Book
Share
IMF History (1972-1978) Volume 3
Chapter

Indicators of Need for Balance of Payments Adjustment

Author(s):
International Monetary Fund
Published Date:
February 1996
Share
  • ShareShare
Show Summary Details

In January 1973 the Research Department prepared the following paper on the use of reserves and of basic balances as possible indicators of the need for balance of payments adjustment. The paper was circulated to Executive Directors and was made available to the Committee of Twenty.

Reserves and Basic Balances as Possible Indicators of the Need for Payments Adjustment

(January 10, 1973)

Introduction

It has been suggested that the reformed international monetary system should utilize objective indicators to promote prompter and more effective adjustment of external disequilibria. One specific proposal, which was advanced by the United States in the Committee of Twenty, envisages international agreement on a “base level” of reserves for each country, which would be bounded by upper and lower “warning points” that would ordinarily signal the need for adjustment and, at a greater distance from the base level, by an “outer point” and a “low point” that would raise the possibility of sanctions being applied as a penalty for the failure to adjust. Another suggestion is to employ the cyclically adjusted basic balance as an indicator of the need for payments adjustment. The purpose of the present paper is to discuss the possible uses of indicators in promoting the adjustment process, with special reference to these two proposals.

In the past the term “indicators” has been used to refer to statistical series that might be used to establish an obligation or presumption that some adjustment action (often a change in par value) should be undertaken. However, the fact that the cyclically adjusted basic balance has been advanced as a possible indicator demonstrates that the term is now being used in a wider sense to include statistical estimates that involve specific adjustments reflecting hypotheses about the operation of the economic system. It is difficult to see that any useful purpose would be served by broadening the meaning of the term even further, such as would be necessary if one were to present as an indicator the results of an elaborate assessment of the possible developments of a country’s balance of payments; such a course might also create semantic confusion. In the present paper, therefore, “indicators” will be defined as statistical series relating to past events that have been subjected to statistical processing of a precisely definable kind.

There are a number of roles in which indicators could be used. First, they might be used to provide a test of the need for adjustment. There is scope for considerable variation in the automaticity of the consequences that would flow from the triggering of an indicator employed in this role. At one extreme, the triggering of the indicator might establish an obligation to undertake a particular adjustment action, such as a change in par value. At the other extreme, the triggering might initiate analysis and discussion of the need for adjustment, without in itself creating any presumption that a specific course of action, or indeed any action, would be warranted. In between, the triggering might establish a presumption that adjustment action should be undertaken, but with the possibility that the indicator might be overridden for sufficient reason: in this event the triggering would shift the onus of proof.

A second possible role is that of allocating the distribution of the responsibility for initiating adjustment actions between countries. In most cases, a conclusion that adjustment is needed to remedy a particular country’s payments problem will be sufficient to establish an obligation for that country to initiate corrective action. However, payments deficits have corresponding surpluses elsewhere; and indicators may help to detect the need for adjustment on both sides of any major imbalance, although not the relative size of the adjustment.

A third possible role is in relation to sanctions as an incentive for countries to follow appropriate adjustment policies and a penalty for failure to adjust. One possibility would be that sanctions might be triggered by a particular indicator. Another approach would be to make the imposition of sanctions consequential on a failure to follow the adjustment policies recommended by the international community in the light of an assessment that was itself triggered by an indicator. The presumptive character of an indicator could be reflected in the automatic imposition of sanctions which could, however, be lifted if assessment led to the conclusion that sanctions were not warranted.

Relations Between Reserve and Basic Balance Indicators

A reserve level indicator and a basic balance indicator differ in three respects:

(1) A reserve level indicator relates to the stock of reserves, whereas the basic balance indicator concerns a flow. However, reserve statistics can also be used to construct a flow indicator, based on changes in reserves: indeed, the U.S. paper refers to reserves both in terms of stocks and flows.

(2) A reserve indicator (whether stock or flow) would be influenced by movements of private short-term capital, whereas the basic balance indicator would attempt to exclude them. (A reserve indicator would also be influenced by official short-term capital movements—reserve borrowing and switches into and out of reserve currencies—if the relevant reserve concept was a gross one, but both a net reserve indicator and a basic balance indicator would be immune to such movements.)

(3) It is envisaged that the basic balance indicator would be adjusted for cyclical and incidental factors, whereas there is no adequate basis for the cyclical adjustment of reserves.

Payments Objectives

The major test of the usefulness of indicators must be the extent to which they are able to provide a measure of present or impending departure of countries’ external positions from an appropriate set of objectives. This immediately raises the question of the nature of the payments objectives that countries should pursue.

Payments objectives have both a flow and a stock dimension. It is generally agreed that flow disequilibria in the balance of payments of a more than temporary character should be avoided. Such disequilibria imply the domestic absorption in investment and consumption by deficit countries of resources from other countries beyond amounts made available by autonomous capital movements and aid flows.

Stock equilibrium is also important. Countries need reserves in order to avoid adjustment to transitory disturbances, to enlarge their freedom of action in dealing with underlying disturbances, and to bridge the time which it takes for corrective measures to have their effects on the payments balance. From these needs may be derived an optimal distribution of any given stock of world reserves at which countries should aim. Since, however, it is of the essence of reserves that their level should fluctuate, an overrigid pursuit of such reserve targets would defeat the very object for which reserves are held. However, it is equally true that reserves will not approximate their target level over time by chance: this will occur only if a reserve target is backed up by a willingness to take corrective action if reserves diverge substantially from their target level as a result of nonreversible factors.

The stock and flow dimensions are intimately interrelated, since a flow disequilibrium is the means by which a stock is changed. Persistent disequilibrium in the balance of payments will therefore cause a maldistribution of reserves, and a severely abnormal reserve situation is often a sign (although it may be a late sign) that persistent disequilibria exist or at least have existed in the past. One can argue that the problem of maintaining reasonable balance in the payments flows is the more important objective, since if persistent imbalances are not corrected they will lead to an increasingly unsatisfactory situation. By contrast, substantial imbalance in reserves, while not optimal, can persist for long periods without necessarily causing serious problems for the functioning of the system if adjustment to imbalances is sufficiently prompt. In that case the main effect of a maldistribution of reserves may be that low-reserve countries are more frequently obliged to use conditional credit from the Fund, or other balance of payments credits, than countries with abnormally high reserves.

There is reason to limit the speed with which deviations of reserve stocks from target levels should be corrected by adjustments which involve imbalances in payments flows. This rests upon the fact that the costs of payments adjustment (i.e., of changing certain types of payments flows) are of particular importance in the current account. Even under the most favorable conditions of price flexibility, or when payments adjustment is effected by exchange rate changes, there will be frictional losses involved in redirecting output between markets, and usually in reallocating resources to change the mix of output as well. If demand management policy is used to engineer the adjustment, the costs of price inflation or output deflation may be superimposed on top of these frictional costs. These costs mean that, even if the stock of reserves is far from its optimal level, it is desirable to allow adequate time for reconstitution, so as to prevent the need for precipitate adjustment policies both at the time when the need for reconstitution is first recognized and at the subsequent time when it has been completed.

These considerations argue in favor of the correction of any maldistribution of reserves to a considerable extent by measures influencing the capital account, including reserve borrowing and lending. Such measures will be the more successful if they are undertaken both by countries with abnormally low reserves and by countries with abnormally high reserves. How far it would be advisable to adjust reserve stocks in this way depends on the costs that may be incurred in influencing the capital account, in the form of the inefficiencies that result from controls or the distortion of interest rates and savings rates; moreover, countries are often reluctant or unable to resort to reserve borrowing on the scale required to restore adequate reserve levels. To the extent that it is necessary to have resort to current account adjustment to correct inadequate or excessive reserve levels, any such adjustment should be spread over a reasonable period in order to minimize adjustment costs.

The above considerations suggest that both stock and flow dimensions are important elements of a country’s external position. Since a country’s stock position can only be changed as a result of the flows over time, the issue of immediate relevance in formulating a payments objective at a particular point in time is that of selecting a flow objective. If reserves are close to target level and the normal reserve increase is taken care of by SDR allocations, a country’s flow objective should be approximately zero. However, where reserves are at some distance from the desired level and circumstances do not justify an adequate correction by a once-and-for-all change in the capital account, it is appropriate to modify the flow target so as to bring reserves to a more satisfactory level over several years. A deviation from the zero flow target could also be appropriate to reduce outstanding balance of payments financing or a country’s foreign assets that were not in the form of reserves.

It would not be sensible to seek balance of payments equilibrium over every short period of time. It would be wasteful to accept the real costs of adjusting the current account in response to factors that are believed to be volatile, transitory, and reversible. Reserves are held for the purpose of enabling countries to avoid such unnecessary adjustments, a task in which they may be assisted by exchange rate movements within sufficiently wide margins around par. It follows that the appropriate flow target is one that is related to the underlying and persistent flows rather than one that includes transitory and reversible elements, such as flows of volatile capital, the effects of discrepancies in the cycle between different countries, or the consequences of strikes or abnormal supply or demand conditions.

The Diagnostic Role

The role traditionally envisaged for indicators is that of assisting diagnosis of the need for adjustment to realize the above-specified payments objectives. In view of the fact that sudden and large adjustments are difficult and disruptive, it is desirable to diagnose emergent disequilibria as early as possible so as to facilitate prompt adjustment. Unfortunately, there is an inherent conflict between maximizing the probability of securing a sufficiently early diagnosis to enable a country to take nondisruptive adjustment action and minimizing the probability of generating false signals. A good indicator is one that will give the best possible trade-off between the two.

1. Indicators of flow disequilibria

One broad indicator of payments flows would be provided by changes in reserves. Its limitation is that, especially in the short run, reserve changes can be very heavily influenced by transitory and reversible flows that should be financed rather than be taken as signifying a need for adjustment. Except insofar as reserve borrowing occurs in connection with such flows, net rather than gross reserves would be the relevant series.

The cyclically adjusted basic balance attempts to exclude flows of short-term capital; the purpose of cyclical adjustment is to extract that part of the basic balance that reflects the current cyclical situation, and that can therefore be expected to be reversed as the cycle develops. Unfortunately, it is far from clear that these statistical adjustments can be made in a way that will extract the transient and reversible parts of payments flows and leave only the underlying and persistent components. The art of cyclically adjusting trade flows is still highly imperfect, and no regular cyclical adjustment is being made either of the invisible or of the long-term capital account. Furthermore, the distinction between long-term and short-term capital has been increasingly eroded and there has never been any compelling reason for expecting volatile funds to be entirely invested in short-term forms or for regarding all flows of short-term capital as potentially reversible.1

Apart from the difficulties of cyclical adjustment and the separation of underlying from reversible capital flows, the basic balance suffers from certain inherent limitations. First, it does not include the effects of past cost and exchange rate changes that are already in the pipeline. Second, it makes no attempt to look forward to predict the effects of probable future trends in competitiveness. These two factors mean that it does not represent the best forecast of the future evolution of the balance of payments that it is possible to make. Of course, it makes no allowance for any need to adjust the reserve stock to its target level.

By their nature, reserve stocks are likely to be relatively slow as indicators of flow disequilibria. They will be carried to extreme values only by flow disequilibria that persist for substantial periods, assuming the trigger points to have been set sufficiently far apart to limit the likelihood of false signals. An indicator based on reserve levels could therefore provide substantial scope for balance of payments disequilibria to become established. For example, a country with a cyclical pattern in its reserve holdings would need to have the trigger points placed at levels that would not bring them into operation by normal cyclical developments. This might, however, mean that these points were unable to identify an emergent disequilibrium whose effect initially ran counter to cyclical developments until the cycle turned, by which time the disequilibrium might be firmly established and difficult to remove.

Thus, if other indicators or periodic assessment were able to spot flow disequilibria at an early stage, reserve stocks might have little function to perform in this connection. If the performance of alternative guides to adjustment was less adequate, a reserve level indicator could at least function as a “long stop” in the sense that it would be sure to be triggered if disequilibria continued uncorrected. However, it might also be triggered, or remain outside the trigger level, even if the underlying disequilibrium had ceased.

One question that would need further consideration in this context is that of the relative merits of gross or net reserves.

2. An indicator of stock disequilibrium

Reserve levels can be used as measures of the position of countries with respect to certain agreed normative criteria for the distribution of reserves. The U.S. paper suggested that a “base level” for countries’ reserves might be defined on the basis of Fund quotas or past actual reserve holdings, which the Fund could modify where it found that the results of such calculations would be inappropriate. Data on past payments variability and other information might be used to determine grounds for modifications of the results of any formula.

The trigger points would need to be placed sufficiently far from a “base level” to ensure that the policy changes needed to prevent their being breached were not so frequent or so severe as to impose costs on the country that exceeded the benefits to the rest of the world from the greater promptness in adjustment. In particular, it would vitiate the basic purpose of holding reserves if they could not fluctuate sufficiently to enable countries to avoid the need to adjust to cyclical and other transitory disturbances.

The system already contains an implicit “low point” in the form of zero (gross) reserves. It might be argued that it would be irrational to raise this low point—except, possibly, by substituting net reserves for gross reserves—because that would oblige a country to hold a certain volume of reserves for purely ornamental purposes. (It is true that countries already aim to hold a positive minimum reserve level in order to preserve confidence, but it would seem likely that a positive trigger point would raise the level of reserves needed to preserve confidence and thereby sterilize part of the reserve stock, even if not the entire part of it below the “low point.”) The “outer point” should be high enough to avoid any risk of its being breached in the course of normal cyclical management. A country which had previously held enough reserves to avoid any serious risk of breaching the existing “low point” of zero would presumably be equally secure against a symmetrical “outer point.” One rule which might therefore provide a basis for discussion is that the “low point” should be zero and the “outer part” should be twice the base level.

It is possible to get some idea of the potential effects of an outer limit set in this manner by examining the historical experience of the 1960s. Accordingly, “base levels” (in terms of shares of world reserves) for the 45 Fund members with quotas of at least $50 million (at the end of 1965, excluding China) were calculated on the assumptions that these base levels would have been distributed in proportion to (a) Fund quotas at the end of 1965, and (b) their average reserves as a percentage of world reserves over the period 1956–60. The data were then examined to see which countries in fact reached a percentage of world reserves equal to double one or the other of these base levels during the period 1961–70.

Before proceeding to the results of these calculations, three general observations need to be made. First, the calculations presented are only two of a variety of possible approaches. Second, any practical attempt to agree on base levels of reserves for operational purposes would no doubt make provision for certain adjustments to take care of the special situations of certain countries. Third, the statistics should not be taken as implying that these countries would have breached the trigger points if such points had been established: to the extent that the establishment of such points had succeeded in its purposes, it would have promoted adjustment that would have tended to avoid such breaches.

The results of these calculations were as follows. Eleven of the 45 countries would have triggered the outer point based on quotas, and 13 based on average past reserves. Seven countries would fall in both groups. Many of the countries in both groups had reserves that exceeded twice the base level in many years.

Some of the countries listed in the accompanying tabulation cannot by any reasonable test be regarded as hoarders of reserves. Some of them appear in Section B because they had inadequate reserve levels in the base period, which they managed to bring to more normal levels in the 1960s. Any standard which is based on past experience is bound to contain a bias of this sort, and for the same reason may fail to reveal cases of long-term reserve hoarding. Some countries may also appear in Section A because they had abnormally low quotas. These brief comments suggest that considerable attention would have to be paid to the selection of any set of base levels and the adjustments that they might require before such data could be used with confidence for operational purposes.

A. Base Levels Based on Fund Quotas
Reserves in Excess of Twice Base Level
CountryNumber

of years
Years
Austria101961–70
Germany, Fed. Rep. of91961–68, 1970
Israel71962–68
Italy61962, 1965–69
Malaysia21969–70
Nigeria11961
Portugal101961–70
Saudi Arabia31965–67
South Africa11968
Spain31964–65, 1970
Thailand71964–70
B. Base Levels Based on Reserves, 1956–60
Reserves in Excess of Twice Base Level
CountryNumber

of years
Years
Austria11967
Chile21969–70
Denmark31964–66
France101961–70
Israel101961–70
Italy11967
Japan21969–70
Norway41967–70
Saudi Arabia81963–70
South Africa31963, 1968–69
Spain101961–70
Thailand41966–69
Yugoslavia11969

3. Indicators and the need for adjustment

A separate paper examines the question of how well a considerable number of indicators would have performed if they had been employed to indicate the need for par value changes by the industrial countries during the 1960s. The two particular indicators primarily discussed in this paper, reserve levels and basic balances, were found to rank among the better performers. But apart from this, the first of these, and to some extent also the second, may be considered as variables that are of interest in their own right.

Figures on countries’ reserve levels bear a relation to the problem of the distribution of reserves in the system. Hence, if agreement could be reached on a set of outer limits for individual countries’ reserves the transgression of these limits would be indicative of an agreed maldistribution of reserves and a need to seek correction. This would be the case whether or not, upon assessment, the reserve levels were found to be indicative of the need for measures to change current account flows.

The significance of the basic balance lies in the fact that it—in contrast, say, to price indices or exchange rates—is at least a first approximation of the underlying balance of payments position that one would like to be able to identify to judge whether a country was in fundamental equilibrium or not. The imperfections of the basic balance from this point of view, even after certain statistical corrections, have already been set out, and stress has been laid on the need to distinguish between what can be made readily available in terms of statistics and what, conceptually, needs to be taken into account in making policy judgments. Nevertheless, though it needs to be handled with due caution, the basic balances may be regarded as not merely a trigger to assessment but also, in most cases, a starting point for that assessment.

The two indicators here discussed, the reserve level and the basic balance, give valuable information on the two aspects of a country’s international economic position that need to be taken into account in devising policy in the framework of the adjustment process. Neither provides by itself an automatic answer to the question whether adjustment action would be required; nor could any mechanical combination of the two indicators perform that function. Their role in relation to such adjustment action can best be described as suggestive of the need for adjustment and as triggers of a full assessment of a country’s position and prospects, in addition to such regular appraisal of balance of payments situations and prospects as may be provided for. If certain indicators are selected to set off this important process of international consultation and explanation, they are obviously believed to carry a presumption that this process will have a certain likelihood of leading to the conclusion that adjustment action would be indicated; without such a presumption it could hardly be worthwhile to be guided by the indicator in undertaking the exercise. At the same time, the assessment itself would determine in any particular case whether this presumption would be confirmed.

The Division of Adjustment Responsibilities

When a need for adjustment is established, it generally follows that a particular country will be identified as out of line and therefore as responsible for initiating adjustment. Occasionally, however, the restoration of equilibrium will be possible by action on the part of either of two countries or groups of countries, and some basis must then be found for dividing the responsibility for initiating adjustment. The question that arises is whether and to what extent indicators could be helpful in connection with the assignment of adjustment responsibilities.

In general, the contribution that indicators could make would be the same as that related to the detection of the need for action by individual countries. They would suggest cases for assessment, and that assessment would lead to a judgment on the need for adjustment by particular countries, and the magnitude of the adjustment required.

In those situations in which there was no clear answer as to the distribution of the need to take adjustment measures, an external consideration, such as the intended effect on the value of the SDR in terms of currencies in general, would have to be employed.2

The signals generated in different countries by certain disturbances would inevitably depend, among other things, on the relative size of the countries involved. If reserve levels, reserve changes, or basic balances were the basis for the indicator, a flow of a given size from one country to another would signal a larger proportional need for adjustment in the smaller than in the larger country. This is not, of course, a problem that is unique to the use of indicators; it would arise just as much in a full assessment. Insofar as the disequilibria caused were judged to be fundamental (rather than transitory) in character, they would tend to require proportionally more adjustment in the smaller than in the larger country. Any other distribution of the adjustments undertaken would tend to generate new disequilibria involving third countries.

The Disciplinary Role

Even when the need for adjustment has been diagnosed and the responsibilities for initiating corrective measures have been divided under agreed rules, there remains the problem of inducing countries to take the appropriate actions. It is in this connection that the possibility of introducing “sanctions” has been mentioned.3

One view is that any “sanctions” that might be introduced should only be imposed, on a discretionary basis, after the need for adjustment had been established and the country in question had declined to introduce corrective measures deemed to be adequate by the Fund. An alternative approach would be to specify in advance certain critical values of an indicator at which international disciplines would automatically be called into play, so as to provide a clear incentive to countries to avoid disequilibrium situations. Under this approach, there would be a need to select an indicator and its trigger points as well as to determine a set of “sanctions.” The concept of presumptive need for correction by the member could be reflected by the automatic imposition of a “sanction” but with the authority for the Fund to lift it.4

The choice of the “sanctions” to be employed might be influenced by the decision as to whether their imposition was discretionary or automatic. It might be argued that it would be difficult to see how drastic penalties could come into force automatically: the intended deterrence would have to rely on the certainty of retribution rather than its severity. Yet it may also prove difficult in practice to decide to impose “sanctions,” especially severe “sanctions”; from this it might be argued that the likelihood of any “sanctions” being applied would be greatest if they did come into effect automatically.

From an economic point of view, two types of “sanctions” might be distinguished: (1) Measures that were intended by themselves to provide a remedy to the situation that was judged to be unsatisfactory, without necessarily implying that any country was at fault. (2) Measures that were intended, by penalizing a country that was judged to be at fault, to induce it to adopt corrective measures.

A loss of access to credit, or of SDR allocations, would be a “sanction” of the second type for a deficit country. An import surcharge against the exports of a persistent surplus country need not be seen as punitive; it would in any event also be remedial in character. Similarly, the freezing of reserves or even cancellation of SDR allocations when reserves exceeded an agreed “outer point” would be corrective of a stock disequilibrium; in that case a further stock corrective could be the injection by the Fund of an equivalent amount of SDRs to be allocated to other members. One point to be borne in mind in this connection is that such measures to correct stock disequilibria may at the same time help to perpetuate flow disequilibria. Less radical measures to deal with excess reserves, intended more to deter than to correct, would be the cessation of payment on certain SDR balances or their temporary blockage.

General Criteria for Indicators

There are certain general considerations that would be relevant whatever the purpose for which an indicator is selected—apart from the obvious one that it should be as accurate as possible as an indicator of the objective which it is intended to indicate. The first is the desirability of having an unambiguous measure. This argues for a comprehensive measure such as reserves, rather than a measure that relies on drawing a statistical line between different capital flows or that involves making adjustments for cyclical or other transitory disturbances. A second consideration is the desirability of having a measure that is not subject to easy manipulation. It is well known that there are a variety of practices that can affect the figures for official reserves, including the exclusion of certain assets, the encouragement of foreign borrowing by banks or semiofficial institutions, the provision of subsidized forward cover to the commercial banks, the encouragement of local financing of foreign trade, central bank investment in ibrd bonds, etc. If reserves are to be used as an indicator, rules would have to be devised to ensure that the figures used would not be unduly influenced by such practices. Finally, there is the desirability of minimizing the impact of capricious events. Since a stock indicator could only be triggered by a persistent failure to take actions that would terminate a flow, whereas a flow can change suddenly for reasons that are impossible to anticipate, this consideration argues for a stock concept. If a flow indicator was nevertheless to be adopted, the basic balance would have the advantage of excluding short-term capital flows. This consideration would seem less important if a stock indicator is chosen, because short-term capital flows generally cumulate to large sums only to reinforce a previous basic imbalance: but it would still seem preferable on this ground to adopt a measure undistorted by shifts of volatile funds.

A major consideration in this general area would be the degree of precision with which an indicator would be defined in the Fund’s Articles or, more generally, in the rules that defined the role that would be attributed to such an indicator. In this context the consequences of the triggering of the indicator would be of considerable importance. If such consequences were to provoke an assessment, or perhaps to raise the question whether an assessment seemed indicated, it might be possible to forgo any very precise definition. In this case, for example, the concept of “gross reserves” as defined for a particular country in the Fund’s International Financial Statistics might be sufficiently precise. But if an indicator were to trigger the automatic imposition of “sanctions,” a much more precise definition might be required lest disagreements on the facts were to raise serious difficulties in practice. But the precise legal definition of concepts such as reserves is not without its own difficulties and pitfalls, as is testified by the Fund’s experience with the obligation to repurchase under Article V, Section 7(b) and Schedule B.

Conclusion

The present paper has attempted to analyze two indicators that had received considerable support at the November 1972 meeting of the Deputies of the Committee of Twenty: reserve levels and basic balances. In the analysis two major distinctions were made. The first of these was between the flow dimension and the stock dimension of payments objectives, the former associated with the concept of fundamental equilibrium and the latter with the optimal distribution of reserves among countries. The second distinction was between the different roles in which indicators can be used—diagnosis, the division of adjustment responsibility, and the disciplinary role. Both of these distinctions are of major importance in the selection of an indicator, and any indicator that may be useful from one point of view, or in one role, will not necessarily be useful from another point of view or in another role. In further discussions on indicators, greater specificity of the context in which a particular indicator is to be used would, therefore, be helpful.

One of the difficult issues that arises with respect to indicators as a diagnostic device is the extent to which they should serve to initiate a process of adjustment or to influence the outcome of that process. The ability of different indicators to provide accurate warning signals of the need for adjustment was subjected to statistical testing in another paper. The data presented showed that both reserve levels and basic balances were successful in signaling the need for exchange rate adjustment in about half the cases where such adjustment was judged by other criteria to have been necessary. In this respect these two indicators outperformed most other indicators. This would still mean that they missed their target in about twice as many instances, either calling for adjustment where this was not judged necessary or failing to call for it in cases where it was. These findings must be seen against the fact, also noted in the paper referred to, that the actual adjustment performance of the countries was far from satisfactory: adjustments did not occur in about three fourths of the cases where they were judged to have been appropriate, although no adjustments took place that were not called for. The objective clearly should be to improve on this performance by increasing the likelihood that adjustments are made when they are needed with a minimum risk of provoking unneeded adjustments. Indicators could perform a function in this connection if they were used as signals that a serious assessment would be appropriate—beyond such regular appraisal of balance of payments situations and prospects as would take place in any event. This assessment would have to be relied upon to provide the diagnosis as to the magnitude and, in some respects, the nature of the action required, questions on which indicators cannot provide guidance; the assessment would thus also determine in any particular case whether the presumption inherent in the indicator was confirmed.

So far as the problem of securing agreement on the division of adjustment responsibilities is concerned, not much additional help can probably be expected than that of signaling the situations that require assessment. When a need for adjustment is established, it generally follows that a particular country will be identified as out of line and therefore as responsible for initiating adjustment. The most significant need to handle other situations is a consensus on the principles that should underlie the distribution of adjustment responsibilities.

As to the disciplinary role, one major question is whether any “sanctions” that may be felt necessary should be imposed as a penalty for failure to follow appropriate adjustment policies or as a consequence of an indicator reaching a certain trigger point. A further area for decision would then be the choice of “sanctions,” including that between corrective and punitive “sanctions.”

If basic balances in fact provided a good measure of the persistent elements in payments flows, one would expect the positive and negative differences between reserve changes and cumulative basic balances to cancel out over time, so that the cumulative difference between the two series over a long span of years would be small. For some countries, however, such as Italy and Switzerland, the cumulative gap in fact grew steadily during the 1960s and reached a very substantial amount by the end of the decade. See Zoran Hodjera, “Basic Balances, Short-Term Capital Flows, and International Reserves of Industrial Countries,” Staff Papers, International Monetary Fund (Washington), Vol. 16 (November 1969), pp. 582–612.

See Reform of the International Monetary System A Report by the Executive Directors to the Board of Governors [reproduced above, pp. 19–56].

The term “sanctions” is used in this section between quotation marks to indicate that it is used in a sense that is somewhat broader than the conventional one. A common understanding of sanctions is that they are measures to ensure the observance of obligations or to punish the violation of them. In the present context, however, a closer approach to a useful definition is that sanctions are those measures that are applied without the need for the consent of the member against which the measure is directed and are regarded by members as censorious. Some of these measures may be applied without the necessity for any finding that a member has neglected its obligations. (See Joseph Gold, “The ‘Sanctions’ of the International Monetary Fund,” American Journal of International Law (Washington), Vol. 66 (1972), pp. 737–62.)

Cf. the provision of Article IV, Section 6, which entails the automatic ineligibility to use the Fund’s resources of a member that changes the par value of its currency despite the objection of the Fund, “unless the Fund determines otherwise.”

    Other Resources Citing This Publication