce specialists (bankers or exchange dealers) to provide a ready market for exchange. In a well-organized free exchange market, the maximum difference for large transactions may be as small as 18 per cent. A substantial difference between buying and selling rates is possible only when the government or the financial authorities limit the freedom to deal in exchange and prescribe the rates at which exchange is to be bought and sold.

A simple economic analysis will show that for a country whose supply of exports and demand for imports have only an insignificant effect on their prices in foreign exchange, the maximum advantage from international trade will be ensured where the same rate of exchange is applied to both exports and imports.1 Every additional dollar’s worth of exports makes possible an additional dollar’s worth of imports, and as long as importers are prepared to buy exchange from exporters, with the entire payment going to exporters, this means a net increase in the advantages from international trade. Exports and, therefore, imports will expand until the point is reached at which the money cost of producing a unit of exports ceases to be less than the money value of a unit of imports. Up to that point, additional exports make possible additional imports at a local value higher than the local costs of producing the corresponding exports.

Import Rate in Excess of Export Rate

The effectiveness of an exchange rate in stimulating exports or restricting imports depends upon the elasticity of supply for exports and the elasticity of demand for imports. Where both supply and demand are elastic, there can be no question of the effectiveness of the exchange rate in adjusting international payments, with a minimum disturbance to local prices of export and import goods. A small change in the exchange rate would then bring about a big reduction in the demand for imports and a large expansion in the supply of exports (at only slightly higher local prices for these goods). Under such conditions, a moderate adjustment in the exchange rate can be very effective in balancing payments.

On the other hand, where the demand for imports is inelastic, only a large change in the exchange rate will reduce the demand to any significant extent. And where the supply of exports is inelastic, a change in the exchange rate can have only a small effect in expanding exports. Where these demand and supply conditions prevail, a change in the exchange rate can restore balance in the international payments of a country with a payments deficit only through a large change in local prices of exports and imports. This raises the question whether in these circumstances the exchange rate is the best instrument for restricting imports, and whether different export and import rates might not be preferable to a single exchange rate.

Where the demand for imports is inelastic, there is a great temptation to seek an adjustment in the payments position through direct restriction of imports. This seems especially attractive in a time of high prices, when the public may be extremely sensitive to measures that involve further price increases. Unfortunately, however, direct restriction of imports does not avoid a rise in price to consumers unless it is accompanied by price control. It merely keeps down importers’ costs while domestic prices rise just as they would have done if the same degree of restriction had been accomplished through a change in the exchange rate. The beneficiaries of direct restriction of imports are the fortunate recipients of import licenses and of official exchange, whose profits will rise enormously.

Where the supply of export goods is inelastic, however, a change in the rate of exchange will cause large increases in the local currency prices of exports and in the profits of exporters. It will not, however, increase substantially the volume of exports or the foreign exchange proceeds from exports. At the same time, the expansion of exporters’ incomes may have a disturbing effect on the stability of the domestic economy. For this reason, too, objections have been raised to the adjustment of international payments by changes in exchange rates when the supply of export goods is inelastic.

Under the conditions described, the case for having one exchange rate to stimulate exports and a different rate (higher in local currency units per dollar) to limit imports can be most favorably presented. In brief, the argument is that direct control produces windfall profits for importers, while exchange depreciation produces windfall profits for exporters. If a wider spread were permitted between export and import rates, it would allow the price system to bring the country’s international payments into balance; and the general public (the government) would be the beneficiary from the rate differential in the form of an exchange tax.

The inelasticity of domestic demand for imports and of domestic supply of exports does not invalidate the conclusion that the maximum advantage from a country’s international trade can be secured through equality between the export and import exchange rates. Rather, it raises the question whether a shift in the distribution of home incomes obtained through an appropriate spread between export and import exchange rates may not yield advantages at home sufficient to counter-balance the advantages lost through a slight restriction of international trade. The argument is similar to that implied in Marshall’s analysis of the effect of a tax on an industry producing at increasing costs. His conclusion was that such a tax falling on producers’ surplus could be to the general advantage.2

The establishment of an exchange rate for imports different from (and exceeding in local currency units per dollar) the export rate is equivalent to a tax on exports or on imports or on both. There is indeed no real distinction between a general tax on exports (assuming the whole of the output of export goods to be sold abroad) and a general tax on imports (assuming the whole of the consumption of these goods to be supplied from abroad). The burden of either tax falls upon both exports and imports. The division of the burden between exporters and importers depends on the elasticity of demand for imports and the elasticity of supply of exports. Since a difference between buying and selling exchange rates is equivalent to a tax, it follows that if the supply of exports is inelastic, this difference shifts income from export producers and importers to the general public. Given a sufficient degree of inelasticity of supply of exports, a country may find that it pays to permit the sacrifice involved in a slight reduction in the volume of international trade so that larger benefits may be secured from a redistribution of home income. That, in turn, will depend upon whether the incomes of export producers and importers are a good source of taxation. And it also depends inversely upon the capacity of the government to capture through other taxes a considerable part of the increase in export profits that would result from a unified exchange rate.

The spread between export and import rates is a tax on a country’s international trade at the same ad valorem rate for all goods. Since the supply of exports is assumed to be inelastic, the generally protective effect of this tax (i.e., in increasing the ratio of home goods to imports) is not large. Its specifically protective effect for any one or more import goods may, of course, be quite large, depending on the capacity of the country to substitute home goods for any one or more import goods.

Does an abnormal spread of this kind between buying and selling rates have detrimental effects on other countries? In the conditions set out above, it does not. If the world supply of exports to the country which adopts such a policy is infinitely elastic, the effect on the export prices of other countries is zero. If the world demand for imports from this country is nearly infinitely elastic, the effect on the import prices of other countries is also zero. And if the home supply of export goods in this country is inelastic, the effect on the volume of trade is small. Under the conditions set out above, no harm can be done to other countries by an abnormal spread in any one country between buying and selling rates for exchange.

The case is different where each of a large group of countries producing raw materials restricts the supply of exports by an abnormal spread. While each country may keep its exports down in this way by only 5 or 10 per cent, the adoption of the same policy in many countries would considerably reduce the aggregate volume of exports and raise the prices of such raw materials. For though world demand may be nearly infinitely elastic for the small part of the total supply coming from any one country, it may at the same time be quite inelastic for the aggregate supply. However, even for many countries to maintain a large spread between their buying and selling rates would not necessarily be objectionable. It would be so only if the prices of the export goods affected were raised to “unreasonable” levels.

Is the assumption of an inelastic supply of exports justified? The production of some raw materials may require a long period of gestation before new investment materially affects output. The short period supply of such exports is likely to be inelastic. But this does not necessarily mean that the elasticity of supply for exports as a group is inelastic. Few countries depend entirely or nearly entirely on a single export. It is unlikely that even in the short run the supply of all of a country’s exports, actual and potential, will be inelastic, and even less likely that over a long period the supply of its exports as a group will be inelastic.

Differential Import Rates

Multiple import rates arise when imports are classified in different categories (lists) and a different exchange rate is applied to each list. Such a system of differential import rates may be the result of a careful policy of encouraging some imports and discouraging others as part of a thoroughgoing program of economic development or of redistribution of real income. It is, however, more likely to be an expedient for meeting an urgent need to restrict international payments; or it may be adopted as an easy means of raising supplementary revenue from taxes on some imports.

When a country finds that its aggregate demand for imports exceeds the foreign exchange receipts available to pay for them, some method of restricting imports is necessary. The simplest way of doing this through the price system is by raising the exchange rate, or at least the import rate. This would involve a rise in the local currency prices of imports and would induce some reduction in the demand for them. Since any excessive demand for imports is likely to have arisen (at least under present conditions) from inflationary forces generated at home, governments will often be reluctant to take this step because of its effects on the cost of living. Multiple import rates, in which so-called essential imports are allowed exchange at a favored rate, appear to be a promising way of restricting imports while keeping down the cost of living.

The typical classification of imports for multiple rates includes one group comprising the “most essential” consumer goods (and possibly machinery and equipment as well), to which a favored rate may be assigned close to or even below parity. A second group comprising “useful” goods may have a rate perhaps one third higher than parity, and a third group comprising “luxury” goods a rate perhaps two thirds higher than parity. The purpose of such a pattern of import rates is to concentrate the reduction of imports in the second and third categories, and thus to reduce aggregate payments while keeping down the local currency price of imports in the favored category.

The tendency is strong in countries using multiple import rates to make the first category, with its favored import rate, much too large. It seldom comprises less than 60 per cent of aggregate potential imports, and in one country, Colombia, it exceeds 90 per cent. While the purpose is to keep down the cost of living, the effect is to increase the difficulty of dealing with the payments problem. It is often impossible to restrict aggregate import demand sufficiently if the higher import rates are confined to a small part of total potential imports. Even when aggregate import demand can be sufficiently restricted by reducing imports in the second and third categories, this requires such high rates for foreign exchange as to undermine faith in the tenability of the existing parity or official rate.

There is one type of favored import rate that may encourage a form of commodity arbitrage to the detriment of the country using multiple import rates. Where the favored import rate is less than the export rate, importers may find it to their advantage to import essential goods at this rate and to re-export them at the export rate. If the difference between the two rates is substantial, the profits from such transactions may encourage a very considerable trade in such re-exports, either unprocessed or slightly processed. This is apparently what happened in Peru with various imported foods. At best, a highly favored import rate of this kind makes it necessary to supervise exports closely to prevent such practices. At worst, it costs a country considerable sums in local currency and induces a shift in the composition of its foreign exchange receipts and payments in the interests of an utterly useless form of trade.

Even if commodity arbitrage of this kind can be avoided, it is difficult to see how any useful function can be performed by an import rate which is less than the going export rate. The country is then using home resources to produce export goods at a cost which exceeds the value of the imports they pay for at the favored rate of exchange. Moreover, the pattern of import rates affects the structure of consumption. Under the assumption that the average import rate is adequate to restrain aggregate imports (which is not always the case), the favored rate is equivalent to a subsidy on the favored imports. There may be a case for subsidizing some forms of consumption; when a country has a payments problem, however, the case for subsidizing the consumption of imports is extremely weak.

Perhaps even more important, the pattern of import rates affects the structure of home production. In Latin American countries, the essential imports in the favored category are often the very goods that the country itself is best able to produce. If imports are allowed to come in at such a rate, the home production of these goods may be discouraged. For example, in Peru, before multiple rates were established, all the meat needed for home consumption could be produced domestically. But when meat was imported at the official rate (6.5 soles per dollar), while less essential imports came in at the certificate rate (say about 16 soles per dollar), the favored import rate for meat kept the price abnormally low and home production to less than half of total consumption.

In Ecuador, there is a similar situation with respect to wheat flour. At a time when Ecuador is hard pressed to meet its foreign payments, imports of wheat flour account for somewhat less than 10 per cent of total imports. Wheat flour is imported under List A at a favored rate of 15 sucres per dollar; if it were included in List B, at 20 sucres per dollar, home production could supply a much larger part of the home demand. The existence of favored import rates has diverted home production in Peru and Ecuador away from industries which could compete with other countries at exchange rates appropriate to balance their international payments.

The harm done by multiple import rates is primarily to the countries that use them; their production is distorted and channeled into less productive industries. But multiple import rates may also do harm to other countries, in particular to countries that export the goods arbitrarily placed in the categories to which penalty import rates are applied. The inclusion of food, raw materials, and equipment in the essential list means that Latin American countries give a subsidy for imports from the United States and other dollar countries. The classification of other consumer goods in the nonessential list means that Latin American countries penalize imports from the European countries which are already heavily burdened with a dollar payments problem.3

Objectives cannot, of course, be condemned on account of the methods whereby they are pursued. Countries with a payments problem may well feel that penalties have to be imposed on imports. But the penalties should, in any event, be imposed in the form which will be most helpful to the country. They might better be imposed on imports for which domestic substitutes would be forthcoming at very close to an appropriate exchange rate than on so-called nonessential goods. When, moreover, balance of payments difficulties are created by inflationary overinvestment, the difficulties will be further increased by offering exchange at preferred rates for imports of machinery and equipment.

Multiple import rates are equivalent to tariffs on all goods included in the categories to which penalty rates are applied. The specific protection afforded may be more or less effective as domestic production is or is not able to provide similar goods. Under any circumstances, they provide a form of general protection by discouraging imports and by raising the proportion of home output in domestic consumption and investment. In part, of course, this general protection is to some extent offset by the subsidies on favored imports. But under the circumstances in which multiple import rates are applied, the protection to home production is haphazard, unrelated to the desirability of encouraging production of various types of goods, and unadjusted to the degree of protection that might call forth the greatest effort in producing such goods efficiently at home.

Differential Export Bates

The use of multiple rates is much less common for exports than for imports.4 Such rates fall into one of two classes: (a) penalty export rates which aim at capturing part of any exceptional profits from exports; (b) favored export rates to encourage marginal exports that would not otherwise be produced and sold.

The taxation of exports as a means of raising revenue is a long-established and widely used practice. More recently, the use of differential exchange rates for this purpose has been increasing. In Iran, for example, the Anglo-Iranian Oil Company returns such of its export proceeds as it requires to meet local expenses. In selling such exchange, it receives the official rate of 32 rials to the dollar, while the usual rate for exports is the certificate rate, now approximately 40 rials to the dollar. In Chile, the largest mining companies return such of their export proceeds as they require to meet local expenses at a rate of 19.37 pesos to the dollar, while the usual rate for other exports is now close to 60 pesos to the dollar. These penalty export rates are used because the profits of the companies are high and, unless contractual arrangements can be revised or higher taxes imposed, government revenue would be diminished if such exports were placed on the same basis as other exports.

If the proposition is accepted that the aggregate revenues received by the government from such enterprises (by way of taxes and exchange profits) are reasonable, both the government and the companies concerned would find it advantageous to have transactions conducted at the going rate of exchange, and to make other arrangements for adjusting taxes to the appropriate level. As long as such enterprises as the mining companies in Chile and the oil company in Iran conduct their business at an inappropriate rate of exchange, there is an inducement to distort their use of home and foreign resources to the detriment of both the country and the companies.

When the mining companies in Chile have the alternative of buying goods in Chile or buying them abroad, there is a strong incentive to buy them abroad whenever they can be bought at a cheaper rate than their local peso equivalent of 19.37 pesos to the dollar. This would mean a considerable loss to the Chilean economy which would have to forego the opportunity to “export” to the copper companies for payment in dollars, because the rate at which such exports must be paid is substantially below the going rate in Chile. A case is cited of the export of locally produced ferro-manganese at a time when the copper companies were importing ferro-manganese.

Penalty export rates have also been frequently proposed, if not in fact frequently used, to meet a situation of a different kind. Exceptionally high prices for exports sometimes give producers very large temporary profits. It has been suggested from time to time that the exchange rate for converting such export proceeds should be less favorable so that part of the exceptional profits may in effect be taxed away. This suggestion is commonly made when depreciation becomes necessary to move some exports, while the competitive position of others remains unimpaired. In such cases it is generally proposed that the penalty export rate should be temporary, to be removed when the market becomes adverse.

It is difficult to insist that taxes should not be imposed on some exports when the burden of taxation can be borne without impairing production. This is the key to the problem. Where there is a payments problem, it is clearly undesirable to restrain the production of additional goods for export. The desirability of a penalty export rate, therefore, depends on the implicit assumption that the supply of exports in the penalty class is quite inelastic. While this may be so, it should not be assumed too easily, and particularly if the inelasticity of supply is a short-period phenomenon. To encourage long-period expansion of output, the going exchange rate should be allowed to have its full impact on supply. Whatever taxes can properly be collected from such export producers should be levied on some basis, such as profits, that would have a minimum effect on the expansion of output.

A common type of multiple export rate provides a favored rate to encourage marginal exports that would not otherwise be produced and sold. The favored export rate in Uruguay (2.35 pesos to the dollar), for example, is considerably higher than the basic export rate (1.52 pesos to the dollar) and the basic import rate (1.90 pesos to the dollar). In effect, such a rate is a subsidy to encourage new export industries or to maintain industries which are in an unfavorable situation. It may be defended as part of a program of economic development (superior perhaps to protective tariffs for infant industries). It is even possible to argue that the apparently favored position of these marginal exports is to a large extent an illusion. If a unitary exchange rate were to be established, these marginal exports would still be needed because of the inelasticity of supply of basic exports.

In Venezuela, exports other than petroleum products are given a favored rate because the high level of wages in the petroleum industry raises the whole level of costs and other industries could not export at the same exchange rate as petroleum. In brief, this amounts to a plea for maintaining less efficient export industries as a supplement to the basic export industry. The justification may be that it is socially desirable to encourage production and employment in agriculture instead of having a one-sided economy. Further, it is always possible that the position of the petroleum industry may be affected by the opening of new fields and the development of synthetics or substitutes. While a sharp adjustment in costs, if necessary by way of depreciation, might then maintain to a greater or less extent the export position of the basic industry, the change in the terms of trade and the loss of exchange receipts would make necessary the development of other industries. It may be desirable, for this reason, to encourage supplementary export industries now so that these risks for the future may be diminished.

From the point of view of other countries, the principal objection to multiple export rates is that they are a form of unfair competition. It is too easy to adjust a multiple export rate every time the competitive position of one or more of a country’s exports deteriorates. Clearly, a country must assure its capacity to provide enough exports to pay for its imports; but when a payments problem has to be solved, application of a favored export rate to only a few special commodities is unfair to other countries producing the same or similar commodities. When a change in the exchange rate is made in this way, its impact is concentrated on a few competing countries. On the other hand, when the exchange rate is depreciated for all exports, the impact is shared by other countries as broadly as the diversity of the economy of the depreciating country permits.

The use of a favored export rate is prima facie prejudicial to the development of other export industries in which the economic advantages of production are greater. It should, therefore, not be too readily assumed that the development of marginal exports by this means ensures the best use of a country’s resources. Moreover, even if it is desirable to encourage the marginal exports, it would be better to do so through direct subsidies from the budget. In this way a more careful weighing of the advantages to be derived from marginal exports would be assured; and a more accurate adjustment of the subsidy to the economic advantages to be gained by maintaining the marginal export industries would be possible.

Multiple export rates do not have a generally protective effect, provided that the expansion of marginal exports attributable to the favored rate is greater than the contraction of basic exports because of the penalty export rate. That is another way of saying that subsidies encourage an expansion of trade while tariffs encourage a contraction of trade. Nevertheless, multiple export rates may be specifically protective for some of the favored industries. Thus, if raw wool is exported at a lower rate and woolen yarn is imported at a higher rate, in terms of domestic currency per unit of foreign exchange, the domestic producer of woolen yarn receives the equivalent of tariff protection equal to the difference between the two rates. Furthermore, if the export rate for woolen yarn is a favored rate, and the export rate for raw wool is a penalty rate, the domestic producer of woolen yarn is given an export subsidy equivalent to the difference between these two export rates, and this subsidy applies not only to the value added by manufacture, but even to the raw wool content of the export.

Concluding Observations

Multiple exchange rates are essentially a system of taxes and subsidies. Like all taxes and subsidies, they affect the distribution of income and the pattern of production and consumption. Their consequences, therefore, are of tremendous importance to a country’s economy. They affect the national economy through their impact on the prices of imports and exports. For this reason, they are of importance also to other countries.

As far as multiple exchange rates constitute a system of taxes and subsidies, affecting home distribution of income, they may or may not be economically desirable. In general, they are based on the assumption that some (or all) exports and some (or all) imports are a desirable source of taxation. In countries with greater facility for administering taxes, it would hardly be thought probable that, even if some (or all) producers of export goods and some (or all) consumers of import goods could reasonably be expected to pay more taxes, the most satisfactory method of increasing government revenues would be through special taxation of these groups. If a tax on producers is desirable, the economically most effective method is to tax the incomes of producers, whether of export goods or home goods; similarly, the most effective tax on consumers would be a tax on either their incomes or their expenditures.

The case for multiple rates as a tax device, however, does not rest on its economic merits. It rests rather on the fact that it is easy to impose, because it can be presented as an exchange device rather than as a taxing device; that it is easy to enforce, being concentrated on payments and receipts through international trade; and that it can be defended as a necessary means of strengthening a country’s payments. These are not good reasons for preferring one type of taxation to another. They may, however, have the merit, in countries with budgetary difficulties, of being better than no additional taxes.

The economic case for multiple rates as a subsidy on consumption is also weak. The distribution of real income through the price system, even when modified by taxation, may not always be equitable, and some further modification through subsidies on consumption may be desirable. But it is difficult to believe that the most constructive method of providing such subsidies is through bargain rates of exchange for imports. It would seem to be preferable to select consumer goods of great importance to the well-being of the poor, whose supply (from either home or import sources) could then be expanded by the most economic means available.

Subsidies on exports are subject to generally similar objections. They may indeed encourage production of goods for export that might not otherwise be produced. It does not, however, follow that they are more desirable than subsidies to encourage the domestic production of goods that would otherwise be imported or of goods that would be neither exported nor imported.

Taxes and subsidies in the form of multiple rates are likely to have undesirable effects on the patterns of production and of consumption. The classification of thousands of import and export goods in a few broad categories to which markedly different exchange rates are applied must necessarily be highly arbitrary, because it is assumed, on the one hand, that “essentiality” is an adequate test of the rate of exchange to be applied to import goods and, on the other, that the supply of most export goods is inelastic.

All this, it may be argued, merely shows that countries are not likely to get from multiple rates either the best pattern of production and consumption or the best system of taxation. But this is equally true of many other economic measures. Provided a country is itself satisfied that the immediate results of multiple rates are desirable, why should the Fund be concerned that the longer run effects are adverse?

The Fund was given the positive duty of passing on multiple rates, not only because of their effects on the member using them, but even more because of their possible effects on other members. As a system of taxes and subsidies on imports and exports, multiple rates affect the trade of other countries. It is true that a similar system of taxes and subsidies could be devised without using multiple rates; but many countries have defended their position with respect to such taxes and subsidies through bilateral trade agreements. They have entered into such agreements on the assumption that the indirect method of doing this by way of exchange rates has been brought under international control through the Fund Agreement. They are, for this reason, entitled to the assurance that the Fund will not approve extensions of multiple rates except where more practicable methods of dealing with the payments problem cannot be put into effect.

From the economic point of view a single rate for both exports and imports, that will balance international payments without restrictions, is the preferred means of correcting a country’s payments position. But suppose a country is unwilling to apply an appropriate rate, adequate to discourage imports, to all of its imports. As a practical matter, it may be better to apply the appropriate rate to 50 per cent or 60 per cent of its imports than to continue an inappropriate rate for all imports. And if a country is unwilling to apply an appropriate rate to all of its exports, it may be better to apply the appropriate rate to a portion of its exports. In such a case, it must be clearly demonstrated that the harm done to other countries by preferential import or export rates is negligible.

A large spread between a single export and a single import rate is probably the form of multiple rate which is least harmful to other countries. The discriminatory features imposed on trade by multiple exchange rates are then kept to a minimum, and the difference between the effective rates and an appropriate parity is unlikely to be very great. And because most countries are more specialized with respect to their exports than with respect to their imports, differential exchange rates are likely to be more harmful to other countries when applied to exports than when applied to imports.

December 1949


Mr. E. M. Bernstein is Director of the Research Department. He was formerly Professor of Economics in the University of North Carolina and Assistant to the Secretary in the United States Treasury. He is the author of Money and the Economic System.


The countries in this category are mainly producers of raw materials whose output is a small part of the total supply. The elasticity of world demand for their exports and of world supply of their imports is very large, and may be assumed to be infinite. The analysis requires modification for large industrial countries whose demand for imports affects the price of raw materials, and whose supply of exports can affect the prices at which they are sold.


Alfred Marshall, The Pure Theory of Domestic Value (London, 1879), p. 33, and Principles of Economics (8th edition, London, 1920), pp. 467-70.


Quantitative controls that limited imports from Europe would have the same detrimental effect on European exports to Latin America.


The classification of an export rate as a penalty rate or as a favored rate is partly a matter of definition. It might be said pro forma that no rate which is at parity is a penalty rate. From the economic point of view, however, any export rate which is less than the average import rate may be said to be a penalty rate. And any export rate which is more than the average import rate may be said to be a favored rate.