Factors Influencing Balance of Payments

The primary determinants of the current account of a country’s balance of payments are (1) relative demand levels in the domestic economy and in its trading partners, and (2) relative competitiveness. This much is not controversial and provides a useful framework for analyzing balance of payments developments—though it has to be recognized that the relevant relationships cannot be established with a high degree of precision. The recent and prospective evolution of the trade and current accounts can be assessed with reference to past and anticipated developments in demand and competitiveness.¹⁰⁰ If forecasts are available for rates of domestic demand growth in the major countries over the relevant time horizon, estimates can be made of how the current account might evolve (other things held equal). Similarly, given an assumed evolution of domestic costs and prices and the pattern of exchange rates, estimates can be made of how these competitiveness factors will affect payments flows.’¹⁰¹ By adjusting the existing payments position for the estimated effects of lagged changes in exchange rates and of prospective changes in cyclical positions, it is thus possible to come to an assessment of the “underlying” balance of payments. The key indicators needed to make such an assessment are as follows: (1) a measure of economic activity (demand or output); (2) a measure of domestic inflation (costs or prices); and (3) a measure of the effective exchange rate.

The first two of these indicators can be projected on a country-by-country basis using standard forecasting techniques. The third (the exchange rate) can be easily observed ex post, but has proved difficult to project in any satisfactory manner. It is nevertheless of considerable interest to develop a framework for analyzing exchange rate developments and prospects, especially if judgments are required concerning whether a given pattern of exchange rates is to be regarded as sustainable or desirable. To do this, it is necessary to understand (and try to develop indicators for) the factors that underlie shifts in capital flows so as to provide a more comprehensive account of international economic interactions. This is a difficult exercise, in part because expectations (which are inherently hard to observe or model) play such an important role in determining the desire to acquire or dispose of external assets. A logical place to begin is by looking at the factors that influence the balance of domestic savings and investment (and therefore the net acquisition of foreign assets). These factors include developments in the fiscal position (which represents the net saving or dissaving of the public sector) as well as developments affecting the willingness of the private sector to save or invest. Indicators that may be useful in this context include: (1) a measure of the overall fiscal position; (2) gross private savings flows; (3) gross private investment; and (4) real interest rates.

Criteria for Assessing Sustainability of Payments Balances

The calculation, as described above, of a set of “underlying current account balances” that is implied by existing policies and exchange rates can be a helpful focus in international surveillance. It can facilitate a dialogue aimed at assessing whether the projected trends can be considered sustainable or not.

A key issue in interpreting underlying payments balances is thus how to establish criteria for what should be considered sustainable. This is a particularly difficult analytical subject, on which it will undoubtedly be necessary to proceed gradually. A basis for approaching the issue is provided by the identity which equates the current account position (surplus or deficit) and the balance between domestic saving and investment. A sustainable current account position could therefore be defined as one in which the domestic savings and investment position of a country is sustainable, given the corresponding preferences of other countries, and the need to avoid an excessive buildup of external liabilities or assets.

The balance between domestic saving and investment can in turn be decomposed into the net financial balance of the public sector and the net financial balance of the private sector. The former is derived from the fiscal balance, which is an important indicator and policy tool in its own right. The latter is influenced by all the factors that affect private saving and investment decisions, including interest rates, the level and growth rate of real incomes, the return on physical capital, and demographic factors. The determinants of gross private saving and investment are subject to empirical estimation, although it has to be recognized that robust relationships are not always easy to establish.

The foregoing analysis suggests that the appraisal of policy interactions among industrial countries could be based on an analytical framework in which “underlying” current account positions (based on existing policies and exchange rates) would be compared with “sustainable” positions (derived from an assessment of the medium-run determinants of savings and investment). It is to be emphasized that international consistency of balance of payments positions is a necessary condition for their sustainability.¹⁰²

It is clear that the prospective evolution of domestic variables is important in appraising both the underlying and the sustainable payments positions. Trends in GNP growth, domestic demand, and relative prices will affect movements in actual current account positions, while policies and developments that influence savings and investment propensities have an important bearing on the payments position that is sustainable over the longer term.

The foregoing implies that the setting and monitoring of objectives with respect to domestic variables is a matter of international concern, at least to some extent. The choice of whether a balance of payments disequilibrium should be corrected by exchange rate movement or a shift in relative rates of economic growth requires, for example, a view on the rate of growth in an individual country that should be considered attainable.

Indicators of Economic Performance

The balance of payments. This variable can be considered either an objective of policy or an intermediate variable. For the major currencies with floating exchange rates, it is not an objective in the sense that these countries have quantified aims for the structure of their balance of payments. However, most countries would probably subscribe to the objective of restoring or maintaining a “sustainable” external payments structure, so as to limit the dangers of protectionist pressures and to minimize the costs and uncertainties that are involved when an unsustainable position emerges and has to be corrected. Moreover, from the point of view of surveillance, the restoration and maintenance of a balance of payments pattern that is adequately consistent with the domestic policies and priorities of all members must be considered a key objective.

A widely used indicator related to the balance of payments is the current account surplus or deficit, and it seems appropriate that the current account balance remain the primary indicator of developments in the external sector.¹⁰³ As noted in the previous section, however, a satisfactory analytical framework for judging the sustainability of a given exchange rate pattern would involve an assessment of the “underlying,” as well as the actual, current account position. This would involve making adjustments for the effects of recent exchange rate changes that had not yet been fully reflected in trade flows, for the impact on imports and exports of cyclical divergences from “normal” employment levels, and for any other special factors affecting payments flows in a given period.

Presentation of underlying balance of payments estimates would facilitate appraisal of the sustainability of the external positions implied by current policies and prospects. A sustainable balance of payments position can be defined as one in which the underlying current account surplus or deficit is matched by capital outflows or inflows that correspond to a country’s desire to accumulate foreign assets or debts, and its capacity to service its external debt out of current foreign exchange earnings. For such a position to be internationally appropriate, it must also be compatible with the savings/investment preferences of other countries, and reasonably full employment of factors of production.

Real output. Perhaps the most widely used indicator of domestic economic performance is the rate of growth of GNP. As an indicator, GNP has the merit of comprehensiveness, widespread familiarity, and comparability across countries. Given the multiplicity of purposes for which it is employed, however, it should not be surprising that it is not equally suited for all purposes. As a measure of welfare, absorption per capita would be a better indicator, while as a measure of efficiency, output per unit of factor input might be superior.

The chief drawback of the GNP indicator for analyzing the international interaction of trends in output and demand, which is the natural focus of multilateral surveillance, is its level of aggregation. In particular, it is often desirable to distinguish the relative contribution of domestic and foreign sources of demand growth to any given change in overall GNP. So while GNP should remain the principal indicator of developments in the real economy, it needs to be supplemented with systematic presentation of developments and prospects in final domestic demand.

A more difficult issue than the choice of which indicator to use is the establishment of criteria by which objectives with respect to growth are to be judged, and performance assessed. Specifically, what rate of growth should be considered sustainable, given the constraints a country faces, its other objectives, and its obligations toward its trading partners? Such a calculation will involve, as a first step, estimating the underlying rate of growth of productive potential. This is not simple to estimate, since it depends not only on the rate of increase in available factors of production (the aggregate supply of labor and capital) but also on hard-to-observe variables such as the quality of factor inputs and the speed of technological progress. A second step would be to judge how large is the gap between existing and potential output levels and how quickly it is feasible to close such a gap. These are also difficult estimates to make, since they depend on factors such as the nature and extent of rigidities in goods and labor markets, the risk of igniting inflationary pressures, and the existence of other objectives (such as fiscal strengthening) that might legitimately constrain governments’ freedom of maneuver. However, techniques for making judgments in these areas exist that enable the attainable rate of growth to be defined as the sum of (1) the growth in underlying capacity and (2) the rate of absorption of economic slack.

Employment. This indicator of economic performance is closely related to GNP growth. There are a variety of possible indicators in the employment field. The rate of increase in numbers employed is sometimes used as an indicator of “success in creating jobs.” By this token, however, an economy with a rapidly growing population and labor force could appear to be more successful in the employment field than one with a slower growing population, even though the latter might have a lower rate of involuntary unemployment. A measure of labor market conditions that avoids this particular difficulty is the rate of unemployment, which is also in many respects a more visible objective for government policy. However, the rate of unemployment is itself not always an effective measure of involuntary unemployment. There may be categories of discouraged jobseekers (longer-term unemployed, young people remaining in education, women remaining in household work) not captured in published unemployment statistics.

As with GNP growth, a labor market indicator such as the unemployment rate needs a standard by which assessments can be made as to whether changes in the unemployment rate are satisfactory or not. A considerable literature exists on the “natural” rate of unemployment or the “nonaccelerating inflation rate of unemployment” (NAIRU). Although neither of these concepts is easy to apply in practice, they do provide a framework in which, for a given institutional setting, the existing unemployment rate can be judged too high (or, possibly, too low). A medium-term objective of policy would presumably be to move toward the NAIRU, and to do so at a pace that does not have seriously adverse consequences for other economic objectives, such as inflation control. It is also possible that governments may have medium-term objectives to reduce their NAIRU, say by structural measures that improve the flexibility of labor markets, and enable price stability to be maintained with a lower level of joblessness. To the extent that such structural goals can be quantified, it should be possible to factor them into medium-term employment objectives.

Inflation. This is another important indicator of economic performance. From a domestic standpoint, the key objective in most countries is probably to come as near as practically feasible to stability in consumer prices. Internationally, it may be more important to identify differentials in cost inflation (and where possible, reduce them), so as to facilitate the task of judging whether movements in nominal exchange rates are appropriate from the perspective of efficient adjustment. While costs and prices often move together, there can be differences in underlying trends. In an economy that has a more rapid rate of productivity increase in its manufacturing than in its service sector (as was the case, for example, in Japan in much of the postwar period), domestic price inflation could be higher than in its trading partners, while cost inflation in traded goods industries might be lower. This need not pose insuperable difficulties for the development and use of indicators, provided that the disparities are properly identified and allowed for.

Efficient adjustment also requires that any trend divergence in production costs in traded goods industries should be compatible with corresponding exchange rate trends. An inflation measure that has proved particularly useful in analysis is the rate of change of unit labor costs in those sectors of the economy that are most exposed to international competition. Being a measure of costs, it is more relevant to international competitiveness, and thus is better able to identify inflationary trends that have implications for balance of payments flows.

The foregoing discussion suggests that the inflation indicator that would be most useful in discussing external imbalances, exchange rate developments, and policy interactions is not necessarily the one that will be most familiar in a domestic economic context. Movements in the consumer price index contain adventitious elements that are only incidentally related to international competitiveness and underlying inflationary pressures. The GNP deflator is a superior indicator in many respects, although it still does not capture some important elements related to international competitiveness. A measure of labor costs per unit of output, perhaps normalized for differences among countries in cyclical position, would have important advantages. Its main drawbacks would be: (1) such an index is available only for the manufacturing sector of the economy; (2) the data from which it is compiled are produced with a lag; and (3) the concept is less familiar to policymakers. To some extent, however, it might be possible to reduce these shortcomings if it were decided to focus on such a measure as a central feature of surveillance.

Indicators of Economic Policy

For expositional convenience, domestic macroeconomic policy can be divided into its monetary and fiscal aspects. In addition, countries that do not maintain either fully floating or rigidly fixed exchange rates have discretion in their management of reserves or exchange rate policies. Finally, the increased focus in recent years on structural policies suggests that it may be desirable to consider the possibilities of developing indicators in this field also.

Monetary policy. The most widely used indicator of monetary policy in the major industrial countries is the rate of growth of some monetary or credit aggregate. The specific aggregate that is used varies from country to country depending on the institutional characteristics of the country concerned and the robustness of empirical relationships.

Several issues arise in developing meaningful indicators for use in cross-country analysis. The first concerns whether the same measure of the money stock should be employed for all countries. On the one hand, it could be said that different definitions of monetary variables would inhibit comparability across countries. Some monetary authorities have selected a monetary target primarily on grounds of controllability; while others have preferred to target an aggregate that is closely linked with developments in the economy, even though the target itself cannot be closely controlled. Differences in criteria for the selection of monetary targets might cause difficulties when the monetary authorities were attempting to concert their policy stance—if, for example, they were trying to keep aggregate growth in the world money supply within some range, or to engage in offsetting policy responses to unwanted exchange rate developments.

Nevertheless, it would seem desirable to adapt the definition of monetary indicators to take account of conditions in particular countries. Generally speaking, when the authorities of a particular country have chosen a monetary aggregate for policy or monitoring purposes, this represents a careful choice based on analysis of the strength and stability of empirical relationships. It is likely to facilitate a meaningful dialogue, without undue loss of effective comparability, if the definitions used for purposes of international surveillance accord with those used for domestic policy formulation.

A second issue in the choice of monetary indicators concerns whether real or nominal variables should be used. It can be argued that it is the real money stock that determines the perceived liquidity of the private sector and therefore influences the willingness to spend. On the other hand, developments in the real money stock can be an ambiguous indicator. Under conditions of rising inflation, wealth-owners will seek to economize on real money balances and the real money stock will be observed to fall. Indeed, a systematic increase in the growth of the nominal money stock, causing an acceleration of inflation, will almost certainly lead to a decline in the real money stock. For this reason, and because the central goal of monetary policy is to keep inflation in check, it seems more appropriate to define monetary indicators (whether used as targets or as monitoring instruments) in nominal terms.

A third potential issue is how to handle unexpected shifts in the demand to hold money balances. If an upward shift in money demand can be identified, measured, and precisely offset by an equivalent adjustment in supply, it could be said that the stance of monetary policy has been kept unchanged. It would therefore be desirable to have an indicator that is adjusted for the effect of known shifts in the demand for money. The difficulty is, of course, that changes in demand are very hard to identify, particularly at the time they are occurring. One possibility would be to use interest rates as an indicator of a change in the balance between supply and demand in money and capital markets. But interest rates are also an ambiguous criterion, since an increase in interest rates may reflect a shift in the demand for money at given income levels, or an increase in credit caused by an incipient increase in nominal output. The implication of the foregoing is that the possible alternatives to the use of money growth rates as a primary indicator of the stance of policy all have drawbacks. A money stock indicator is still probably the best indicator of monetary policy in countries that use monetary targets, but it must be used with caution and in the light of surrounding developments.

Not all countries use monetary targets, of course. In countries with fixed exchange rates, the domestic money stock cannot be closely controlled by the authorities, since liquidity can enter or leave the country via overall balance of payments surpluses or deficits. Even in countries where there is greater freedom for exchange rate movement, monetary policy may be managed in the light of exchange rate and interest rate developments rather than to achieve target monetary objectives. For countries that do not have objectives for monetary aggregates, an alternative indicator of changes in monetary conditions will be necessary. One possibility, which would have the merit of facilitating international comparisons, would be to use interest rate differentials with major international currencies.

Fiscal policy. While the medium-term goal of monetary policy can be defined as the restoration and maintenance of an appropriate degree of price stability, the aims of fiscal policy are both macroeconomic and structural. At the macroeconomic level, governments have objectives for the budget deficit, related to the need for economic stabilization, as well as to a desire to limit the government’s claims on the saving available for private investment. At the structural level, there may also be objectives for the structure and level of taxation and expenditure, with a view to enhancing incentives for efficient resource allocation, and limiting the absolute volume of real resources absorbed by the government.

In choosing how the macroeconomic objective relating to fiscal balance should be defined, a number of questions arise. These include: whether fiscal objectives should be established for the entire public sector, the general government, or for the central government alone; whether the objective should be for the actual fiscal deficit or for the “underlying” deficit (that is, whether the “policy-induced” fiscal change should be separated from the change attributable to purely cyclical factors); and whether or not the deficit that is monitored should be inflation corrected. Beyond these definitional issues, of course, lies the more fundamental question of what the fiscal objective should be, and whether independently determined objectives (such as zero deficits) are internationally consistent in terms of their impact on balance of payments flows.

The choice of whether the fiscal deficit that is measured should extend beyond the central government depends, to a considerable extent, on the scope of public sector activities that fall under the control of the fiscal authorities. Whichever level of government is selected as the primary focus for the fiscal indicator, however, it is important that the value of the indicator not be undermined by shifts in the classification of transactions (for example, from central to local government) that have substantial effects on the chosen indicator but little economic significance.

Concerning the choice of actual or cyclically corrected fiscal deficits, it is relevant to note that most governments express their aggregate fiscal objective in terms of some stated level for the actual fiscal deficit. This, indeed, is the indicator that most accurately measures the extent of the government sector’s claims on financial markets. However, movements in the actual fiscal deficit can give a misleading impression of the thrust of fiscal policy when output is growing significantly faster or slower than its medium-term trend. In a phase of relatively rapid cyclical expansion, revenues tend to grow faster than expenditures and thus the budget deficit can shrink, even though policy-related fiscal measures may be tending to increase the deficit. Similarly, when the economy is stagnating, weak tax revenues may cause a budget deficit to widen even though the measures introduced by the authorities may in themselves be tending to strengthen the fiscal position. It seems desirable therefore to use indicators both of the actual and of the cyclically adjusted fiscal position.¹⁰⁴

Adjustment for inflation poses some potentially more difficult issues. On the one hand, it can be argued that inflation effectively reduces the value of a government’s outstanding debt with effects on wealth-holders that are similar to those of taxation. On the other hand, inflation adjustment can sometimes define away a real problem, as when the fiscal deficit itself is an important source of the pressures that give rise to inflation. In circumstances when inflation is low and stable, inflation adjustment is not needed to permit a meaningful comparison of fiscal conditions across countries or over time.

A more difficult issue than defining the fiscal deficit to be monitored is that of evolving criteria for assessing the appropriateness of countries’ medium-term objectives in the fiscal field. While a country’s objective for its fiscal deficit is intimately related to its domestic social and political priorities, it also has international implications. Changes in public sector saving or dissaving have consequences for the overall saving/investment balance in a country, and thus for its balance of payments situation. This is an area in which technical criteria are not easy to establish, and in which, therefore, it will be particularly useful to generate a multilateral dialogue.

Exchange market policies. Possible indicators in this area include the exchange rate and some measure of exchange market intervention. Since most of the major industrial countries pursue flexible exchange rate policies, the exchange rate itself is not regarded as a proximate instrument of policy. It is better viewed as an intermediate variable, and as such it is discussed further in the following section. Exchange market intervention, on the other hand, is a policy instrument that can be measured by the size of reserve movements over given intervals. It should be remembered, however, that exchange market intervention can sometimes take place in ways that do not affect reserves (for instance through borrowing and lending in foreign exchange by public sector entities). Moreover, the impact of intervention depends importantly on whether it is sterilized or nonsterilized. Concerning the criteria according to which exchange market intervention might be assessed, it seems reasonable to assume that reserve levels would, over time, tend to move toward some stable relation to external transactions (say, imports). For most major countries, it is unlikely that there would be any systematic ex ante intention to accumulate reserves, or run them down, in significant quantities. There would therefore be no prior standard (other than “no change”) against which actual reserve movements would be compared. It would, however, be useful to compare intervention activities by one country with any offsetting activity on the part of trading partners; and to view exchange market intervention in the light of accompanying movements in exchange rates and interest rates.

Structural policies. Indicators of these policies are, by their nature, hard to devise. The structural policies that have been the focus of most attention in recent years have been those relating to deregulation, labor market rigidities, and trade restrictions. With regard to deregulation, it is possible to list the number of regulations eliminated or modified, and it is possible to provide analytical judgments about the effects of deregulation (in terms of reduced prices, for example, or an increased volume of transactions in the markets concerned). On the whole, it seems more practical to analyze the process of deregulation through ad hoc empirical studies, rather than through an attempt to devise specific and quantified indicators of objectives and performance. Labor market rigidities are thought to be manifested in a variety of ways, including inadequate flexibility of wages; wage levels that are too high; lack of geographical and occupational mobility; and inadequate training facilities. While many of these factors are not amenable to quantitative measurement through indicators, it does seem desirable to use the indicator of unit labor costs described earlier to help in reaching judgments of whether developments in real wage rates are warranted by movements in labor productivity. As far as trade restrictions and protectionism are concerned, there is again no fully satisfactory way of developing quantitative indicators. There is thus little alternative to continuing to analyze the impact of trade restrictions in qualitative (albeit specific) terms, while to the extent possible including judgments of their quantitative significance.

Indicators of Intermediate Variables

The channels by which policy variables affect the ultimate objectives of policy are not direct, nor are they fully predictable. For these reasons, it can be of value to develop indicators of those intermediate variables through which policy works to influence more fundamental objectives. Intermediate variables are not, in general, controlled directly by the policy authorities. However, they can be used to check whether the behavioral assumptions underlying the formulation of policy are an adequate representation of reality, and whether economic developments are following their anticipated path during the interval before measures have their effect on ultimate objectives. Intermediate variables can also be used to identify emerging problems of international consistency of policies.

Interest rates and exchange rates. An important channel by which policies influence ultimate economic objectives is through their effect on conditions in financial markets. In this connection, key roles are played by money and capital markets and the foreign exchange market. It is therefore of considerable interest to monitor developments in interest rates and exchange rates.

With regard to interest rates, a major determinant of individual saving and investment decisions is probably the level of the real interest rate. Notwithstanding the fact that inflationary expectations can only be measured indirectly, it nevertheless seems desirable to use some estimate of the real interest rate as the primary indicator for monitoring purposes. A helpful approximation in this connection, that is generally not seriously misleading, is to deduct from the nominal interest rate the rate of change in the GNP deflator over some recent period. While the level of real interest rates is of importance as an indicator of the incentive to save and invest, it is less significant than interest differentials in determining incentives to capital movements, and therefore exchange rate pressures. In monitoring real interest rates, it is therefore necessary to pay attention to international differentials in rates as well as to absolute levels.

As far as exchange rates are concerned, the indicator that is most relevant for purposes of international competitiveness and adjustment is the real effective exchange rate. This can be obtained by combining a measure of the nominal effective exchange rate (using currency weights derived from, say, the Fund’s Multilateral Exchange Rate Model)¹⁰⁵ with a relative inflation estimate from the inflation indicator described above.

Since both interest rates and exchange rates are intermediate variables, it would be unrealistic to expect policy authorities to prescribe in advance any precise path for their expected evolution. Nevertheless, to the extent that the analysis of other indicators reveals underlying disequilibria in foreign exchange and capital markets, this analysis could signal the direction of possible changes in interest and exchange rates.

Saving and investment balances. As already implied, movements in interest rates and exchange rates are in turn influenced by underlying shifts in domestic saving and investment. Indeed, it is movements in interest and exchange rates that give causal content to the identity that makes the current account of the balance of payments equal to domestic savings minus domestic investment.

Any ex ante inconsistency at the global level between countries’ balance of payments objectives or forecasts must be reflected in a similar global inconsistency between projected saving and investment trends. It may therefore be useful to keep track of actual or expected trends in savings and investment, in order to provide advance warning of possible inconsistencies. As noted earlier, the aggregate saving/investment balance of a country can be divided into the balance attributable to the public sector and that attributable to the private sector. The financial position of the public sector is, of course, the fiscal balance and has been discussed above. This can be complemented, for analytical purposes, with supplemental indices of saving and investment in the private sector.