The Choice of an Exchange Rate Regime 1

Richard Bart, and Chorng-Huey Wong
Published Date:
June 1994
  • ShareShare
Show Summary Details
Manuel Guitián

I. Introduction

Exchange rate management, which is among the most discussed economic policy subjects, is one on which a consensus is hard to gather. To a large extent, this state of affairs reflects the complex interrelationship between exchange rate management and domestic economic policy, as well as the importance of these two policy areas for economic performance.

Some contentious points come to mind in the context of the seminar. Its very title, Exchange Rate Policies in Developing and Transition Economies, provides a good illustration of the absence of consensus that I have just mentioned. The title conveys, at least by implication, the idea that exchange rate policy varies or can vary depending on the type of economy. I, for one, do not think such a proposition is valid. Certainly, the workings of a particular exchange rate policy or regime depend on (and will differ with) the characteristics of the economy in which it is being pursued or established. But the fundamental tenets of exchange rate policy itself apply to any and all types of economy. Yet the notion that there are exchange rate policies that apply to or are more appropriate for only one type of economy is fairly widespread.1 My own impression is that there is a unity, a oneness, to exchange rate policy which can get lost in arguments that describe this important policy area as if it acted differently in different economies. Therefore, although the seminar focuses on developing and reforming economies, the major issues I will raise and the main points I will make in my remarks have wider applicability.

Exchange rate management has numerous dimensions and can be viewed from many perspectives. Most, if not all, of them can be classified into two broad categories: the macroeconomic aspect, which is concerned with issues of domestic financial stability; and the microeconomic aspect, which focuses instead on the question of the international competitiveness of the economy. Much of the absence of consensus that I noted at the outset reflects the prevalence of intellectual or analytical positions that stress one or the other of these two aspects. Those who focus on the macroeconomic dimensions of exchange rate policy emphasize the importance of establishing a clear and credible anchor as one element of a policy strategy aimed at domestic price level stability, a nominal variable. Although the argument is nowadays phrased in terms of the role of the exchange rate as a nominal anchor, it nevertheless addresses an age-old concern. The reasoning underlying the usefulness of this type of nominal anchor—or of the pursuit of hard currency options—closely parallels the case made in earlier periods for the exchange rate’s important role in ensuring discipline in domestic financial management.2

In contrast, those who focus on the microeconomic dimensions of exchange rate policy stress the importance of maintaining international competitiveness in any economy. The argument points accurately to the need in open economies to keep a viable, sound balance of payments position, a need that requires those economies to be and remain competitive or, in other words, to pursue an exchange rate policy geared to a real variable. This line of reasoning can also be traced far back in the literature on exchange rates.

Indeed, these two different perspectives or angles are modern versions of the relative weight given in the classical potential conflict between domestic and foreign conditions or exigencies to the attainment of internal balance (price level stability with sustainable employment and growth rates) as opposed to external balance (balance of payments viability).3 An important difference between current and previous analyses, though, is the international setting in which they take place. A prominent feature of present discussions of exchange rate management, and more generally, of domestic economic policies, is the presence of a closely interdependent world economic space. Trade and current account flows, which have been increasingly liberalized over the past four decades, have linked national economies, but perhaps more importantly, the growing international capital movements that have been progressively freed during the last ten years have strengthened and consolidated those links. As a result, the permeability of national boundaries has increased markedly, to the detriment of their ability to keep national economic domains separate.4 Consequently, although conflicts between internal and external balance or stability can and do continue to arise, their potential duration is becoming increasingly limited—a feature of the current world economic setting that warrants attention from the standpoint of economic policy formulation in general and from the perspective of the choice of an exchange rate regime in particular.

The plan of my remarks will be the following: Section II focuses on the nature of and the main considerations involved in the choice of an exchange rate regime. Section III examines the implications of this choice for the design and implementation of domestic economic policy; in turn, this examination provides the basis for a discussion of the dilemmas that must be confronted when selecting an exchange rate system (Section IV). The linkage of exchange rate management with other domestic policies and the consequent dilemmas are considered from the standpoint of currency convertibility, a concept that has acquired renewed interest with the prospective integration into the international system of the previously centrally planned economies (Section V). Finally, the central points and their relevance for developing and reforming economies are laid out as concluding remarks in Section VI. Table 1 outlines schematically various types of existing exchange rate arrangements.

II. Making the Choice

The choice of an exchange rate regime revolves around two crucial issues: the relationship of national economies to the global system and the degree of activism envisaged for domestic economic policies. With regard to the first issue, the choice of an exchange rate regime amounts to the expression of a national preference for either an open or a closed system. The interaction of national economies can be described as consisting of two basic elements: the point of intersection, represented by the balance of payments; and the terms of intersection, represented by the existing exchange arrangements, in particular the exchange rate regime. Both elements are, of course, interdependent, as any economic relationship between a quantity variable (balance of payments flows) and a price variable (exchange rate, exchange regime) must be.

Table 1:Types of Exchange Rate Regimes
Basically Fixed Regimes: Pegged Exchange Rates
  • Vis-à-vis a single currency: economies that peg to major international currencies with no or rare parity adjustments; economies that announce a prearranged schedule of exchange rate adjustments against the currency of the peg (the exchange rate changes, but at a fixed pace).

  • Vis-à-vis a currency basket: economies that peg to a basket of currencies of their main trading partners or to standardized currency composites such as the European currency unit (ECU) or the SDR.

  • Within pre-established margins: economies that peg to a single currency or a currency basket within certain (typically narrow) margins.

  • Fixed but adjustable peg: the arrangement that prevailed under the Bretton Woods par value system.

Basically Flexible Regimes: Adjustable and Flexible Exchange Rates
  • Indicators: economies that adjust their currencies automatically to changes in selected indicators, such as developments in the real effective exchange rate.

  • Managed float: economies that adjust their exchange rates frequently on the basis of judgments made following developments in variables such as reserves and the payments position.

  • Independent float: economies that let markets and market forces determine the exchange rates for their currencies.

The available choices for exchange rate systems lean either toward a fixed exchange rate and/or a flexible exchange rate. The connection I make between this choice and the preference for an open or closed system is based on the following considerations. Selecting a fixed exchange rate arrangement is equivalent to accepting a constraint on national economic policies. Of all the combinations of domestic economic policy stances and mixes that a country, taken in isolation, can in principle adopt, a fixed exchange rate most effectively limits the range of possibilities. The country will be able to adopt only those policy combinations consistent with maintaining the fixed exchange rate; domestic policy formulation thus becomes endogenous and subject to the exchange rate commitment. In a nutshell, this option is tantamount to placing an international constraint on national economic policies. In this generic sense, the choice of a fixed exchange rate regime is equivalent to a preference for an open system, in that no insulation is sought from the interactions between the national component and the world system.

In contrast, a flexible exchange rate arrangement, in principle, indicates a desire to accept no constraint on the pursuit of any particular domestic economic policy package. Whatever the effects of the policies undertaken may be, exchange rate fluctuations will keep them within the domestic domain. And correspondingly, whatever the consequences for the national economy of policies elsewhere, exchange rate adjustments will keep them outside the domestic domain. In fact, this option is tantamount to keeping national economic policy free from international constraints. In this broad sense, the preference for a flexible exchange regime reveals a corresponding preference for a closed system, in that the exchange rate arrangement that has been chosen will insulate the national economy from the international environment.

Therefore, the issue at stake in the choice of an exchange rate regime goes well beyond the technical aspects of exchange rate management. In effect, it entails a preference for internationally or nationally based systems and reflects the relative importance given to international and national considerations and objectives. The types of interaction between the part and the whole that will develop under a fixed or a flexible exchange rate are quite different. Under the former, the effects of internal and external policy impacts and shocks will be disseminated across the system at large. Under the latter, in contrast, each component of the international system remains insulated so that policy impacts and shocks do not carry across national boundaries.

Let me now discuss the issue of the degree of activism envisaged or sought for domestic economic policies. The differences on this score are quite marked. A fixed exchange rate, because it represents a commitment that sets a constraint (as already noted) on national policies means that domestic economic policies cannot be pursued independently. This is a feature typical of open systems, which, because of the inevitability of leakages, cannot but discourage independent action. A flexible exchange rate, in contrast, is a policy instrument that can be used to keep the scope of domestic policy action unconstrained by the economy’s participation in the international system. Accordingly, national policies can be pursued actively without concern for the outside world, and this arrangement is therefore akin, in its basic features, to a closed system.5

So far, I have described the nature of and the considerations involved in the choice of an exchange rate regime in the broadest of senses, because I believe it is in that context that relevant issues must be considered. There are, however, other criteria to guide the choice, and I now turn to these.

The criteria in question are technical in character and focus on the following three issues: the types of disturbance to which economies are exposed, the structural characteristics of those economies, and the commonality of the risks to which they are subject and the objectives that they pursue.6 Conventional wisdom relates the comparative advantage of the two exchange rate regimes to the types of disturbances to which the economy is most likely to be subject. Formal analyses conducted on the basis of a downwardly rigid cost-price model of output determination exhibit the insulating properties of a market-determined flexible exchange rate when an economy experiences external nominal shocks. Whatever effects these shocks could have had on the foreign price level have been countered by compensating changes in the exchange rate, shielding the domestic economy from the disturbance.7

As for domestic disturbances, the traditional line of reasoning distinguishes them as either real or nominal in nature, on the grounds that this is the relevant distinction for purposes of choosing an exchange rate regime. Domestic nominal shocks, such as those originating from money market imbalances, are best handled under a fixed exchange rate system. For example, a disturbance that creates an excess supply of domestic currency will lead to a balance of payments deficit that will restore balance to the money market through international reserve losses; the disturbance, in these circumstances, does not spill over to the real economy, which is protected by the fixed exchange rate regime. In contrast, domestic real shocks, such as those from imbalances in the goods market, are best coped with under a flexible exchange rate arrangement, because shocks to domestic demand will lead to changes in the exchange rate that will bring about offsetting movements in foreign demand, so that domestic output is not severely affected.

Interesting though these analytical findings are likely to be, the reality is that all economies confront both nominal and real shocks. Yet a shift in exchange rate regime in response to the nature of shocks is clearly an unworkable proposition. Apart from the uncertainty such a variable course of action would entail, there is the pragmatic issue of determining the nature of the disturbance affecting the economy, not always an easy task. In purely practical terms, selecting an exchange rate regime on this basis is far less suitable, despite the conceptual validity of this distinction.8

A second set of considerations to guide the choice between exchange rate regimes focuses on the structural characteristics of the economies in question. These considerations, developed in an effort to ascertain the scope of an optimum currency area, focus on the degree of mobility of labor (and more generally, of factors of production) among the various economies in the system—that is, on a measure of economic openness in terms of the markets for production inputs. The established conclusion is that a high degree of mobility of factors of production advocates in favor of the use of a fixed exchange rate. The reasoning has been extended to openness in terms of (and to the diversity of product mix in) the market for goods and services to lead to the same conclusion: open economies with these characteristics are better off fixing the rate of exchange.9

A third set of considerations with a bearing on the choice of exchange rate regime stresses the extent to which national economies either share common policy aims, exhibit similar policy attitudes, or both. The reasoning in this context leads of course to the inference that the more (less) common the policy objectives and the more (less) similar the policy attitudes, the more efficient a fixed (flexible) exchange rate regime will be.

In contrast, those systems based on the structural (openness) or attitudinal (commonality of objectives and policies) characteristics of economies are more operational. It is indeed possible to determine the openness of an economy, just as it is to ascertain how widespread and widely shared are its policy instruments and objectives. And consequently, criteria such as these can be (and have been) used to select an exchange rate arrangement.10

Fundamentally, all these various sets of criteria (including, to a certain extent, those based on the nature of shocks) can be subsumed in the more general considerations of the choice between open and closed systems. International mobility, openness, and common aims, all of which advocate for fixed exchange rates, are typical features of an open system. In contrast, their absence, which favors flexible exchange rates, is representative of closed systems.

III. Ramifications for Economic Policy

In order to illustrate the differences between the two types of exchange regimes, this section traces the relationship between exchange rate management and other domestic economic policies in the context of the correction of imbalances. In the domain of economic policy, establishing and maintaining appropriate economic incentives in general and price incentives in particular are of great relevance, if at times controversial. Their relevance, and also their contentiousness, reflect the complexities and difficulties involved in designing and implementing policy actions that affect key variables in the economy, such as the exchange rate or the structure of interest rates. They also reflect the critical role these variables play in allocating resources among economic sectors and agents.

At times, caveats are voiced concerning the dangers of undue reliance on price signals as the main instrument to guide resource allocation and use. These caveats typically stress the notion that price incentives place too much emphasis on the role of market forces and private enterprises. Yet the issue to be confronted is the need to ensure that economic decision making is based on the accurately measured opportunity cost of scarce resources. For this purpose, prices are relevant to all economies and to all decision makers. The acknowledgment that the pricing system serves a central function based on the interplay of market forces has become increasingly accepted over the last decade, as more and more countries have come to realize that keeping an appropriate structure of relative prices and costs in the economy contributes directly to its productive capacity and efficiency.

Generally, domestic imbalances can be traced to the pursuit of macroeconomic policies that are not compatible with the economy’s potential level and rate of growth of production. Such imbalances, unless corrected promptly, result in movements of domestic prices and costs that diverge from those occurring abroad. As a consequence, the economy’s ability to compete internationally is put at risk, and the efficiency of domestic resource allocation is impaired by the emergence of distortions in the structure of relative prices. When circumstances such as these are allowed to persist, inflation and balance of payments pressures build up, and the rate of growth soon falters.

Among the possible means of correcting situations of this nature is the well-known option of an exchange rate adjustment, be it policy induced (as would be the case in a fixed exchange rate regime) or market determined (as would be the case under a flexible exchange rate arrangement). Exchange rate adjustments are highly visible events, and as such they can lead to controversy and resistance. The typical conceptual rationalization of such resistance is the argument that an imbalance created by inadequate domestic macroecnomic management should be rectified by correcting macroeconomic policy. In the abstract, this reasoning is correct. But whether or not a policy strategy that excludes exchange rate adjustments is superior from an efficiency and welfare standpoint to one that includes them is a question that cannot be settled a priori.11 A theoretical proposition of great importance in this context warrants stressing: there is no lasting trade-off between exchange rate adjustments and appropriate domestic financial policies. In the context of a flexible exchange rate, what this means is that exchange rate depreciations are no substitute for and cannot replace adequate policies. And in the context of a fixed exchange rate, the implication is that devaluation does not eliminate the need for domestic financial restraint. Exchange rate adjustments, though, have a bearing on the degree of strictness in domestic policy.

As briefly noted earlier, when domestic prices and costs (or their rates of increase) are rigid in the downward direction, relying exclusively on domestic demand management to restore balance to the economy may be unnecessarily costly in terms of employment and output. The adjustment will require a decline in the level or rate of increase of domestic prices and costs relative to those prevailing abroad, and, with the postulated downward rigidity, such a decline will place the brunt of the adjustment on output and employment. In these circumstances, it has been argued, it would be more appropriate to accompany demand policy with an exchange rate adjustment.12 This course of action does not require a decline in the domestic cost-price structure; relative price balance (i.e., competitiveness) is restored by raising the domestic currency price of internationally traded goods rather than by lowering the domestic price of non-traded goods.

These considerations illustrate the extent to which exchange rate management and macroeconomic policy are interrelated. But the association is closest between exchange rate and monetary management. This association will now be examined not only to bring out the limitations of each of these policies, but also to highlight their potential for mutual reinforcement. In the process, the appropriate variable for monetary management will be identified, an issue for which the presence or absence of an external constraint is of relevance, as is the type of exchange rate regime in effect.

From the perspective of the money market, there are two main sources of liquidity in an open economy: the credit extended by the central bank and, more broadly, by the banking system to the rest of the economy through purchases of domestic assets; and the supply of money that the banking system provides through purchases of international reserves and foreign assets.

For purposes of stabilization, balance is required in the money market, calling for the monetary expansions that actually take place to be commensurate with the growth in the demand for money. This is the flow equilibrium dimension of the money market, which has as a broad counterpart the prevalence of balance between expenditure and income in the economy. In addition to flow equilibrium, stabilization requires balance in the stocks in the system—that is, the actual and the desired stocks of money and international reserves must be equal. In the absence of such stock-flow balance, adjustments will take place that may call for policy action.

Absence of external constraint

In a hypothetical closed economy, there is only one source of liquidity: purchases of domestic assets by the central bank or the banking system at large. In this setting, an excess expansion of credit, or the existence of credit flows that surpass the rate at which the economy is willing to increase its money holdings, tend to raise expenditure over income. Prices and, temporarily, output increase as a result. The demand for cash balances rises with them, helping to restore balance to the money market and contribute to stability in the economy at large. This, of course, is the sequence of events only if the excess supply in the money market does not persist. Otherwise, prices continue rising (inflation takes off), and output stops growing as resources are fully employed. As a consequence, the persistent money market imbalance perpetuates inflation; if the degree of imbalance remains unchanged, so does inflation, because the value of the excess nominal money holdings is precisely eroded by given, constant price increases, and actual and real cash balances will once again coincide.

Conceptually, the closed economy can be compared with an open economy that operates under a freely fluctuating exchange rate regime, where the sources of liquidity are domestic and foreign asset purchases by the central bank or banking system. If the exchange rate is allowed to fluctuate freely on the basis of market forces, the central bank will not intervene in the foreign exchange market, and the link between foreign exchange flows and the money supply process will be severed. A money market imbalance in this setting, besides fueling inflation, creates excess demand for foreign exchange and pressure for an exchange rate depreciation. Price increases, rising output, and devaluations may correct the imbalance, but only if the causes do not persist. If they do, inflation will continue and depreciation will follow. In this scenario, the respective rates and the variability in both prices and the exchange rate are determined by the degree and variation of the imbalance in the money market.

In these two situations, the monetary authorities determine the nominal quantity of money, while the rest of the economy determines its real value. Any incipient or persistent differences between the two are eliminated either by a change in the price levels and exchange rate or by ongoing inflation and depreciation.

Presence of external constraint

In open economies with a fixed rate of exchange or with limited exchange rate flexibility, a link is established between the money supply and the balance of payments that must be taken into account in the formulation of monetary policy. In this setting, both the monetary authorities and the economy as a whole influence the nominal supply of money in the system through balance of payments outcomes that result from the interaction of policy with economic performance.

Achieving stability in open economies requires coincidence between domestic credit expansion and money demand growth, together with equilibrium in the balance of payments. These two factors are important because a given rate of growth in the demand for money is compatible with different balance of payments results, and therefore money market equilibrium need not coincide with external balance.

Transmission of monetary impulses

The channels through which monetary policy operates reflect the behavior of the different sectors in the economy. In general, as the latter grow, so does the demand for money, reflecting a decision by the community to devote part of its growing income and wealth to increasing its money balances. Resources are thus provided to the banking system, which channels them to finance activities where, for a time, outlays exceed income. In this way, saving and investment flows are brought together by bank intermediation. Whenever the banks expand credit in amounts that differ from those the holders of cash balances make available, prices, output, and the balance of payments adjust. Thus, in designing monetary policy, it is possible to render the monetary authorities’ international reserve objectives compatible with the evolution of the demand for money in the economy. In this context, it has been argued that monetary policy in the open economy constitutes a powerful tool for aggregate demand and balance of payments management. This said, however, it should be pointed out that deviations of monetary policy from economic fundamentals cannot be sustained and that the effectiveness of monetary policy in this regard has often been overstressed.

Capital movements

The analysis so far has proceeded without considering capital account transactions, the introduction of which adds a market for domestic and foreign securities to the economy. The presence of capital flows, however, does not alter the relationships already discussed. If anything, it strengthens the link between domestic credit, the balance of payments, and the international reserve position. With capital mobility, for example, excessive domestic credit expansion relative to the growth in the demand for money will induce a current account deficit, a net capital outflow, and hence a decrease in net international reserves. The balance of payments thus continues to be one of the channels through which the supply of money is adjusted to demand. An important consequence of capital movements that warrants early mention is the intimate link they create between monetary and exchange rate policies. Any inconsistency between these two types of policy will be rapidly eliminated by the free flow of capital.

IV. Policy Interactions and Dilemmas

An important issue in the implementation of monetary policy in the context of any exchange rate regime is the type of instrument variable that should be used. A well-established tenet in this regard is that domestic credit expansion is the appropriate variable to monitor, as in open economies the money supply is outside the monetary authorities’ control. The logic behind this tenet is that the rate of domestic credit expansion is closely linked to aggregate expenditure and demand. The demand for money, on the other hand, is related to income and wealth. Therefore, choosing domestic credit expansion as the monetary policy instrument is simply recognizing that the counterpart to an imbalance between credit and money increases is a discrepancy between expenditure and income flows.

Instances of excessive money growth due to external (foreign exchange) surpluses have stimulated arguments in favor of using the money supply as the policy instrument, a view also supported by the direct relationship between money and prices (a relationship, it must be noted, that keeps implicit the links between money and spending and between spending and prices). Clearly important in this context is determining whether the balance of payments surpluses are the result of an increase in the demand for money (in which case the resulting monetary growth could not be deemed excessive) or whether they are due to other factors (in which case policy dilemmas could arise, such as the need to choose between controlling inflation and maintaining competitiveness). The outcome of external surpluses or of unanticipated net capital inflows will, of course, depend on the exchange rate regime in existence, providing yet another illustration of the closeness between money and exchange rates.

A foreign exchange surplus develops following a sequence of events that depend on the nature of exchange arrangements. With a fixed exchange rate, both net international reserves and the money supply rise at a higher-than-anticipated rate. As already noted, no inflationary pressure need arise if the surplus is caused by money demand developments. In similar circumstances, but with a flexible exchange rate, there is no monetary expansion owing to the incipient balance of payments surplus. Instead, the exchange rate appreciates, and downward pressure is exerted on inflation or the price level. Credit and money developments coincide in these circumstances, but credit expansion is still the policy instrument used to keep expenditures under control. With a partially flexible exchange rate, a combination of the two above sequences of events takes place: some monetary expansion, a measure of exchange rate appreciation, and downward price pressure occur; net international reserves increase, but not at the same rate as the balance of payments surplus. Other things being equal, once the unexpected external outcome works itself throughout the economy, real cash balances will be the same in all situations. Nominal money balances, however, will differ: they will be highest in the economy with a fixed exchange rate and lowest in the economy with a freely fluctuating exchange rate.

When the external surplus is not the result of money demand developments, a different sequence of events unfolds. With a fixed exchange rate, the resulting monetary expansion is voluntarily held, expenditures rise (so that the surplus diminishes with a growing current account deficit), and prices tend to increase. With a flexible exchange rate, the exchange rate appreciation tends to eliminate the surplus. With incomplete exchange rate flexibility, some monetary expansion and upward price pressure take place, together with a measure of exchange rate appreciation.

The adjustment is not open ended, however. In the first (fixed exchange rate) case, pressure on aggregate demand and prices deflects more and more spending abroad, eliminating the impact of the foreign exchange surplus on the domestic money market. In the second (flexible exchange rate) case, exchange rate appreciation initiates diversion toward the rest of the world. In the third (intermediate) case, higher domestic prices and lower foreign prices (in domestic currency terms) divert outlays abroad. In all three cases, the various adjustments in the economy eliminate the foreign exchange surplus.

One conceivable by-product of unanticipated foreign exchange surpluses is that in the process of reacting to the disturbance, the exchange rate may become unrealistic and require direct adjustment, supporting adaptations of other economic policies, or both. An exchange rate appreciation is most likely to occur whenever the foreign exchange inflows do not respond to developments in fundamental economic factors but rather are a manifestation of inflation abroad.

When the initial situation is one of external and internal equilibrium, exchange rate appreciation caused by unexpected foreign exchange accumulation can pose a dilemma in terms of maintaining the competitiveness of the economy. When the magnitude and duration of the external inflow are limited, the extent of real exchange rate appreciation may be small or even negligible, requiring no specific policy reaction. However, should the inflows become significant or persist for a relatively long period, domestic policy adjustments will be required either to prevent an erosion of or to restore competitiveness. Preventive action to safeguard the economy’s competitiveness in such circumstances requires early identification of disturbances (such as the foreign exchange inflow just described) so that they can be sterilized. In this case, low rates of domestic credit expansion relative to the growth in the demand for domestic cash balances can help to avert the erosion in competitiveness from the outset.

Such timely identification of a disturbance and its characteristics is unlikely, however, and in most instances the task of economic management will be to restore a lost margin of competitiveness. The goal of restoring competitiveness also calls for a policy of restraint in domestic credit expansion in relation to the growth in money demand in order to offset the price and exchange rate effects that the unanticipated external inflow has caused. As already discussed, the degree of actual restraint in domestic credit expansion will depend, among other things, on whether or not the policy mix includes an adjustment in the nominal exchange rate. While domestic credit and fiscal restraint are usually required in the presence of unanticipated capital inflows, they provide primarily short-run solutions. More fundamental changes in the economy are required, if the capital flows persist, to reestablish global balance. These changes must include some means of enhancing the economy’s adaptability, mainly through flexibility in the markets for and prices of factors of production, particularly those for labor.

Related areas where adverse consequences from unexpected foreign exchange surpluses may arise are output and production. The sequence of events that unfolds in an economy experiencing an unanticipated foreign exchange surplus varies, depending on the nature of the surplus as well as on the prevailing exchange rate regime. To the extent that the foreign exchange surplus influences domestic expenditure levels, domestic prices and the exchange rate will be affected, and with them output and activity. Conceptually, it is possible to eliminate effects on output by postulating the existence of symmetry in the behavior of domestic prices and costs—that is, symmetry in the sense that such prices and costs can rise or fall with the same ease or, in other words, that supply price elasticities are the same when prices are rising and when they are falling. This presumption of symmetrical behavior in prices and costs permits the argument to overlook impacts on production, because, under these conditions, although output varies with movements in prices and costs in one direction, such a variation is virtually offset by equivalent price and cost movements in the opposite direction, so that the equilibrium level of output is not affected. As noted earlier, however, such a presumption does not always hold, and there is evidence that downward price and cost rigidities prevail in most economies. Under those circumstances, a process that restores balance to the economy through reductions in the level or the rate of increase of prices and costs is likely to entail output losses. This fact has led some authors to advocate exchange rate flexibility.

In essence, the dilemmas discussed in this section are but reflections of the two basic perspectives from which exchange rate policy is typically analyzed, its macroeconomic or its microeconomic aspects. Views on exchange rate management that reflect concerns over maintaining competitiveness or over output developments are based on the belief that exchange rate policy (supported, to be sure, by other domestic economic policies) can and should be used actively to attain real objectives. The implication here is that real economic objectives can be attained and maintained using nominal policy instruments. In addition, these views imply that nominal exchange rate adjustments have real effects durable enough to make them worth pursuing. It is worth recalling that this line of reasoning is based on the assumption that domestic prices and costs are rigid in the downward direction.

The alternative view of exchange rate policy, though equally concerned with competitiveness and output developments, is based on the proposition that nominal exchange rate adjustments do not lead to sustained real economic changes. This is an extension to the exchange rate domain of a disbelief in the existence of money illusion.13 In other words, this view postulates that in circumstances requiring real economic adjustment (typically, a variation in output, employment, or real income), economic behavior that varies depending on the nominal policy instruments used cannot prevail for long. Consequently, proponents of this view argue that the best role for the exchange rate is to provide a nominal anchor for the economy as a means of attaining price stability or, more generally, keeping domestic price performance in line with the evolution of prices abroad. A development that gives credence to this view is the prevalence of wage indexing in inflationary settings. The potential for exchange rate adjustments to influence real variables is rapidly eroded in the presence of indexed wages, because the resulting real wage rigidity severs any link between exchange rate changes and competitiveness.14

V. The Issue of Convertibility

The subjects of foreign exchange management and exchange rate policy that I have been discussing so far are fundamental components of any strategy aimed at economic stabilization, adjustment, or reform. Their importance is, of course, even larger in the context of strategies aimed at all three. In the process, it has become clear that besides playing a critical role in economic policy in general, foreign exchange management and exchange rate policy are inextricably linked to monetary management, both conceptually and in practice.

Now I will turn to a related subject, currency convertibility, which bears heavily on issues of foreign exchange management and has significant implications for exchange rate policy15. I will discuss the subject from three perspectives: that of the formerly socialist economies, that of the market-based economies, and that of the International Monetary Fund, for which exchange rate policy and exchange arrangements are a central part of its mandate and responsibilities.

At the time of the IMF’s inception in the years just after World War II, full currency convertibility, important though it was, was not considered an urgent or immediate goal of economic policy. For well-known reasons, it was not until the late 1950s that the currencies of most industrial countries were given the status of convertibility. The reasons for the lengthy transition included the understandable focus on the immediate and urgent task of reconstruction, which took precedence over that of economic opening; the prevalence of extensive controls on and interference with international transactions, which required dismantling; and a belief in the importance and effectiveness of government policy, which tended to override the relevance of price signals as guides for economic behavior.

The first consideration, reconstruction, was clearly an urgent matter that had priority; economic policy, therefore, was aimed at and became subordinate to its attainment. The second consideration, lifting controls and restrictions on international transactions, was not only an endeavor that contributed to the task of reconstruction (in that it helped increase the efficiency of economies) but an important ingredient of exchange rate policy and a step toward currency convertibility. This was the period of the Bretton Woods par value system, when exchange rates, though adjustable, were fixed; therefore, the elimination of controls on exchange transactions was equivalent to the introduction of currency convertibility. The third consideration, belief in the effectiveness of government policy, may also have influenced the speed and scope of convertibility. It was clearly behind the decision to confine convertibility to a limited set of transactions (those on current accounts).

The IMF actively sought the elimination of trade and other current account restrictions, a goal enshrined in its Articles of Agreement. The focus on current transactions was appropriate at the time, if only because of their predominance in the balance of payments. The era of dominant capital flows had yet to arrive, and member obligations under the Articles of Agreement extended only to current account convertibility.

At present, it is clear that the reforming economies place a higher priority on the introduction of currency convertibility as an economic policy aim than did the industrial countries of the postwar period, in at least two respects: timing (most of the transitional economies plan to make their currencies convertible in a relatively short time); and scope (they make less of a distinction between, and an implicit or explicit sequence of, current and capital transactions). In addition, capital movements have been increasingly liberalized in industrial and developing countries, so that the scope of de facto convertibility has broadened significantly.16 In today’s global economic environment, the notion of establishing currency convertibility promptly and with respect to all transactions seems eminently reasonable, particularly for those economies in the throes of reform, for several reasons. First, the previously centrally planned economies are engaged in a process of deep economic, political, and social transformation with two underlying themes or aims. One is the curtailment of the dominant role of the state or the government; the other, related aim is the move toward a free market and away from the controls and restrictions characteristic of central planning. From the standpoint of these aims, the case for establishing full convertibility during the early stages of reform is certainly strong. A convertible currency takes away much of the scope for discretionary government action on exchange rate policy and the exchange regime. It also implies the removal of exchange restrictions and controls over the international transactions required to integrate these economies with the international system.

This said, however, the adoption of convertibility does carry with it a number of important prerequisites that link it to the other areas of reform. Typical preconditions for the establishment of a convertible currency, besides the adoption of an appropriate exchange rate level or regime, include sound macroeconomic policies—that is, a set of domestic financial policies which, by keeping the level and growth rate of demand in the economy in line with developments in its productive capacity, ensure a sustainable balance of payments position and help maintain domestic price stability. These are aims of the reform process itself, and, if attained, they will pave the way for currency convertibility. Thus, effective reform is important to convertibility, and convertibility is also important to reform, because an appropriate exchange rate and a sound currency constitute the flip side of a sustainable balance of payments position. In addition, the proper environment for convertibility allows market incentives to guide economic behavior. Creating such an environment entails liberalizing prices and opening the economy—actions that are also essential components of the reform process. A third important prerequisite for convertibility is the availability of an adequate stock of international reserves. These reserves not only will allow the economy to withstand policy or exogenous shocks that could threaten convertibility but will buttress confidence in the exchange regime in general and the permanence of convertibility in particular.

Generally speaking, convertibility is best contemplated after an economy has been stabilized, liberalized, and opened up, and after it has accumulated a healthy stock of international reserves. This, of course, is a tall order—so tall, in fact, that waiting for these conditions to materialize may be the best prescription for never introducing convertibility. But an argument can also be made that there is no need to wait for such stringent conditions to be met. Currency convertibility at an early stage may help induce, or may even force, the adoption of sound macroeconomic policies, the liberalization of prices, and the opening of the economy. Under such circumstances, it is also likely that international reserve earnings will accumulate to further support the convertibility action.

At stake here is an age-old controversy on how to approach the presence of a constraint: one view contends that a constraint should not be imposed or accepted until conditions are right. An alternative view asserts that, on the contrary, it should be imposed and accepted before appropriate conditions develop, if only because its mere presence will bring about those conditions. The matter is one of balance, of course; the right approach is neither to wait for all conditions to be completely right nor to move ahead when they are clearly impossible. With subjects such as convertibility, the controversy can be seen in a variety of lights. To accept convertibility early on is to remove exchange rates and monetary policy from the realm of politics; to wait for the right conditions, in contrast, is to opt for discretion in policymaking and risk making policy a prey to politics. On the other hand, haste may impair sustainability—hence the need for balance in this matter.

Convertibility is typically understood to mean freedom to use or exchange a currency at a given exchange rate. And indeed, this is the clearest concept. But the notion of convertibility has further dimensions that can render it compatible with other than fixed exchange rate regimes. The most comprehensive concept of convertibility involves several criteria. A currency’s convertibility implies that its holder may use it without restrictions for any purpose; it also implies that the currency can be exchanged freely for any other currency and that this exchange, as already noted, can always be made at a given exchange rate. Full convertibility obtains when these three criteria are met, but partial convertibility is also possible. In particular, different exchange rate regimes can be compatible with convertibility, although the broadest concept of convertibility operates with a fixed exchange rate.

In general, convertibility, interpreted as the ability of residents and nonresidents of a country to freely exchange domestic for foreign currency without limit, is consistent with any exchange regime that has exchange restrictions. This interpretation specifically includes a flexible exchange rate regime, which should, in principle, ease the introduction of convertibility, since conceptually such a regime needs no restrictions to support the exchange rate. In practice, however, few governments have been willing to give their currencies total freedom or to eschew the use of restrictions on international transactions. As is the case with the argument contending that exchange rate flexibility renders international reserves redundant—accurate in principle but unrealistic in practice—flexible exchange rates did not lead to convertibility because restrictions remained and did not become redundant.

VI. Conclusions

A variety of criteria can be used to justify the choice of a particular exchange rate regime. In this paper, I have argued that the process of making the decision is best served by following fundamental criteria that go beyond the purely technical characteristics of different exchange rate regimes. In this vein, my argument has been that the essence of the choice of a particular exchange arrangement lies in the national preferences it reflects.

One of these preferences corresponds to the decision to form part of an open system and the other to the perception that national economies, though interacting in an international system, derive more advantages by keeping themselves relatively closed. With a fixed exchange rate, an economy expresses its willingness to withstand the consequences of external shocks and disturbances and, correspondingly, to disseminate across the system at large the effects of its own domestic economic management. Under a fixed exchange rate regime, domestic policies become endogenous, and their effects are subject to leakages that substantially reduce the scope for divergence between national and international policies and objectives, if they do not eliminate it altogether. The large measure of economic interaction typical of open systems offers the prospect of enhancing individual component welfare, if on balance most governments pursue appropriate policies—that is, policies consistent with the constraints of interdependence. In such a setting, the costs of individual country policy errors are shared, as are the benefits of individual country policy efficiency. But these policies also raise the possibility of malfunction if countries do not focus sufficiently on the costs of their own policy shortcomings (because their dissemination across the system will tend to lower the incentive to avoid incurring such costs), concentrating instead on the benefits that can be derived from other countries’ adept management. Such moral hazard and free rider syndromes plague open systems constantly.

The other course of action corresponds to deciding whether there are advantages to organizing national economic behavior on the basis of pre-established norms, or rules, which exceed the benefits that can be derived from the primary use of discretion. Here again, preference for a rules-based system leads toward fixed exchange rate regimes. Indeed, a fixed exchange rate is in itself a rule that exacts a measure of domestic policy consistent with the policies prevailing elsewhere. Such consistency is both a feature and a requirement of open systems, in that it implies giving precedence to maintenance of the rule over those national interests that would conflict with it.

The choice of flexible exchange rate arrangements exhibits the opposite preferences. This choice goes in the direction of keeping national economies relatively contained, closed, and insulated within the system as a whole. Rather than being based on the willingness to share, flexible exchange rates are based on the general philosophy that exchange rate flexibility will provide a shield against outer disturbances and a boundary to contain the effects of domestic policies, which typically are perceived to exceed in quality those abroad. This particular choice also indicates a preference for a system based on discretion rather than on rules. Each economy is seen as free to pursue its own national objectives rather than as obligated to constrain them to the exigencies of interdependence. The essence of this choice, in contrast to that involving fixed exchange rates, is the belief that it leaves each economy independent to handle its own affairs. The belief, of course, presupposes that linkages are limited and that vulnerability to external shocks can be contained.

These considerations bring to the forefront the difference between the two types of regimes that is perhaps most often mentioned: domestic economy policy dependence (fixed exchange rate) versus domestic economic policy independence (flexible exchange rate). There is, of course, a measure of validity in arguments that link flexible exchange rates with policy independence and the ability to pursue national objectives separate or different from those pursued in the system as a whole. It is less clear, though, that such independence can endure.

Clearly a country that, on balance, adopts and implements better policies than its neighbors most of the time will be well advised to opt for a flexible exchange rate, which permits a relatively closed economic attitude or discretion in economic policy. But just as clearly, there will be an incentive for those countries that generally pursue less appropriate policies to opt for a fixed exchange rate and establish a direct link to the well-managed economy, based on a well-defined and transparent rule. Thus, even if economies that exhibit good policy performance strive for independence, their method of attaining it may not be straightforward.

If, on the contrary, the economy opting for discretion and independence generally pursues less appropriate policies than its peers (and the key question here is why it would choose this course of action, which reins in the consequences of policy errors), limits to such independence and discretion are likely to arise rapidly, created by resource flows from the mismanaged to more stable economies. In such circumstances, using a flexible exchange rate regime to attain policy independence is likely to prove more successful in theory than in reality.

What this line of reasoning hints at is that limits to independence and national discretion exist, and that, in general, the possibilities for closing off a national economy within an integrated environment like that of the global economy are quite narrow. The choice of a flexible exchange regime gives an economy scope to diverge from the rest of the system only up to a point, and by not too large a margin. Floating exchange rates do not give full independence, and total dependence does not characterize a fixed exchange rate regime.

An issue that often crops up in this connection is the size of the economy. It is generally argued that relatively small countries can hardly benefit from a flexible rate, since developments in their economy are typically dominated by events in large countries. There is some truth in this reasoning, and to that extent it could be contended that flexible exchange rates provide little, if any, advantage to many developing and reforming economies. Of course, this line of reasoning presumes that, in general, the large economies display a relatively stable policy environment. This said, size, like the nature of the shocks, is a technical criterion unlikely to suffice as the basis for choosing an exchange rate regime. Rather, such technical criteria need to be accompanied by broader factors, such as the quality of economic management and the prevailing attitude toward open systems, in order to ensure an efficient choice.


Anthony Lanyi

Manuel Guitián’s paper reminded me of the opening sentences of Tolstoy’s novel Anna Karenina, which begins this way: “All happy families are alike. Each unhappy family is unhappy in its own way.” For indeed, Mr. Guitián’s message is that there are two kinds of countries: the happy ones, which have fixed exchange rates and convertibility, open economies, and stable macroeconomic policies; and the unhappy ones, which have flexible exchange rates, are insulated from the international economy, and choose their own paths—more likely than not, those of financial ill-discipline and restriction.

Hindsight has taught me, however, that there may be just as many similarities between unhappy families as there are between happy families. In this sense, I have doubts about the linkage Mr. Guitián suggests between fixed exchange rates and openness and between flexible exchange rates and “closedness.” I am not sure whether positing such a linkage is a normative or empirical statement. Yet it is interesting to consider what sort of correspondence exists between exchange arrangements and a measure of openness and “closedness.”

Included in my discussion are all the IMF member countries, both developed and developing. My table shows the member countries (except for the 15 new members from the former Soviet Union) grouped by types of exchange rate regimes, as classified by the IMF. “Fixed” or “pegged” exchange rates are linked to either a single currency, the SDR, or a composite basket of currencies. “Limited flexibility” refers either to countries that participate in the European exchange rate mechanism (ERM) or to some states in the Middle East that peg their currencies to the U.S. dollar but allow for a wider range of flexibility. The “more flexible” group includes those countries with a managed float and those that adjust the value of their currency according to a set of indicators. And finally, there is the group of so-called “independently floating” countries.

In the left-hand column are two categories. One includes the countries that have accepted Sections 2,3, and 4 of Article VIII of the IMF Articles of Agreement—that is, the commitments to maintain a current account regime free of restrictions, to avoid multiple currency practices, and to ensure “convertibility” in the sense the Articles suggest which really means convertibility for current account transactions. However, I agree entirely with Mr, Guitián that the distinction between current account convertibility and capital account convertibility, which may have been meaningful in the world of 1946, is not very meaningful in today’s world of computers and internationally integrated markets.

Exchange Rate Arrangements1


of total


of all



of total


of all



of total


of all



of total


of all

Article XIV5263.432.700.00.01866.711.31645.710.186
Article VIII3036.618.915100.09.4933.35.71954.311.973
Sources: IMFrInternational Financial Statics (September 1992), and IMF Secretary’s Department

Excluding Cambodia, for which no current information is available

ERW, flexible single-currency peg

Managed float; adjusted to indicators

Sources: IMFrInternational Financial Statics (September 1992), and IMF Secretary’s Department

Excluding Cambodia, for which no current information is available

ERW, flexible single-currency peg

Managed float; adjusted to indicators

The other category of countries, the Article XIV countries, comprises those that have not yet dared make the commitment to Sections 2, 3, and 4 of Article VIII. They are not sure whether they want to keep their hands off various kinds of controls over current transactions or to commit themselves to never indulging in multiple exchange rates.

The table shows that there is not necessarily a clear relationship between fixed rates and open economies. There are many economies—and have been for many years—that use current account restrictions and import controls and that nevertheless find it convenient or useful to have fixed exchange rates. They do not dare open up their economies to convertible regimes. At the same time, many countries (and not just developed countries) have flexible rates and are nevertheless open economies in a meaningful sense: they have convertibility and do not restrict current or, in many cases, capital transactions. There are no apparent reasons why these countries prefer to have flexible exchange rates. (Milton Friedman’s argument for flexible exchange rates, written back in the early 1950s, includes the thought that flexible exchange rates preclude the need for import controls and restrictions. Perhaps this argument is somewhat dated in terms of liberalizing current and capital accounts, but it is still an interesting consideration.)

There is, of course, another factor at work in exchange rate systems. The major players in the system are floating their currencies against each other, and it is very hard to imagine a world in which Germany, Japan, and the United States would commit themselves to rates fixed against each other. Such a world existed, of course, before 1971, but in today’s circumstances, such a world would imply a kind of coordination of the economic policies of different countries that would be difficult to define (in terms of achieving an optimal result) and also, for political reasons, to implement.

My table suggests that most of the countries with fixed rates are, in fact, Article XIV, not Article VIII countries. Why have most Article XIV countries chosen fixed rates? Part of the reason, perhaps, lies in the conflict between different criteria. Mr. Guitián has mentioned two criteria: first, the need to achieve internal and external balance, and second, the desire to encourage international trade and investment through a stable exchange rate. Over the last three decades, a number of the former colonial countries in Africa and South Asia that had heavily controlled trade and exchange regimes have chosen to fix their rates against the currency of either the former mother country, another large country, or a basket of currencies to assure stability in foreign trade and encourage the participation of investors from industrial countries. Another reason for the fixed rates may lie in the need to adapt to an international environment within which major currencies float against each other, world economic conditions are changeable, and countries want the freedom to adjust the value of their currencies when an unexpected shock occurs. If its currency is pegged to a composite, countries may prefer to describe themselves as “managed floaters” in order not to tie themselves down vis-à-vis the IMF or vis-à-vis its own residents, but the exchange regime may, in fact, be relatively stable.

In terms of the “technical” aspects of the choice of an exchange rate regime, there is, of course, a spectrum of regimes to choose from, ranging from fixed to completely market-determined floats. Countries may have a very clear idea where in the spectrum they would like to be, but even within each part of the spectrum important choices have to be made—for instance, between pegging and following some kind of managed flexibility with a rule. Some countries have considered these choices fairly important, for reasons particular to their economies or foreign trade situations.

Peter Wickham

Manuel Guitián’s paper on the choice of an exchange rate regime provides much food for thought. My remarks will focus on parts of the paper that carry the arguments somewhat further than I would like.

Mr. Guitián’s theme concerns two elements affecting the nature of the choice involved. The first is the relationship of national economies to the global system, the second the degree of activism envisaged for domestic economic policies. The paper characterizes a fixed exchange rate arrangement as reflecting a decision to be open to the global system by accepting an international constraint on national economic policies. This regime is seen as limiting the range or mix of domestic economic policies a country can adopt. On the other hand, the choice of a flexible exchange rate regime, at least in principle, reflects the desire for a more closed economic system with no international constraints on the adoption of domestic economic policies.

This argument has deep roots in the literature on the merits of fixed versus flexible exchange rates. In 1953, Milton Friedman argued strongly in favor of flexible exchange rates, which he said would allow each country to achieve “unrestricted free trade and the freedom… to pursue internal stability after its own lights.”1 An outgrowth of this argument was that flexible exchange rates had insulating capabilities, whereas (as Mr. Guitián argues) fixed rates implied that the effects of internal and external policy impacts and shocks would be disseminated across national boundaries. The question I wish to pose asks how far this characterization can be pushed. Countries as diverse as Canada, Bolivia, and The Gambia have floating exchange rates. Austria, Botswana, and Malaysia operate under pegged exchange rate arrangements. To what extent do these choices reflect the desire to remain an open or a closed system, to follow independent policies, or to insulate an economy from certain shocks?

I would agree with Mr. Guitián that the choice of an exchange rate regime must go beyond purely technical criteria to issues of a more normative nature, such as a government’s reputation, credibility, and commitment. I am less comfortable with the characterization of the choice of regime as one involving preferences for openness and policy activism and would argue, as Mr. Guitián in fact does in his closing remarks, that the scope to indulge in insular preferences has narrowed over time and may be rather restricted in today’s international environment. It has not, however, disappeared altogether.

Another way of viewing the difference between a floating and fixed exchange rate regime is to note that in the former, the money market clears through the float with an exogenous money supply, while in the latter, the market clears through an exogenous exchange rate with a fluctuating money supply. So in one sense we can view the choice of a floating rate as a vote for monetary “independence.” But what does monetary or, more generally, financial policy independence and exchange rate flexibility buy, and does it also involve costs?

As I have already indicated, much of the early debate on this issue was based on the idea that floating rates could insulate an economy from certain outside disturbances, so that monetary policy could be directed toward short-run stabilization. However, subsequent analysis, which embraced the idea of closely integrated markets for financial capital, tended to undermine this idea. Over time, a standard classification emerged on the comparative advantage of fixed and floating exchange rates in dealing with various types of disturbances, but as Mr. Guitián suggests in his paper, it is difficult to apply the analysis operationally. For optimal exchange rate intervention, the authorities would need to have some idea of the size and frequency of various shocks (including covariance properties) and would have to extract information from movements in observable variables. But phenomena such as exchange rate overshooting and the poor track record of models in explaining exchange rate movements during the post-Bretton Woods period make the possibility of acquiring such information unlikely.

If the picture with regard to short-run stabilization remains controversial, other ramifications of exchange rate regimes are clearer. A floating or flexible exchange rate does provide for longer-run monetary autonomy and domestic price management. In a country with a floating rate, inflation is determined by domestic monetary policy. In a country with a fixed rate, however, inflation depends on the monetary policies of those countries sharing the common standard. In other words, a floating rate is arguably compatible with complete monetary autonomy, but a fixed exchange rate is not, since it depends on some form of explicit or implicit monetary harmonization with, or subordination to, foreign constraints on economic policy. In this limited sense, I am in agreement with Mr. Guitián’s characterization of the nature of the two types of regimes.

But how can, and how should, the monetary autonomy a floating rate offers be exercised? In some cases, such autonomy has been exercised in ways that can be considered only as less than optimal or economically rational. At least one explanation of why the Bretton Woods system broke down and the major industrial countries began to float is that the United States, a key player in the cooperative game, began in the 1960s to pursue an expansionary monetary policy. This policy, which spilled over internationally, was at odds with the inflationary preferences of other important players. The float allowed Germany, for example, to close itself off and reassert its monetary sovereignty and control over the domestic inflation rate. On the other hand, Latin America provides numerous instances of governments adopting monetary policies (often closely linked to fiscal inadequacies) and generating rates of inflation totally inconsistent with any degree of stability in their exchange rates.

In light of the above, can it be argued that some countries would be better off subordinating control of a major policy instrument to external discipline by pegging their exchange rates? One problem with this solution is that a government too weak to follow a monetary policy capable of achieving a reasonable rate of inflation is likely to find it very difficult to submit itself to the monetary and fiscal discipline needed to maintain a fixed peg. Thus, even if the benefits of adopting a fixed peg are clear, the authorities may not be able to commit themselves credibly to it. This factor has led to difficulties in a number of countries that have attempted to bring down their inflation rates after inflationary episodes by pegging their exchange rates, a move that in effect sets up the exchange rate as a nominal anchor in stabilization programs. In such cases, the central bank lacks credibility, and it is often very difficult to convince the public that the authorities will not renege on their commitment to the new peg. This lack of trust is often reflected in high nominal (and hence real) interest rates, which can then depress investment and make fiscal adjustment more difficult. A further complicating factor is that inflationary inertia and the difficulties of changing inflationary expectations may make convergence of the domestic inflation rate with one consistent with the peg a somewhat protracted affair. The tendency for the real exchange rate to appreciate in such circumstances tends to further threaten the peg’s credibility. A number of transition economies have faced similar dilemmas during stabilization episodes in the aftermath of price liberalization. Poland, for example, was able (aided by a stabilization fund and a significant devaluation) to use the exchange rate as a nominal anchor in the first stages of its stabilization and reform program. When inflation convergence remained elusive, the authorities moved to a crawling peg. Romania and Bulgaria, on the other hand, had neither high levels of foreign reserves nor a stabilization fund. Given the difficulties of establishing financial discipline in a reform environment, the governments did not view a fixed exchange rate peg as a viable option and so instituted floats instead.

A considerable number of countries continue to opt for some form of flexible exchange rate arrangement, suggesting that the requirements of inflation convergence and the loss of monetary sovereignty associated with a fixed peg are politically very difficult to accept. But how the issues I have been discussing interact with and are affected by the exchange regime is an area worthy of further analysis, particularly in light of the recent turmoil in currency markets.

Summary of Discussion

The discussion of Manuel Guitián’s presentation centered on three main issues: the choice between a fixed nominal exchange rate regime and a regime that allows the nominal rate to be adjusted for inflation in order to maintain an unchanged real rate; considerations that would favor a flexible rate system; and the impact of destabilizing capital flows on the exchange rate. Opinion was divided on the choice between maintaining a fixed nominal or real rate. Since the equilibrium real rate is generally unknown, and since a number of participants were reluctant to endorse a policy of using the nominal exchange rate to obtain a real exchange rate objective, there was substantial support for a fixed nominal rate. On the other hand, it was thought likely that in the presence of large internal imbalances and minimal foreign exchange reserves, a fixed nominal rate would have to be changed soon after the initial peg, touching off expectations of further changes. In these circumstances, several speakers favored a fixed real rate.

Some viewed the exchange rate as a price that, like any other price, should be free to adjust to shifts in supply and demand. Proponents of this view regarded a fixed rate as interference with the market’s rationing function. Opinion was again divided on when to use a floating arrangement: some felt it was best suited to situations in which domestic imbalances were large, but others thought a flexible rate without appropriate supporting domestic policies would lead to accelerating inflation and a vicious circle.

A related issue was the choice of a hard currency peg in a fixed rate arrangement. The consensus was that the appropriate choice would be the major currency most widely used in trade by the pegging country. It was also pointed out that two escape valves were available to countries choosing fixed nominal rates: temporary restrictions on current payments and the possibility of using IMF resources. Even if the fixed rate was not set at the equilibrium value, a country could, under some circumstances, avoid a devaluation.

Discussing the problem of destabilizing capital flows, participants noted that a country would probably not be able to defend a fixed parity in the presence of speculative capital flows without some assistance from other countries. The exchange rate mechanism (ERM) of the European Monetary System was cited as providing a good example of coordinated action (in this case, by France and Germany) to defend a currency (the franc) against a speculative attack by private capital. The central banks of France and Germany were able to handle the volume of private capital flows and maintain the currency alignment, despite the fact that monetary policies diverged among the ERM countries. Regarding the liberalization of the capital account in developing and transition economies, it was generally agreed that if opening the capital account would lead to a large outflow, it would be advisable to liberalize the capital account gradually.


    AghevliBijanB.MoshinS.Khan and PeterJ.Montiel.1991. Exchange Rate Policy in Developing Countries: Some Analytical Issues.IMF Occasional Paper No. 78. Washington, D.C.: International Monetary Fund.

    ArgyVictor and PaulDe Grauweeds.1990. Choosing an Exchange Rate Regime: The Challenge for Smaller Industrial Countries.Washington, D.C.: International Monetary Fund.

    BalassaBela.1987. “Effects of Exchange Rate Changes in Developing Countries.”The Indian Journal of Economics68 (October): 20321.

    CordenW. Max.1990. “Exchange Rate Policy in Developing Countries.”World Bank Working Paper WPS 412. Washington, D.C.: The World Bank.

    FloodRobert P. and Nancy P.Marion.1991. “Exchange Rate Regime Choice.”IMF Working Paper 91/90.Washington, D.C.: International Monetary Fund.

    FrankelJeffreyA.1992. “Monetary Regime Choices for a Semi-Open Economy.”Paper presented to a conference on Monetary Policy in Semi-Open Economies inSeoul, KoreaNovember 6–71992.

    FrenkelJacobA.MorrisGoldstein and PaulR.Masson.1991. Characteristics of a Successful Exchange Rate System. IMF Occasional Paper No. 82. Washington, D.C.: International Monetary Fund.

    FriedmanMilton.1953. Essays in Positive Economics.Chicago: University of Chicago Press.

    GilmanMartinG.1990. “Heading for Currency Convertibility.”Finance and Development27 (September): 3234.

    GreeneJoshuaE. and PeterIsard.1991. Currency Convertibility and the Transformation of Centrally Planned Economies. IMF Occasional Paper 81. Washington, D.C.: International Monetary Fund.

    GuitiánManuel.1973. “Credit Versus Money as an Instrument of Control.”IMF Staff Papers 20 (November): 785800.

    GuitiánManuel.1976. “The Effects of Changes in the Exchange Rate on Output, Prices, and the Balance of Payments.”Journal of International Economics6 (February): 6574.

    GuitiánManuel.1988. “The European Monetary System: A Balance between Rules and Discretion.”In Policy Coordination in the European Monetary System IMF Occasional Paper No. 61.Washington, D.C.: International Monetary Fund.

    GuitiánManuel.1992a. “Monetary and Exchange Rate Policies for Stabilization and Reform.”Paper presented to a seminar on Transition and Adjustment at the Joint Vienna InstituteOctober5–81992.

    GuitiánManuel.1992b. “Capital Account Liberalization: Bringing Policy in Line with Reality.”Paper presented to a conference on Monetary Policy in Semi-Open Economies inSeoul, KoreaNovember6–71992.

    GuitiánManuel.1992c. Rules and Discretion in International Economic Policy. IMF Occasional Paper No. 97. Washington, D.C.: International Monetary Fund.

    International Monetary Fund. 1992. Developments in International Exchange and Payments Systems.World Economic and Financial Surveys. Washington, D.C.

    JohnsonHarryG.1973. Further Essays in Monetary Economics.CambridgeMass.: Harvard University Press.

    KenenPeterB.1969. ‘The Theory of Optimal Currency Areas: An Eclectic View.” InMonetary Problems of the International Economyedited by RobertA.Mundell and AlexanderK.Svoboda.Chicago: University of Chicago Press.

    McKinnonRonaldI.1963. “Optimum Currency Areas.”American Economic Review51 (September): 71725.

    MundellRobertA.1961. “A Theory of Optimal Currency Areas.”American Economic Review51 (September): 65764.

    MundellRobertA.1968. International Economics.New York: MacMillan Company.

    PolakJacquesJ.1991. “Convertibility: An Indispensable Element in the Transition Process in Eastern Europe.” InCurrency Convertibility in Eastern Europeedited by JohnWilliamson.Washington, D.C.: Institute for International Economics.

    WilliamsonJohned.1991. Currency Convertibility in Eastern Europe.Washington, D.C.: Institute for International Economics.

Note: In preparing these remarks, I have relied extensively on a large amount of research and policy work in the exchange rate management area conducted by the IMF over a long period of time. And with regard to various specific sections in this paper, I have drawn on a previous lecture given last October at the Joint Vienna Institute; see Guitìán (1992a).

For example, this is the case in much of the work published by the International Monetary Fund: see, among others, (Aghevli, Khan, and Montieì 1991); Frenkel, Goldstein, and Masson (1991); Corden (1990); and Balassa (1987).

For an enlightening discussion of this subject, see Friedman (1953). In presenting what is generally acknowledged as a classical statement of the case for flexible exchange rates, Friedman lists arguments that view the exchange rate as an indicator of inflation and that consider this variable a constraint on a government’s ability to undertake inflationary actions. Friedman, of course, proceeds to present a case against this type of reasoning, but there can be no doubt that the logic of such lines of thought is equivalent to what lies behind modern discussions of exchange rate anchors and hard currency policy options: see also, in this general context, Johnson (1973, chapter 8).

For an excellent analysis of this subject, see Mundell (1968, chapter 11).

For a recent discussion of the consequences for economic policy of the predominance of capital flows, see Guitián (1992b) and Corden (1990). And a discussion of the relevance of porous national boundaries for the design and implementation of economic policy can be found in Guitián (1992c, section II).

The analogy carries in the following sense: a possible definition of a closed economy is one in which national income and expenditure, ex post, must coincide. A flexible exchange rate, if not tampered with and in the absence of capital flows, will ensure that balance obtains in the current account of the balance of payments, rendering national expenditure and income equal; see Guitián (1973).

For fuller discussions, see Aghevli, Khan, and Montiel (1991); Flood and Marion (1991); and Frankel (1992). A collection of interesting articles discussing technical and empirical aspects of the choice of exchange rate regime will be found in Argy and De Grauwe (1990).

See Friedman (1953) for an incisive analysis of these issues. I will point out here that the justification for using a downwardly rigid model of prices is that, to quote Friedman, if “internal prices were as flexible as exchange rates, it would make little economic difference whether adjustments were brought about by changes in exchange rates or by equivalent changes in internal prices. But this condition is clearly not fulfilled” (p. 165).

It has been argued that since both nominal and real shocks afflict all economies, the best choice is an exchange rate regime that exhibits an intermediate degree of flexibility (see Flood and Marion (1991)). I am not persuaded by this reasoning, which implies that average suboptimal responses to both types of shocks are preferable to optimal responses to one type of disturbance combined with subpar suboptimal responses to the other type of shock. This is, of course, an empirical question.

The classic article on the subject of optimum currency area is by Mundell (1961), and the analysis was subsequently extended by McKinnon (1963) and Kenen (1969); see, for a brief summary of the various intellectual strands, Guitián (1988).

For example, goods, factors of production, market openness, and commonality of aims—as well as policy—are behind the establishment of the exchange rate mechanism (ERM) of the European Monetary System (EMS). Absence of factors of production—in particular, labor mobility—and different policies and aims are behind the independent float among Europe, Japan, and the United States.

A point worth noting here is that the economic consequences of corrective policy packages that exclude exchange rate adjustments are not likely to coincide with those that include them. There is, of course, an empirical question involved here, which has already been discussed earlier (see the quotation from Milton Friedman in note 7 above).

Correction of an imbalance with no exchange rate adjustment requires, other things being equal, a relatively more stringent domestic policy stance than is called for when the strategy includes such an adjustment. How durable this trade-off is, though, depends on how often it is used. Acceptance of frequent devaluations to relax the degree of policy restrictiveness will erode the credibility of the strategy, tending to eliminate the trade-off or render it less favorable; see Guitián (1976).

This point has been insightfully and compactly made by Mundell in his “Monetary Dynamics of International Adjustment under Fixed and Flexible Exchange Rates” (1968, Chapter 11), as follows: “The argument is based on money illusion: the community is unwilling to accept variations in real income through changes in money prices, but it will accept the same changes in real income through adjustments in the exchange rate” (p. 152).

The consequences of wage indexation for the effectiveness of exchange rate changes have been amply discussed by Corden (1990).

Currency convertibility is a subject that has recently regained widespread interest, particularly in the context of reforming economies. Analyses of convertibility will be found in Gilman (1990), Greene and Isard (1991), and Williamson (1991); but see also Polak (1991).

For an extensive discussion of developments in international capital flows and their consequences for economic policies, see Guitián (1992b); see also Corden (1990).

Milton Friedman, “The Case for Flexible Exchange Rates,” in Essays in Positive Economics (Chicago: University of Chicago Press, 1953), pp. 157–203.

    Other Resources Citing This Publication