The 35 adjustment programs supported by the Fund with upper credit tranche stand-by or extended arrangements approved in 1983 were examined in some detail. A high proportion of the programs—71 percent—included action on the exchange rate.

The 35 adjustment programs supported by the Fund with upper credit tranche stand-by or extended arrangements approved in 1983 were examined in some detail. A high proportion of the programs—71 percent—included action on the exchange rate.

Ten arrangements approved in 1983 were in support of adjustment programs that did not include exchange rate action. Six of these involved members of currency unions—Central African Republic, Grenada, Mali,20 Niger, Senegal, and Togo—while in two others, Liberia and Panama, the U.S. dollar is legal tender. Because exchange rate changes in such countries would involve major institutional changes or joint action with other countries, there is a strong presumption that, if at all possible, adjustment should be carried out without resort to exchange rate action. The programs with Guatemala and Haiti also excluded exchange rate action, because the need was not considered great enough at the time the programs were designed to require departure from the long-standing parities vis-à-vis the U.S. dollar exhibited by their currencies. The ten cases where no exchange rate action was taken are discussed separately below.

The other 25 programs involved exchange rate action. In 11, frequent small depreciations were already a feature of exchange rate policy, though each of these programs somewhat modified previous policy, and in most there was a significant discrete devaluation at the outset. (See Table 3, details of which are discussed below.) In the remaining 14 countries, the currency was pegged to another currency or currency basket. In eight of these, one or more substantial depreciations had taken place in the previous year or two. Only six programs involved a change in a long-standing peg. Fifteen of the 25 arrangements followed immediately, or not long after previous arrangements. In several of these, the policy to be followed was largely a continuation of that implemented in the previous program. However, in most of them, problems in achieving the objectives of the previous program, either because of difficulties in policy implementation or because of further deterioration of the external environment, occasioned major shifts in exchange rate policy.

Table 3.

Mechanism of Exchange Rate Adjustment in Programs Approved in 19831

article image
article image
Source: Country authorities.

Classified by exchange rate policy prior to change in policies associated with program.

This section begins with a discussion of the considerations which formed the basis for changes in exchange rate policies that were incorporated in the 1983 programs. This is followed by discussion of the mechanism through which the exchange rate adjustment was implemented, including adaptations of policies in response to the particular circumstances of countries. The paper does not attempt to assess the effectiveness of the exchange rate actions. Such an assessment would have to be conducted within the broader frame-work of a case-by-case analysis of each adjustment program as a whole, and is beyond the scope of this study.

Besides references in the text to the experience of individual countries, the case studies in the appendix illustrate the variety of circumstances of countries seeking Fund support and the range of approaches that can be taken in selecting a set of policies to eliminate economic imbalances and adequately strengthen the balance of payments. One study is on Ghana, where the need for exchange rate action was evident from the outset, but the appropriate amount needed to be established. The other studies are on Sri Lanka, where there was a long history of economic adjustment involving exchange system liberalization and where the examination of exchange rate questions was thus particularly detailed, and Ecuador, where market mechanisms played a large role in exchange rate adjustment.

The Process of Formulating Exchange Rate Policies

In considering the process through which exchange rate adjustments were determined in the 1983 programs, it is important to bear in mind that, when work on an adjustment program formally begins, a general understanding of the nature of the problem usually already exists. Indeed, often there has already been extensive detailed analysis. Accordingly, objectives and constraints facing the authorities have generally been defined, and there is usually a general idea of the feasible range of demand management and other policies. From this starting point, an analytical framework is developed which considers exchange rate policies in relation to other policies and in relation to the whole range of program objectives.

This section does not attempt to deal with the full complexity of the formulation of exchange rate policies or the overall design of adjustment programs. Rather, it focuses on three main aspects: the circumstances facing the countries at the time they undertook exchange rate adjustment; the major ways in which exchange rate action was expected to improve the allocation of resources, both to correct the balance of payments and, more generally, to improve the overall functioning of the economy; and, finally, the way in which the precise amount of exchange rate action was determined, once the broad orders of magnitude had been established through the general analysis of the need for adjustment.

Circumstances Facing Countries Undertaking Exchange Rate Adjustment

Most of the 25 countries that followed an active exchange rate policy in their 1983 adjustment programs had, like most other developing countries, been affected by a deterioration in external conditions. This deterioration was due to a weakening of the external terms of trade, a rise in international interest rates, and an increasing reluctance of financial institutions to increase their exposure to developing countries. Moreover, most had followed financial and exchange rate policies that had resulted in a loss of external competitiveness. Table 4 presents a few indicators which capture certain essential aspects of the external performance of these countries. The data refer to 1982, the year in which work began on the design of most of the programs approved in 1983. The indices in the table measure changes since 1978, when most developing countries faced external conditions considerably more favorable than in the early 1980s.

Table 4.

Selected External Sector Indicators: Countries That Took Exchange Rate Actions in 1983 Programs

(Indices are based on 1978 = 100, unless otherwise noted)

article image
Sources: Country authorities; and Fund staff estimates.

Includes official transfers.

July 1981 to June 1982.

Represents official rate only.

Base year 1979 = 100.

July 1982 to June 1983.

Excludes gold, most of which is pledged.

All but three of the 25 countries following an active exchange rate policy had experienced a significant deterioration in terms of trade betweeen 1978 and 1982. The exceptions were Argentina and Ecuador, which were self-sufficient in energy, and Zimbabwe, which benefited from the lifting of trade sanctions on its exports and imports. The deterioration in the terms of trade had exceeded 10 percent in all but one of the other countries. In 10 countries, the deterioration had been particularly severe, exceeding 20 percent.21 In addition, developments since 1980 in the international capital markets adversely affected a large group of countries that had relied on external commercial borrowing to finance their current account deficits.22 The rise in real interest rates had sharply increased the burden of debt service for a large number of these countries, particularly those in the Western Hemisphere. At the same time, beginning in 1982, foreign commercial banks sharply curtailed the external financing provided to the major borrowers among these countries.

The deterioration in external conditions suggested a need for exchange rate and demand management policies consistent with an enhancement of competitiveness, but most of the 25 countries had experienced excess demand and real appreciation of their currencies.23 In seven of the countries, the real appreciation since 1978 had exceeded 10 percent. A beginning had been made toward real depreciation in previous programs for nine countries, but in only five of them had the real depreciation been more than 10 percent.24

Given the combination of deteriorating terms of trade and lack of real depreciation, most of the countries concerned were encountering severe current account problems.25 Eighteen had deficits exceeding 5 percent of gross domestic product in 1982. Such deficits were likely to be unsustainable for most, given their deteriorating access to foreign private capital, the unfavorable prospects for concessional aid, and problems of capital flight. In many of these countries, moreover, the deficit was held to that level only through severe demand restraint or exchange and trade restrictions. Of the countries with smaller deficits, Argentina and Brazil were encountering such severe deterioration in their access to capital markets that even seemingly moderate deficits were not financeable. In Ghana, Sudan, and Uganda, small deficits reflected a lack of access to foreign capital and intensive use of exchange restrictions. Two other countries with relatively modest deficits, Korea and Turkey, had both adopted outward-looking development strategies which required close attention to the maintenance of competitiveness. Moreover, Korea was reducing import duties, while Turkey had a very large external debt and was engaged in dismantling many of the restrictions that had developed during a prolonged period of strain in the balance of payments.

The severity of the crises facing most of these countries is also evident from their low levels of international reserves. In several, reserves had been virtually exhausted. In 18 countries, reserves amounted to the equivalent of less than three months of an already depressed level of imports. Another indication of the severity of the problems confronting these countries is the behavior of import volumes. Between 1978 and 1982, import volume declined in 16 of the 25 countries, in most of them by substantial amounts.26 In view of the relatively limited potential for import substitution in most of these countries, declines in import volumes tended to be associated with declines in economic activity and levels of consumption. This particular development perhaps best illustrates the circumstances calling for an active exchange rate policy. Such a policy would place a relatively greater emphasis on the expansion of exports and on a more market-oriented allocation of imports as a means of achieving external adjustment, thereby reducing the need for policies that are detrimental to growth.

Most of the countries had succeeded in increasing their export volumes somewhat. In seven, however, export volumes declined between 1978 and 1982. The decline exceeded 10 percent in Ecuador, Kenya, Madagascar,27 and Zaïre. In eight countries, the increase in export volume exceeded the average for all non-oil developing countries (29 percent), but still this was not enough in most cases to avoid serious current account problems. The largest increase (87 percent) was registered by Turkey which, having started with a severe external imbalance and a low level of exports, had made considerable progress toward adjustment by 1982. A major factor in Turkey’s strong export performance was the real depreciation of 35 percent that had taken place since the end of 1979.

Expected Consequences of Exchange Rate Action

Given the severity of the situations that most of the countries faced, it is not surprising that exchange rate adjustment needed to be considered. Before a firm conclusion could be reached, however, the way the adjustment would contribute to achieving the country’s policy objectives in each particular instance needed to be examined in detail. This subsection provides a summary of the expected effects of exchange rate adjustment emphasized in particular cases, either as stated explicitly in the documents presenting the request for a stand-by or extended arrangement to the Fund’s Executive Board, or as indicated implicitly in analyses of balance of payments problems in the request papers or consultation reports. The description, which is illustrative rather than exhaustive, focuses on highlights of particular cases. In practice, of course, each favorable effect would be present to some degree in almost all cases, but the relative magnitudes vary widely.

At a general level, most of the papers presenting the requests for arrangements stressed the importance of exchange rate adjustment both as a means of reducing the balance of payments deficit and as a means of improving resource allocation—objectives which are, of course, closely related. In connection with these objectives, the papers also emphasized the need to improve the competitiveness of the country’s tradable goods sectors.

Given the fact that for most countries there had already been a sharp reduction or, at best, little increase in imports in recent years, programs seldom emphasized the role of exchange rate adjustment in reducing import demand in the short run (aside from the way in which its expenditure reduction effects facilitated demand restraint more generally). The opportunity for import substitution over the medium term, when it depended on new investment to create additional production capacity, was commonly noted. In most cases, assessment of the impact of the exchange rate on import demand was made difficult by the fact that the recent behavior of imports had been strongly influenced by, on the one hand, excess aggregate demand that was to be corrected by the adjustment program and, on the other hand, restrictions that had prevented actual imports from reflecting changes in the amount demanded.

The main improvement in the trade balance was usually expected to come from recovery or new growth of exports. In most instances, detailed studies of key exports were an essential component of the underlying analysis. Often the need to restore the profitability of key export industries was the most striking indication of the need to undertake exchange rate action. This was the case in Bangladesh, Ghana, Sri Lanka, Sudan, and Zambia. In primary producing countries, moreover, it was considered essential that there be some diversification into manufactured exports. This was generally predicated on new investment and was expected to take time, so that, as with import substitution, the objective was of a longer-range character. By contrast, where the main objective was to reverse a previous decline in traditional primary exports, the response was often expected to be rapid.

Invisible transactions were also expected to respond to exchange rate action. For some types of transactions, such as tourism and direct investment, enhancement of competitiveness was emphasized. Other types involved more complex considerations of short-run expectations. Problems with emigrant remittances or capital flight, for example, were often considered to be due to a widely held opinion that the exchange rate was unsustainable, which offered a strong incentive to avoid repatriating funds before the expected exchange rate adjustment. Worker remittances were emphasized in Bangladesh, Portugal, Sudan, and Uganda. Capital flight was a widespread problem. In some countries where there had been large capital outflows, as in Argentina, Ecuador, and Uruguay, reversal of these flows was considered to be a means of obtaining a quick improvement in the balance of payments. Overinvoicing and underinvoicing of trade transactions were common means of capital flight in these countries and many others.

Rationalization of the trade and payments system so as to remove excessive import protection and encourage exports was a major reason for undertaking exchange rate action in several countries. In Turkey, in fact, the major reason for further real depreciation was to facilitate liberalization of the exchange system, although there was also a need to deal with large debt obligations in the medium term. The removal of recently imposed restrictions was often emphasized, as in Morocco and Portugal. Steps to reduce effective protection were often taken, notably in Korea. By contrast, temporary increases in restrictions were adopted in some particularly difficult situations, such as Sudan, though the long-term goal of liberalization was reaffirmed. In some cases, such as Kenya, import duties were increased as a partial substitute for stronger action on the exchange rate. In all instances of multiple rates, the eventual goal was to unify the system. Exchange rate action also was expected to help to eliminate payments arrears.

While this survey necessarily focuses on how exchange rate adjustment in itself was expected to improve the balance of payments and resource allocation, in each case these effects were considered in the context of an overall policy program. In most, it was clear from the outset that it was not feasible to bring about the needed adjustment in relative prices through demand management policies alone. In a few cases, however, where the misalignment of relative prices was not particularly large or entrenched, the decision to take exchange rate action emerged only after the sort of detailed consideration that this section summarizes.

Determining the Amount of Exchange Rate Adjustment

The analytical process described above often established not only the need for exchange rate action, but also the approximate extent of the needed change. It still remained to establish a definite figure for the amount of adjustment, or to select a mechanism which would determine the amount.28 A simple solution was to permit the rate to be set largely by market forces, as was done in the programs in Chile,29 Uruguay, and Zaïre. In other programs discussed below, the adjustment incorporated lesser elements of market determination. In most, however, management of the rate continued and the amount of adjustment was determined through an examination of indicators.

Balance of payments models incorporating elasticities with respect to the exchange rate played a limited role in specifying the amount of exchange rate adjustment in the 1983 programs. Although experience amply demonstrates the effectiveness of exchange rate adjustment in changing exports and imports, estimates of elasticities, as noted in Chapter IV, are subject to wide margins of error. This is particularly the case in developing countries, where accurate data are often lacking. In countries where there has been rapid structural change, quantification of the potential impact of exchange rate action is particularly imprecise. In Chile, for example, estimates of import and export elasticities were available but were not considered to be very useful, given the major changes that had occurred in recent years in the economic structure and policy framework. Where countries already have a fairly developed manufacturing sector, the responsiveness of production can be gauged with more confidence. Estimates of elasticities were more useful in Korea, for example, than in many other countries.

In 15 adjustment programs,30 indicators of competitiveness based on real effective exchange rate indices were the primary basis for the amount of exchange rate action. In some countries, such as Malawi, an earlier program was considered to have brought the exchange rate to an appropriate level, and the 1983 program involved restoration of the real effective exchange rate to the previously acceptable level. Where no such convenient reference point was available, it was necessary to select some historical period. Because most of the countries faced terms of trade that were extremely adverse by historical standards, it was clear that the real rate had to be considerably lower in a number of cases than it had generally been in the past. One alternative was to match the real rate with the lowest rate that had occurred in recent experience. In Brazil, for example, the initial devaluation was sufficient to restore the real rate to that recorded following the major devaluation of early 1979.

In many of these programs, the consideration of competitiveness indicators was supplemented by detailed studies of the profitability of key export industries (where one or two primary exports predominated) or by the elasticities estimated for manufacturing exports in general, such as in Korea. In Zambia, copper mining was so critical for both external and fiscal balance that the amount of exchange rate adjustment was determined largely on the basis of restoring the profitability of that industry.

These results were checked against developments in parallel market rates. Four programs went further, and took as the basis for action the convergence of official and legal parallel markets. In Argentina, the initial action was based on unification of the dual exchange system at the more depreciated of the two rates. In Sudan, the free market accounted for one third of transactions, and the initial exchange rate action was based on narrowing by 70 percent the gap between the official rate and the rate prevailing at that time in the parallel market. In the Dominican Republic, where change in the official rate was considered infeasible, a limited depreciation was accomplished by transferring certain transactions to the parallel market. In Uganda, both convergence and shifts of transactions took place.31

In Turkey, the amount of real devaluation was chosen as a reasonable figure consistent with the objective of a gradual liberalization of the exchange and trade system. In Portugal, the amount of the adjustment took into account the need to restore confidence and stem capital flight.

In most programs where the exchange rate was permitted largely to reflect market forces, there continued to be some degree of management of the parallel market or restrictions on access to it, so that the exchange rate was not a true market rate. The appropriateness of the rate thus continued to be checked against the other indicators used in exchange rate assessment.

While a particular basis for the amount of action in each of the 1983 programs can thus be identified, it must be kept in mind that the general order of magnitude of the action was established within the analytical framework that formed the basis for the overall design of the program, and that the role of these indicators was generally limited to selecting a figure in the appropriate range. Given the uncertainties inherent in economic analysis and forecasting, the decision on the amount of devaluation will always be, to some extent, arbitrary. Indicators such as those described above may be extremely useful in reaching a decision, provided that their limitations are kept in mind during application.

In a number of programs, part of the exchange rate adjustment was postponed until later stages. (See the discussion below on the mechanism of adjustment.) As in programs requiring lesser amounts of adjustment, these compromises came at some cost in terms of forgone output and unemployment.

The Mechanism of Adjustment

Ideally, full adjustment of the exchange rate takes place at the outset of a program. This sends a decisive signal to producers of tradable commodities on which they can base their production plans, and at the same time signals the authorities’ intention to adhere to policies consistent with maintaining the new rate. This assumes that inflation is not expected to erode the real effective exchange rate during the program period.32

In most of the 1983 adjustment programs, inflation was expected to be a continuing, though gradually moderating, problem. Most programs, therefore, envisaged further action during the program to maintain the real depreciation achieved by the initial devaluation. In some programs, the subsequent action was to go further, as part of a strategy that called for part of the real depreciation to be carried out during the program, rather than at the outset. Such gradual action generally leads to greater losses in terms of income and employment for a given amount of real depreciation. Where there is particular uncertainty as to the appropriate amount of adjustment, gradual action may be desirable, though again at the risk of prolonging the adjustment period or running up against financing constraints.

Most of the 1983 programs involved a substantial exchange rate action near the beginning of the adjustment period, sometimes implemented through a gradual process. Often the exchange rate already had been adjusted before the Fund approved the stand-by or extended arrangement (Table 3). Table 5 indicates the change in the real effective exchange rate index from the month before the exchange rate action began to the month the arrangement was approved (or the subsequent month, where substantial action took place in the month of approval); it also indicates the subsequent adjustment during the program period.33 Even where little change occurred, often there was a significant nominal adjustment that at least prevented real appreciation from occurring.34

Table 5.

Implementation of Exchange Rate Policy Programs Approved in 1983: Developments in Real Effective Exchange Rates Indices

article image
Source: Fund staff estimates.

Base year 1978 = 100.

Change between the month preceding initial adjustment and the month of Board approval.

Change between the month of Board approval and the final month of the arrangement (or December 1984 where the arrangement had not expired by that month).

The month following Board approval has been taken as the reference month so as to reflect fully the effects of an exchange rate action taken in the month of Board approval.

The value of the index refers to December 1984 in the cases of arrangements which had not expired by that date.

Official rate only.

The 1983 stand-by arrangement was immediately preceded by a 3-year stand-by arrangement.

In most of the programs, the initial exchange rate action achieved the full desired real depreciation, but subsequent adjustments were also necessary in some cases to maintain the real rate in the face of inflation, or were expected to occur in response to market forces. Argentina and Brazil adopted a policy of adjusting frequently the U.S. dollar rate broadly in line with the domestic inflation rate. (Full adjustment of the rate in line with domestic inflation would have the effect of producing some real depreciation against the reference currency, as there would be no offset for inflation in the United States. The real effective depreciation would depend on the behavior of the dollar relative to the currencies of other trading partners.) For Chile, daily depreciations were expected to be carried out on the basis of relative inflation. The mechanism that Ghana used was similar, except that it involved quarterly adjustments pn the basis of relative inflation during the preceding quarter. For Bangladesh, Malawi, and Sri Lanka, the objective was to prevent appreciation of the real effective exchange rate index. However, because these countries had relatively moderate inflation, adjustments of the rate were expected to occur only from time to time. In Kenya, Korea, Mauritius, Solomon Islands, Western Samoa, and Zimbabwe, the exchange rate was to be monitored with the aim of preserving competitiveness. Exchange rate policy in Ecuador, Portugal, and Zambia was to be determined largely on the basis of anticipated inflation. In Uruguay, the exchange rate was initially determined without central bank intervention to give time for the exchange rate to find a stable equilibrium level; subsequently, the exchange rate was expected to be managed flexibly. Morocco’s policy involved a series of discrete actions during the program period.

In the remaining programs, exchange rate policy aimed at a phased real depreciation, sometimes following an initial adjustment. In some of these, the phased approach was integral to the design of the program. Turkey’s liberalization of restrictions, for example, was to be accompanied by a phased depreciation in addition to adjustments to offset inflation. In Madagascar’s program, a planned phasing of the exchange rate action helped effect a faster pass-through to prices. In other examples of phased real depreciation, it was recognized that the change could mean slower progress toward achieving full economic adjustment. In the Philippines, there was particular concern about the social and political consequences of a substantial initial devaluation, so it was planned that the action would be phased during the program period. In Sudan, it was expected that further adjustments of the exchange rate would be made from time to time. The basic adjustment mechanism in the Dominican Republic, Uganda, and Zaïre involved multiple exchange rates.

The principal elements in each program’s adjustment mechanisms are as set out above, but other elements were often present that reflected the particular circumstances of the country. Besides the three multiple exchange rate cases already noted, other countries using multiple rates as part of their adjustment mechanisms included Ecuador, Ghana, and Sudan. As with the first three countries, the planned adjustment mechanism in these latter cases generally involved movement toward unification at the more depreciated rate. In Korea and Turkey, exchange rate policy took into account the implications for competitiveness of realignments of major currencies. Given the critical importance of copper exports for the Zambian economy, exchange rate policy was to be kept under review in the context of developments in that sector.

In practice, 15 programs achieved a real depreciation in excess of 10 percent by the end of the program period;35 7 had real depreciations less than 10 percent, and 3 had real appreciations. (Over the whole period from 1978, 12 had depreciations in excess of 10 percent, while 5 had real appreciations.) In 7 instances, the initial action was reversed or substantially eroded by subsequent real appreciation. Where real depreciation fell short of expectations, weaknesses in implementation were sometimes the cause. External events, such as the unexpected appreciation of the U.S. dollar, were also a factor in some cases.

Formulation of Exchange Rate Policies in Special Circumstances

The importance attached to particular policy objectives and institutional arrangements varies from country to country. The importance of price stability, for example, is generally accepted, but the vigor with which countries pursue that objective varies, depending on the real or perceived trade-offs with other objectives. Similarly, some countries attach particular importance to stability of the exchange rate because of their institutional arrangements. There is also wide variation in the extent to which countries use planning mechanisms and direct controls—as opposed to relying on market forces—to manage their economies. Central planning is the strongest of this type of institutional arrangement. This section examines some of the issues involved in the formulation of exchange rate policies in such special circumstances.

Close Links to Other Currencies

After the end of the par value system, most countries gradually have come to place less emphasis on exchange rate stability as an overriding objective of economic policy. Nevertheless, a number of countries undertaking Fund-supported adjustment programs have preferred to follow demand management and related policies consistent with maintaining an existing peg, even in cases where it might be argued that adjustment could be achieved more efficiently with exchange rate action.

The different preferences of countries with regard to maintaining an existing peg are, of course, always reflected in the mix of policies. Exchange rate stability, however, continues to have particular importance for members of currency unions and countries with long-standing close links to a major currency. Countries in these categories that did not adjust their exchange rates in adjustment programs undertaken in 1983 are listed in Table 6, which includes data on certain indicators corresponding to those presented in Table 4 for the other program countries. (As noted previously, the Dominican Republic undertook exchange rate action by expanding its parallel market, though the traditional parity with the U.S. dollar was maintained for the official rate.) In each of these programs, the appropriateness of the level of the exchange rate was considered, and it was concluded that a feasible adjustment program could be designed without exchange rate action. All countries in these categories, except Panama, suffered severe deterioration of the terms of trade, but in other respects their situation was generally not as critical as in the countries that took exchange rate action.

Table 6.

Selected External Sector Indicators: Countries That Did Not Take Exchange Rate Action in 1983 Programs

(Indices are based on 1978 = 100)

article image
Sources: Country authorities; and Fund staff estimates.

Including official transfers.

July 1982 to June 1983.

October 1981 to September 1982.

For four countries with strong ties to the U.S. dollar, reversal of the real appreciation that had occurred between 1978 and 1982 was predicated largely on an expectation that the U.S. dollar would decline relative to other major currencies. In Haiti, where the gourde had been pegged to the U.S. dollar at an unchanged rate since 1929, financial policies were considered adequate to deal with the external imbalance. The improvement in public finances, together with a tightness of credit policies under an immediately preceding stand-by arrangement, had eased the pressures on the balance of payments and markedly reduced the discount on the gourde in the parallel market that had developed. In Guatemala, no significant relative price misalignment was believed to exist at the time the program was designed, and the external imbalance was to be tackled with a tightening of financial policies and the introduction of export incentives through fiscal measures. In Liberia and Panama, where the dollar was legal tender, changing the peg would have involved a major institutional change. In Panama, moreover, the problem was neither a deterioration of the terms of trade nor embedded inflation, so adjustment through demand management was preferable in any case.

The currencies of five African countries were pegged to the French franc in the context of monetary unions.36 The maintenance by these countries of exchange systems free of restrictions on current transactions had made it possible to avoid major cost-price distortions. Real appreciation since 1978 thus had generally been small (negative in the case of Senegal).

Another country that did not take exchange rate action was Grenada, whose currency, the East Caribbean dollar, is issued by the East Caribbean Central Bank. The current account deficit and substantial real appreciation, coming at a time when the terms of trade had deteriorated severely, made a strong case for devaluation. Grenada’s high rate of inflation and a large share of countries other than the United States in its trade and tourism meant that its real effective appreciation was considerably larger than those of the other members of the East Caribbean Central Bank. However, in light of the major institutional change that an adjustment of the exchange rate would have required, a program was developed that did not include initial exchange rate action. The question of a possible subsequent exchange rate action by Grenada was to be considered by the authorities in consultation with the other members of the East Caribbean Central Bank.

Countries with a History of High Inflation

At the opposite extreme with regard to exchange rate flexibility are those countries with a long experience of high inflation that protect external competitiveness through frequent quasi-automatic devaluations. In such countries, the usual political resistance to devaluation is substantially reduced, and there is little difficulty in incorporating an active exchange rate policy into the adjustment program. It is generally understood from the outset that continuing depreciation will at least offset inflation, and there is little problem in accelerating the rate slightly to produce a real depreciation.

While the desirability of real depreciation in such cases, or at least the avoidance of real appreciation, is not in doubt, the question may be raised as to whether some degree of nominal exchange rate stability should be an important goal of exchange rate policy. Stability can be brought about through, for example, strong incomes policy in association with financial restraint, as opposed to a continued sanctioning of high rates of inflation through rapid depreciation. The relative efficiency of these quite different approaches to adjustment goes beyond exchange rate policy into the area of the overall design and implementation of adjustment programs, and is thus beyond the scope of this paper. However, there is no doubt that programs should plan for an eventual return to exchange rate stability. In fact, the 1983 programs of high inflation countries generally did plan for a gradual slowing of inflation, implying a declining rate of nominal depreciation of the currency. In practice, however, many of them did not succeed in achieving this objective.

Planned Economies

All countries use administrative action to limit the extent that market forces influence economic developments. To the degree that such administrative controls prevail, prices, production, trade, and income distribution may be more or less insulated from the effects of exchange rate changes. An important aspect of exchange rate policy is thus the extent to which changes in the exchange rate are passed through to the rest of the economy. In designing adjustment programs involving exchange rate action, such considerations always need to be kept in mind.

In this respect, there is no sharp distinction between countries formally described as having “planned economies” and others. Hungary, for example, in some respects permits the exchange rate to have more pervasive domestic effects than do many countries that are nominally market economies but that insulate key domestic producer and consumer prices from developments abroad. Other planned economies also increasingly emphasize market-related mechanisms.

It is nonetheless true that for planned economies, limits on the use of market mechanisms are often an integral part of the institutional structure, while for most other developing countries such limits tend to arise as ad hoc responses to particular developments, such as changes in income distribution, that are seen as undesirable from a social or political point of view. One important aspect of this difference is that ad hoc administrative responses tend to be biased against the development of a country’s potential to benefit from the gains from trade. The result is an increasingly inward-looking pattern of growth. On the other hand, planned economies often make deliberate efforts to expand their external sectors. However, unless decisions for such expansion are based on a proper appraisal of opportunity costs, it may not be carried out efficiently. In any event, the fact that market mechanisms are, in principle, allowed only a limited role in such economies may have some effects on the formulation of exchange rate policies.

Because no arrangements with planned economies were approved in 1983, this paper reviews the formulation of exchange rate policies in the 1982 and 1984 Fund-supported adjustment programs of Hungary and the 1981 program of Romania. In Hungary, the experience was similar to that in many of the 1983 programs reviewed above. The principal elements in the quantification of the exchange rate action were the profitability of key export industries, as well as estimates of elasticities which were used to calculate the exchange rate consistent with the external current account objectives. Developments in the real effective exchange rate index were considered, but the extensive restructuring of domestic relative prices that had taken place meant that the usefulness of the index was limited.

Romania, by contrast, did present a number of special features, although there, too, the adjustment program was developed in the context of a major price and exchange reform initiated earlier. A key element of this reform was the abolition of the “equalization” system that had previously been used to insulate the domestic prices of traded goods from developments in international prices, effectively setting a separate variable exchange rate for each commodity. This was replaced by a system of 27 exchange rates, which meant that changes in world prices would be reflected directly in the profits and losses of trading enterprises, provided they were not offset by adjustments in taxes and subsidies.

Adjustments to the exchange rate and further simplification of the system took place in the second and third years of the adjustment program, leading to a unified commercial exchange rate on July 1, 1983. The amount of adjustment was largely based on developments in the real effective exchange rate index, though interpretation of the index was complicated by the price reforms which were going forward. It was believed that the increase in the real effective exchange rate index understated the loss of competitiveness, in view of the existence of repressed inflation. The implication of the institutional reforms being undertaken by Romania was that the exchange rate changes were to be passed through to domestic prices to a significant degree. The complexity of the pricing system, however, together with the absence of detailed commodity analyses, made it exceptionally difficult to develop and monitor policies in this key area. In addition, it was recognized that the system of planning would limit the flexibility of the export sector to respond to exchange rate adjustments.

While a full appraisal of the formulation of exchange rate policies in planned economies would go well beyond the scope of this paper, experience suggests that the special circumstances of such countries pose few conceptual difficulties. There are, however, some problems of an essentially informational character. Where a country relies on market mechanisms in achieving the desired effects of exchange rate changes, it is possible to predict with reasonable confidence the responses of the economy, even without detailed sector-by-sector information. To the extent that planned economies themselves make use of market mechanisms, the same conclusion applies. Where decisions on prices and production are made administratively, however, it is necessary to be able to trace on a case-by-case basis the ways in which the exchange rate will enter into those decisions. This may imply a need for detailed analysis of the planning process for each industry or sector. At a more general level, a thorough understanding of the way the overall plan is formulated and executed is required to provide an adequate basis for appraising the consistency of exchange rate policies with other policies.